Adding P4D2_2018_East Folder (#116)

* Copying P4D2 Fall 2017 into P4D2 2018 East.

* Updated P4D2_2018_East VM.  Added vagrant URL workaround, cdrom to VM.  Updated to latest commits of BMV2, p4c, PI.  Known issue with p4runtime exercise.

* Applied patch from @antoninbas in  and updated solution
This commit is contained in:
robh2
2018-03-04 03:50:22 -05:00
committed by Nate Foster
parent 89277b5542
commit 97f31560ca
103 changed files with 8930 additions and 0 deletions

View File

@@ -0,0 +1,49 @@
# P4 Tutorial
## Introduction
Welcome to the P4 Tutorial!
We've prepared a set of exercises to help you get started with P4
programming, organized into four modules:
1. Introduction and Language Basics
* [Basic Forwarding](./basic)
* [Basic Tunneling](./basic_tunnel)
2. P4 Runtime and the Control Plane
* [P4 Runtime](./p4runtime)
3. Monitoring and Debugging
* [Explicit Congestion Notification](./ecn)
* [Multi-Hop Route Inspection](./mri)
4. Advanced Data Structures
* [Source Routing](./source_routing)
* [Calculator](./calc)
5. Dynamic Behavior
* [Load Balancing](./load_balance)
## Obtaining required software
If you are starting this tutorial at the Fall 2017 P4 Developer Day, then we've already
provided you with a virtual machine that has all of the required
software installed.
Otherwise, to complete the exercises, you will need to either build a
virtual machine or install several dependencies.
To build the virtual machine:
- Install [Vagrant](https://vagrantup.com) and [VirtualBox](https://virtualbox.org)
- `cd vm`
- `vagrant up`
- Log in with username `p4` and password `p4` and issue the command `sudo shutdown -r now`
- When the machine reboots, you should have a graphical desktop machine with the required
software pre-installed.
To install dependencies by hand, please reference the [vm](../vm) installation scripts.
They contain the dependencies, versions, and installation procedure.
You can run them directly on an Ubuntu 16.04 machine:
- `sudo ./root-bootstrap.sh`
- `sudo ./user-bootstrap.sh`

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,173 @@
# Implementing Basic Forwarding
## Introduction
The objective of this exercise is to write a P4 program that
implements basic forwarding. To keep things simple, we will just
implement forwarding for IPv4.
With IPv4 forwarding, the switch must perform the following actions
for every packet: (i) update the source and destination MAC addresses,
(ii) decrement the time-to-live (TTL) in the IP header, and (iii)
forward the packet out the appropriate port.
Your switch will have a single table, which the control plane will
populate with static rules. Each rule will map an IP address to the
MAC address and output port for the next hop. We have already defined
the control plane rules, so you only need to implement the data plane
logic of your P4 program.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the
> reference.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`basic.p4`, which initially drops all packets. Your job will be to
extend this skeleton program to properly forward IPv4 packets.
Before that, let's compile the incomplete `basic.p4` and bring
up a switch in Mininet to test its behavior.
1. In your shell, run:
```bash
make run
```
This will:
* compile `basic.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`)
configured in a triangle, each connected to one host (`h1`, `h2`,
and `h3`).
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, and `10.0.3.3`.
2. You should now see a Mininet command prompt. Open two terminals
for `h1` and `h2`, respectively:
```bash
mininet> xterm h1 h2
```
3. Each host includes a small Python-based messaging client and
server. In `h2`'s xterm, start the server:
```bash
./receive.py
```
4. In `h1`'s xterm, send a message to `h2`:
```bash
./send.py 10.0.2.2 "P4 is cool"
```
The message will not be received.
5. Type `exit` to leave each xterm and the Mininet command line.
Then, to stop mininet:
```bash
make stop
```
And to delete all pcaps, build files, and logs:
```bash
make clean
```
The message was not received because each switch is programmed
according to `basic.p4`, which drops all packets on arrival.
Your job is to extend this file so it forwards packets.
### A note about the control plane
A P4 program defines a packet-processing pipeline, but the rules
within each table are inserted by the control plane. When a rule
matches a packet, its action is invoked with parameters supplied by
the control plane as part of the rule.
In this exercise, we have already implemented the the control plane
logic for you. As part of bringing up the Mininet instance, the
`make run` command will install packet-processing rules in the tables of
each switch. These are defined in the `sX-commands.txt` files, where
`X` corresponds to the switch number.
**Important:** A P4 program also defines the interface between the
switch pipeline and control plane. The commands in the files
`sX-commands.txt` refer to specific tables, keys, and actions by name,
and any changes in the P4 program that add or rename tables, keys, or
actions will need to be reflected in these command files.
## Step 2: Implement L3 forwarding
The `basic.p4` file contains a skeleton P4 program with key pieces of
logic replaced by `TODO` comments. Your implementation should follow
the structure given in this file---replace each `TODO` with logic
implementing the missing piece.
A complete `basic.p4` will contain the following components:
1. Header type definitions for Ethernet (`ethernet_t`) and IPv4 (`ipv4_t`).
2. **TODO:** Parsers for Ethernet and IPv4 that populate `ethernet_t` and `ipv4_t` fields.
3. An action to drop a packet, using `mark_to_drop()`.
4. **TODO:** An action (called `ipv4_forward`) that:
1. Sets the egress port for the next hop.
2. Updates the ethernet destination address with the address of the next hop.
3. Updates the ethernet source address with the address of the switch.
4. Decrements the TTL.
5. **TODO:** A control that:
1. Defines a table that will read an IPv4 destination address, and
invoke either `drop` or `ipv4_forward`.
2. An `apply` block that applies the table.
6. **TODO:** A deparser that selects the order
in which fields inserted into the outgoing packet.
7. A `package` instantiation supplied with the parser, control, and deparser.
> In general, a package also requires instances of checksum verification
> and recomputation controls. These are not necessary for this tutorial
> and are replaced with instantiations of empty controls.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, your message from
`h1` should be delivered to `h2`.
### Food for thought
The "test suite" for your solution---sending a message from `h1` to
`h2`---is not very robust. What else should you test to be confident
of your implementation?
> Although the Python `scapy` library is outside the scope of this tutorial,
> it can be used to generate packets for testing. The `send.py` file shows how
> to use it.
Other questions to consider:
- How would you enhance your program to support next hops?
- Is this program enough to replace a router? What's missing?
### Troubleshooting
There are several problems that might manifest as you develop your program:
1. `basic.p4` might fail to compile. In this case, `make run` will
report the error emitted from the compiler and halt.
2. `basic.p4` might compile but fail to support the control plane
rules in the `s1-commands.txt` through `s3-command.txt` files that
`make run` tries to install using the Bmv2 CLI. In this case, `make run`
will log the CLI tool output in the `logs` directory. Use these error
messages to fix your `basic.p4` implementation.
3. `basic.p4` might compile, and the control plane rules might be
installed, but the switch might not process packets in the desired
way. The `/tmp/p4s.<switch-name>.log` files contain detailed logs
that describing how each switch processes each packet. The output is
detailed and can help pinpoint logic errors in your implementation.
#### Cleaning up Mininet
In the latter two cases above, `make run` may leave a Mininet instance
running in the background. Use the following command to clean up
these instances:
```bash
make stop
```
## Next Steps
Congratulations, your implementation works! In the next exercise we
will build on top of this and add support for a basic tunneling
protocol: [basic_tunnel](../basic_tunnel)!

View File

@@ -0,0 +1,162 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_IPV4 = 0x800;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
/* TODO: add parser logic */
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
/* TODO: fill out code in action body */
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
/* TODO: fix ingress control logic
* - ipv4_lpm should be applied only when IPv4 header is valid
*/
ipv4_lpm.apply();
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
/* TODO: add deparser logic */
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,53 @@
#!/usr/bin/env python
import sys
import struct
import os
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import ShortField, IntField, LongField, BitField, FieldListField, FieldLenField
from scapy.all import IP, TCP, UDP, Raw
from scapy.layers.inet import _IPOption_HDR
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class IPOption_MRI(IPOption):
name = "MRI"
option = 31
fields_desc = [ _IPOption_HDR,
FieldLenField("length", None, fmt="B",
length_of="swids",
adjust=lambda pkt,l:l+4),
ShortField("count", 0),
FieldListField("swids",
[],
IntField("", 0),
length_from=lambda pkt:pkt.count*4) ]
def handle_pkt(pkt):
if TCP in pkt and pkt[TCP].dport == 1234:
print "got a packet"
pkt.show2()
# hexdump(pkt)
sys.stdout.flush()
def main():
ifaces = filter(lambda i: 'eth' in i, os.listdir('/sys/class/net/'))
iface = ifaces[0]
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,4 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:00:01:01 1
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:02:02:00 2
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:03:03:00 3

View File

@@ -0,0 +1,4 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:01:02:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:00:02:02 1
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:03:03:00 3

View File

@@ -0,0 +1,4 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:01:03:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:02:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:00:03:03 1

View File

@@ -0,0 +1,41 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
from scapy.all import sendp, send, get_if_list, get_if_hwaddr
from scapy.all import Packet
from scapy.all import Ether, IP, UDP, TCP
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def main():
if len(sys.argv)<3:
print 'pass 2 arguments: <destination> "<message>"'
exit(1)
addr = socket.gethostbyname(sys.argv[1])
iface = get_if()
print "sending on interface %s to %s" % (iface, str(addr))
pkt = Ether(src=get_if_hwaddr(iface), dst='ff:ff:ff:ff:ff:ff')
pkt = pkt /IP(dst=addr) / TCP(dport=1234, sport=random.randint(49152,65535)) / sys.argv[2]
pkt.show2()
sendp(pkt, iface=iface, verbose=False)
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,176 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_IPV4 = 0x800;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,16 @@
{
"hosts": [
"h1",
"h2",
"h3"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["s1", "s2"], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,166 @@
# Implementing Basic Tunneling
## Introduction
In this exercise, we will add support for a basic tunneling protocol
to the IP router that you completed in the previous assignment. To do so,
we will define a new header type to encapsulate the IP packet and
modify the switch to perform routing using our new header.
The new header type will contain a protocol ID, which indicates the type
of packet being encapsulated, along with a destination ID to be used for
routing.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the
> reference.
## Step 1: Run the (incomplete) starter code
The starter code for this assignment is in a file called `basic_tunnel.p4`
and is simply the solution to the IP router from the previous exercise.
Let's first compile this code to and send a packet between two end hosts
to ensure that the IP routing is working as expected.
1. In your shell, run:
```bash
make run
```
This will:
* compile `basic_tunnel.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`)
configured in a triangle, each connected to one host (`h1`, `h2`,
and `h3`).
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, and `10.0.3.3`.
2. You should now see a Mininet command prompt. Open two terminals
for `h1` and `h2`, respectively:
```bash
mininet> xterm h1 h2
```
3. Each host includes a small Python-based messaging client and
server. In `h2`'s xterm, start the server:
```bash
./receive.py
```
4. In `h1`'s xterm, send a message to `h2`:
```bash
./send.py 10.0.2.2 "P4 is cool"
```
The packet should be received at `h2`. If you examine the received
packet you should see that is consists of an Ethernet header, an IP
header, a TCP header, and the message. If you change the destination IP address
(e.g. try to send to `10.0.3.3`) then the message should not be
received by h2.
5. Type `exit` or `Ctrl-D` to leave each xterm and the Mininet command line.
Each switch is forwarding based on the destination IP address. Your
job is to change the switch functionality so that they instead decide
the destination port using our new tunnel header.
### A note about the control plane
A P4 program defines a packet-processing pipeline, but the rules
within each table are inserted by the control plane. When a rule
matches a packet, its action is invoked with parameters supplied by
the control plane as part of the rule.
For this exercise, we have already added the necessary static control
plane entries. As part of bringing up the Mininet instance, the
`make run` command will install packet-processing rules in the tables
of each switch. These are defined in the `sX-commands.txt` files,
where `X` corresponds to the switch number.
**Important:** A P4 program also defines the interface between the
switch pipeline and control plane. The commands in the files
`sX-commands.txt` refer to specific tables, keys, and actions by name,
and any changes in the P4 program that add or rename tables, keys, or
actions will need to be reflected in these command files.
## Step 2: Implement Basic Tunneling
The `basic_tunnel.p4` file contains an implementation of a basic IP router.
It also contains comments marked with `TODO` which indicate the functionality
that you need to implement. A complete implementation of the `basic_tunnel.p4`
switch will be able to forward based on the contents of a custom encapsulation
header as well as perform normal IP forwarding if the encapsulation header
does not exist in the packet.
Your job will be to do the following:
1. **NOTE:** A new header type has been added called `myTunnel_t` that contains two 16-bit fields: `proto_id` and `dst_id`.
2. **NOTE:** The `myTunnel_t` header has been added to the `headers` struct.
2. **TODO:** Update the parser to extract either the `myTunnel` header or `ipv4` header based on the `etherType` field in the Ethernet header. The etherType corresponding to the myTunnel header is `0x1212`. The parser should also extract the `ipv4` header after the `myTunnel` header if `proto_id` == `TYPE_IPV4` (i.e. 0x0800).
3. **TODO:** Define a new action called `myTunnel_forward` that simply sets the egress port (i.e. `egress_spec` field of the `standard_metadata` bus) to the port number provided by the control plane.
4. **TODO:** Define a new table called `myTunnel_exact` that perfoms an exact match on the `dst_id` field of the `myTunnel` header. This table should invoke either the `myTunnel_forward` action if the there is a match in the table and it should invoke the `drop` action otherwise.
5. **TODO:** Update the `apply` statement in the `MyIngress` control block to apply your newly defined `myTunnel_exact` table if the `myTunnel` header is valid. Otherwise, invoke the `ipv4_lpm` table if the `ipv4` header is valid.
6. **TODO:** Update the deparser to emit the `ethernet`, then `myTunnel`, then `ipv4` headers. Remember that the deparser will only emit a header if it is valid. A header's implicit validity bit is set by the parser upon extraction. So there is no need to check header validity here.
7. **TODO:** Add static rules for your newly defined table so that the switches will forward correctly for each possible value of `dst_id`. See the diagram below for the topology's port configuration as well as how we will assign IDs to hosts. For this step you will need to add your forwarding rules to the `sX-commands.txt` files.
![topology](./topo.png)
## Step 3: Run your solution
Follow the instructions from Step 1. This time when you send a packet from
`h1` to `h2` try using the following command to send a packet that uses
our new `myTunnel` header.
```bash
./send.py 10.0.2.2 "P4 is cool" --dst_id 2
```
You should see a packet arrive at `h2` which contains the `MyTunnel` header.
Also note that changing the destination IP address will not prevent the packet
from arriving at `h2`. This is because the switch is no longer using the IP header for routing when the `MyTunnel` header is in the packet.
> Python Scapy does not natively support the `myTunnel` header
> type so we have provided a file called `myTunnel_header.py` which
> adds support to Scapy for our new custom header. Feel free to inspect
> this file if you are interested in learning how to do this.
### Food for thought
To make this tunneling exercise a bit more interesting (and realistic)
how might you change the P4 code to have the switches add the `myTunnel`
header to an IP packet upon ingress to the network and then remove the
`myTunnel` header as the packet leaves to the network to an end host?
Hints:
- The ingress switch will need to map the destination IP address to the corresponding `dst_id` for the `myTunnel` header. Also remember to set explicitly set the validity bit for the `myTunnel` header so that it can be emitted by the deparser.
- The egress switch will need to remove the `myTunnel` header from the packet after looking up the appropriate output port using the `dst_id` field.
### Troubleshooting
There are several problems that might manifest as you develop your program:
1. `basic_tunnel.p4` might fail to compile. In this case, `make run` will
report the error emitted from the compiler and halt.
2. `basic_tunnel.p4` might compile but fail to support the control plane
rules in the `s1-commands.txt` through `s3-command.txt` files that
`make run` tries to install using the Bmv2 CLI. In this case, `make run`
will log the CLI tool output in the `logs` directory. Use these error
messages to fix your `basic_tunnel.p4` implementation or forwarding rules.
3. `basic_tunnel.p4` might compile, and the control plane rules might be
installed, but the switch might not process packets in the desired
way. The `/tmp/p4s.<switch-name>.log` files contain detailed logs
that describing how each switch processes each packet. The output is
detailed and can help pinpoint logic errors in your implementation.
#### Cleaning up Mininet
In the latter two cases above, `make` may leave a Mininet instance
running in the background. Use the following command to clean up
these instances:
```bash
make stop
```
## Next Steps
Congratulations, your implementation works! Move onto the next assignment
[p4runtime](../p4runtime)!

View File

@@ -0,0 +1,197 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
// NOTE: new type added here
const bit<16> TYPE_MYTUNNEL = 0x1212;
const bit<16> TYPE_IPV4 = 0x800;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
// NOTE: added new header type
header myTunnel_t {
bit<16> proto_id;
bit<16> dst_id;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
// NOTE: Added new header type to headers struct
struct headers {
ethernet_t ethernet;
myTunnel_t myTunnel;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
// TODO: Update the parser to parse the myTunnel header as well
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4 : parse_ipv4;
default : accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
// TODO: declare a new action: myTunnel_forward(egressSpec_t port)
// TODO: declare a new table: myTunnel_exact
// TODO: also remember to add table entries!
apply {
// TODO: Update control flow
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
// TODO: emit myTunnel header as well
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,20 @@
from scapy.all import *
import sys, os
TYPE_MYTUNNEL = 0x1212
TYPE_IPV4 = 0x0800
class MyTunnel(Packet):
name = "MyTunnel"
fields_desc = [
ShortField("pid", 0),
ShortField("dst_id", 0)
]
def mysummary(self):
return self.sprintf("pid=%pid%, dst_id=%dst_id%")
bind_layers(Ether, MyTunnel, type=TYPE_MYTUNNEL)
bind_layers(MyTunnel, IP, pid=TYPE_IPV4)

View File

@@ -0,0 +1,43 @@
#!/usr/bin/env python
import sys
import struct
import os
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import ShortField, IntField, LongField, BitField, FieldListField, FieldLenField
from scapy.all import IP, TCP, UDP, Raw
from scapy.layers.inet import _IPOption_HDR
from myTunnel_header import MyTunnel
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def handle_pkt(pkt):
if MyTunnel in pkt or (TCP in pkt and pkt[TCP].dport == 1234):
print "got a packet"
pkt.show2()
# hexdump(pkt)
# print "len(pkt) = ", len(pkt)
sys.stdout.flush()
def main():
ifaces = filter(lambda i: 'eth' in i, os.listdir('/sys/class/net/'))
iface = ifaces[0]
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,9 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:00:01:01 1
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:02:01:00 2
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:03:01:00 3
table_set_default myTunnel_exact drop
table_add myTunnel_exact myTunnel_forward 1 => 1
table_add myTunnel_exact myTunnel_forward 2 => 2
table_add myTunnel_exact myTunnel_forward 3 => 3

View File

@@ -0,0 +1,9 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:01:02:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:00:02:02 1
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:03:02:00 3
table_set_default myTunnel_exact drop
table_add myTunnel_exact myTunnel_forward 1 => 2
table_add myTunnel_exact myTunnel_forward 2 => 1
table_add myTunnel_exact myTunnel_forward 3 => 3

View File

@@ -0,0 +1,9 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:01:03:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:02:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:00:03:03 1
table_set_default myTunnel_exact drop
table_add myTunnel_exact myTunnel_forward 1 => 2
table_add myTunnel_exact myTunnel_forward 2 => 3
table_add myTunnel_exact myTunnel_forward 3 => 1

View File

@@ -0,0 +1,53 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
import argparse
from scapy.all import sendp, send, get_if_list, get_if_hwaddr, hexdump
from scapy.all import Packet
from scapy.all import Ether, IP, UDP, TCP
from myTunnel_header import MyTunnel
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def main():
parser = argparse.ArgumentParser()
parser.add_argument('ip_addr', type=str, help="The destination IP address to use")
parser.add_argument('message', type=str, help="The message to include in packet")
parser.add_argument('--dst_id', type=int, default=None, help='The myTunnel dst_id to use, if unspecified then myTunnel header will not be included in packet')
args = parser.parse_args()
addr = socket.gethostbyname(args.ip_addr)
dst_id = args.dst_id
iface = get_if()
if (dst_id is not None):
print "sending on interface {} to dst_id {}".format(iface, str(dst_id))
pkt = Ether(src=get_if_hwaddr(iface), dst='ff:ff:ff:ff:ff:ff')
pkt = pkt / MyTunnel(dst_id=dst_id) / IP(dst=addr) / args.message
else:
print "sending on interface {} to IP addr {}".format(iface, str(addr))
pkt = Ether(src=get_if_hwaddr(iface), dst='ff:ff:ff:ff:ff:ff')
pkt = pkt / IP(dst=addr) / TCP(dport=1234, sport=random.randint(49152,65535)) / args.message
pkt.show2()
# hexdump(pkt)
# print "len(pkt) = ", len(pkt)
sendp(pkt, iface=iface, verbose=False)
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,215 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_MYTUNNEL = 0x1212;
const bit<16> TYPE_IPV4 = 0x800;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header myTunnel_t {
bit<16> proto_id;
bit<16> dst_id;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
myTunnel_t myTunnel;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_MYTUNNEL: parse_myTunnel;
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_myTunnel {
packet.extract(hdr.myTunnel);
transition select(hdr.myTunnel.proto_id) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
action myTunnel_forward(egressSpec_t port) {
standard_metadata.egress_spec = port;
}
table myTunnel_exact {
key = {
hdr.myTunnel.dst_id: exact;
}
actions = {
myTunnel_forward;
drop;
}
size = 1024;
default_action = drop();
}
apply {
if (hdr.ipv4.isValid() && !hdr.myTunnel.isValid()) {
// Process only non-tunneled IPv4 packets
ipv4_lpm.apply();
}
if (hdr.myTunnel.isValid()) {
// process tunneled packets
myTunnel_exact.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.myTunnel);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

Binary file not shown.

Binary file not shown.

After

Width:  |  Height:  |  Size: 92 KiB

View File

@@ -0,0 +1,16 @@
{
"hosts": [
"h1",
"h2",
"h3"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["s1", "s2"], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,112 @@
# Implementing a P4 Calculator
## Introduction
The objective of this tutorial is to implement a basic calculator
using a custom protocol header written in P4. The header will contain
an operation to perform and two operands. When a switch receives a
calculator packet header, it will execute the operation on the
operands, and return the result to the sender.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`calc.p4`, which initially drops all packets. Your job will be to
extend it to properly implement the calculator logic.
As a first step, compile the incomplete `calc.p4` and bring up a
switch in Mininet to test its behavior.
1. In your shell, run:
```bash
make
```
This will:
* compile `calc.p4`, and
* start a Mininet instance with one switches (`s1`) connected to
two hosts (`h1`, `h2`).
* The hosts are assigned IPs of `10.0.1.1` and `10.0.1.2`.
2. We've written a small Python-based driver program that will allow
you to test your calculator. You can run the driver program directly
from the Mininet command prompt:
```
mininet> h1 python calc.py
>
```
3. The driver program will provide a new prompt, at which you can type
basic expressions. The test harness will parse your expression, and
prepare a packet with the corresponding operator and operands. It will
then send a packet to the switch for evaluation. When the switch
returns the result of the computation, the test program will print the
result. However, because the calculator program is not implemented,
you should see an error message.
```
> 1+1
Didn't receive response
>
```
## Step 2: Implement Calculator
To implement the calculator, you will need to define a custom
calculator header, and implement the switch logic to parse header,
perform the requested operation, write the result in the header, and
return the packet to the sender.
We will use the following header format:
0 1 2 3
+----------------+----------------+----------------+---------------+
| P | 4 | Version | Op |
+----------------+----------------+----------------+---------------+
| Operand A |
+----------------+----------------+----------------+---------------+
| Operand B |
+----------------+----------------+----------------+---------------+
| Result |
+----------------+----------------+----------------+---------------+
- P is an ASCII Letter 'P' (0x50)
- 4 is an ASCII Letter '4' (0x34)
- Version is currently 0.1 (0x01)
- Op is an operation to Perform:
- '+' (0x2b) Result = OperandA + OperandB
- '-' (0x2d) Result = OperandA - OperandB
- '&' (0x26) Result = OperandA & OperandB
- '|' (0x7c) Result = OperandA | OperandB
- '^' (0x5e) Result = OperandA ^ OperandB
We will assume that the calculator header is carried over Ethernet,
and we will use the Ethernet type 0x1234 to indicate the presence of
the header.
Given what you have learned so far, your task is to implement the P4
calculator program. There is no control plane logic, so you need only
worry about the data plane implementation.
A working calculator implementation will parse the custom headers,
execute the mathematical operation, write the result in the result
field, and return the packet to the sender.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, you should see the
correct result:
```
> 1+1
2
>
```
## Next Steps
Congratulations, your implementation works! Move on to
[Load Balancer](../load_balance).

View File

@@ -0,0 +1,250 @@
/* -*- P4_16 -*- */
/*
* P4 Calculator
*
* This program implements a simple protocol. It can be carried over Ethernet
* (Ethertype 0x1234).
*
* The Protocol header looks like this:
*
* 0 1 2 3
* +----------------+----------------+----------------+---------------+
* | P | 4 | Version | Op |
* +----------------+----------------+----------------+---------------+
* | Operand A |
* +----------------+----------------+----------------+---------------+
* | Operand B |
* +----------------+----------------+----------------+---------------+
* | Result |
* +----------------+----------------+----------------+---------------+
*
* P is an ASCII Letter 'P' (0x50)
* 4 is an ASCII Letter '4' (0x34)
* Version is currently 0.1 (0x01)
* Op is an operation to Perform:
* '+' (0x2b) Result = OperandA + OperandB
* '-' (0x2d) Result = OperandA - OperandB
* '&' (0x26) Result = OperandA & OperandB
* '|' (0x7c) Result = OperandA | OperandB
* '^' (0x5e) Result = OperandA ^ OperandB
*
* The device receives a packet, performs the requested operation, fills in the
* result and sends the packet back out of the same port it came in on, while
* swapping the source and destination addresses.
*
* If an unknown operation is specified or the header is not valid, the packet
* is dropped
*/
#include <core.p4>
#include <v1model.p4>
/*
* Define the headers the program will recognize
*/
/*
* Standard Ethernet header
*/
header ethernet_t {
bit<48> dstAddr;
bit<48> srcAddr;
bit<16> etherType;
}
/*
* This is a custom protocol header for the calculator. We'll use
* etherType 0x1234 for it (see parser)
*/
const bit<16> P4CALC_ETYPE = 0x1234;
const bit<8> P4CALC_P = 0x50; // 'P'
const bit<8> P4CALC_4 = 0x34; // '4'
const bit<8> P4CALC_VER = 0x01; // v0.1
const bit<8> P4CALC_PLUS = 0x2b; // '+'
const bit<8> P4CALC_MINUS = 0x2d; // '-'
const bit<8> P4CALC_AND = 0x26; // '&'
const bit<8> P4CALC_OR = 0x7c; // '|'
const bit<8> P4CALC_CARET = 0x5e; // '^'
header p4calc_t {
bit<8> op;
/* TODO
* fill p4calc_t header with P, four, ver, op, operand_a, operand_b, and res
entries based on above protocol header definition.
*/
}
/*
* All headers, used in the program needs to be assembled into a single struct.
* We only need to declare the type, but there is no need to instantiate it,
* because it is done "by the architecture", i.e. outside of P4 functions
*/
struct headers {
ethernet_t ethernet;
p4calc_t p4calc;
}
/*
* All metadata, globally used in the program, also needs to be assembled
* into a single struct. As in the case of the headers, we only need to
* declare the type, but there is no need to instantiate it,
* because it is done "by the architecture", i.e. outside of P4 functions
*/
struct metadata {
/* In our case it is empty */
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
P4CALC_ETYPE : check_p4calc;
default : accept;
}
}
state check_p4calc {
/* TODO: just uncomment the following parse block */
/*
transition select(packet.lookahead<p4calc_t>().p,
packet.lookahead<p4calc_t>().four,
packet.lookahead<p4calc_t>().ver) {
(P4CALC_P, P4CALC_4, P4CALC_VER) : parse_p4calc;
default : accept;
}
*/
}
state parse_p4calc {
packet.extract(hdr.p4calc);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr,
inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action send_back(bit<32> result) {
/* TODO
* - put the result back in hdr.p4calc.res
* - swap MAC addresses in hdr.ethernet.dstAddr and
* hdr.ethernet.srcAddr using a temp variable
* - Send the packet back to the port it came from
by saving standard_metadata.ingress_port into
standard_metadata.egress_spec
*/
}
action operation_add() {
/* TODO call send_back with operand_a + operand_b */
}
action operation_sub() {
/* TODO call send_back with operand_a - operand_b */
}
action operation_and() {
/* TODO call send_back with operand_a & operand_b */
}
action operation_or() {
/* TODO call send_back with operand_a | operand_b */
}
action operation_xor() {
/* TODO call send_back with operand_a ^ operand_b */
}
action operation_drop() {
mark_to_drop();
}
table calculate {
key = {
hdr.p4calc.op : exact;
}
actions = {
operation_add;
operation_sub;
operation_and;
operation_or;
operation_xor;
operation_drop;
}
const default_action = operation_drop();
const entries = {
P4CALC_PLUS : operation_add();
P4CALC_MINUS: operation_sub();
P4CALC_AND : operation_and();
P4CALC_OR : operation_or();
P4CALC_CARET: operation_xor();
}
}
apply {
if (hdr.p4calc.isValid()) {
calculate.apply();
} else {
operation_drop();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.p4calc);
}
}
/*************************************************************************
*********************** S W I T T C H **********************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,97 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
import re
from scapy.all import sendp, send, srp1
from scapy.all import Packet, hexdump
from scapy.all import Ether, StrFixedLenField, XByteField, IntField
from scapy.all import bind_layers
import readline
class P4calc(Packet):
name = "P4calc"
fields_desc = [ StrFixedLenField("P", "P", length=1),
StrFixedLenField("Four", "4", length=1),
XByteField("version", 0x01),
StrFixedLenField("op", "+", length=1),
IntField("operand_a", 0),
IntField("operand_b", 0),
IntField("result", 0xDEADBABE)]
bind_layers(Ether, P4calc, type=0x1234)
class NumParseError(Exception):
pass
class OpParseError(Exception):
pass
class Token:
def __init__(self,type,value = None):
self.type = type
self.value = value
def num_parser(s, i, ts):
pattern = "^\s*([0-9]+)\s*"
match = re.match(pattern,s[i:])
if match:
ts.append(Token('num', match.group(1)))
return i + match.end(), ts
raise NumParseError('Expected number literal.')
def op_parser(s, i, ts):
pattern = "^\s*([-+&|^])\s*"
match = re.match(pattern,s[i:])
if match:
ts.append(Token('num', match.group(1)))
return i + match.end(), ts
raise NumParseError("Expected binary operator '-', '+', '&', '|', or '^'.")
def make_seq(p1, p2):
def parse(s, i, ts):
i,ts2 = p1(s,i,ts)
return p2(s,i,ts2)
return parse
def main():
p = make_seq(num_parser, make_seq(op_parser,num_parser))
s = ''
iface = 'h1-eth0'
while True:
s = str(raw_input('> '))
if s == "quit":
break
print s
try:
i,ts = p(s,0,[])
pkt = Ether(dst='00:04:00:00:00:00', type=0x1234) / P4calc(op=ts[1].value,
operand_a=int(ts[0].value),
operand_b=int(ts[2].value))
pkt = pkt/' '
# pkt.show()
resp = srp1(pkt, iface=iface, timeout=1, verbose=False)
if resp:
p4calc=resp[P4calc]
if p4calc:
print p4calc.result
else:
print "cannot find P4calc header in the packet"
else:
print "Didn't receive response"
except Exception as error:
print error
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,256 @@
/* -*- P4_16 -*- */
/*
* P4 Calculator
*
* This program implements a simple protocol. It can be carried over Ethernet
* (Ethertype 0x1234).
*
* The Protocol header looks like this:
*
* 0 1 2 3
* +----------------+----------------+----------------+---------------+
* | P | 4 | Version | Op |
* +----------------+----------------+----------------+---------------+
* | Operand A |
* +----------------+----------------+----------------+---------------+
* | Operand B |
* +----------------+----------------+----------------+---------------+
* | Result |
* +----------------+----------------+----------------+---------------+
*
* P is an ASCII Letter 'P' (0x50)
* 4 is an ASCII Letter '4' (0x34)
* Version is currently 0.1 (0x01)
* Op is an operation to Perform:
* '+' (0x2b) Result = OperandA + OperandB
* '-' (0x2d) Result = OperandA - OperandB
* '&' (0x26) Result = OperandA & OperandB
* '|' (0x7c) Result = OperandA | OperandB
* '^' (0x5e) Result = OperandA ^ OperandB
*
* The device receives a packet, performs the requested operation, fills in the
* result and sends the packet back out of the same port it came in on, while
* swapping the source and destination addresses.
*
* If an unknown operation is specified or the header is not valid, the packet
* is dropped
*/
#include <core.p4>
#include <v1model.p4>
/*
* Define the headers the program will recognize
*/
/*
* Standard ethernet header
*/
header ethernet_t {
bit<48> dstAddr;
bit<48> srcAddr;
bit<16> etherType;
}
/*
* This is a custom protocol header for the calculator. We'll use
* ethertype 0x1234 for is (see parser)
*/
const bit<16> P4CALC_ETYPE = 0x1234;
const bit<8> P4CALC_P = 0x50; // 'P'
const bit<8> P4CALC_4 = 0x34; // '4'
const bit<8> P4CALC_VER = 0x01; // v0.1
const bit<8> P4CALC_PLUS = 0x2b; // '+'
const bit<8> P4CALC_MINUS = 0x2d; // '-'
const bit<8> P4CALC_AND = 0x26; // '&'
const bit<8> P4CALC_OR = 0x7c; // '|'
const bit<8> P4CALC_CARET = 0x5e; // '^'
header p4calc_t {
bit<8> p;
bit<8> four;
bit<8> ver;
bit<8> op;
bit<32> operand_a;
bit<32> operand_b;
bit<32> res;
}
/*
* All headers, used in the program needs to be assembed into a single struct.
* We only need to declare the type, but there is no need to instantiate it,
* because it is done "by the architecture", i.e. outside of P4 functions
*/
struct headers {
ethernet_t ethernet;
p4calc_t p4calc;
}
/*
* All metadata, globally used in the program, also needs to be assembed
* into a single struct. As in the case of the headers, we only need to
* declare the type, but there is no need to instantiate it,
* because it is done "by the architecture", i.e. outside of P4 functions
*/
struct metadata {
/* In our case it is empty */
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
P4CALC_ETYPE : check_p4calc;
default : accept;
}
}
state check_p4calc {
transition select(packet.lookahead<p4calc_t>().p,
packet.lookahead<p4calc_t>().four,
packet.lookahead<p4calc_t>().ver) {
(P4CALC_P, P4CALC_4, P4CALC_VER) : parse_p4calc;
default : accept;
}
}
state parse_p4calc {
packet.extract(hdr.p4calc);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr,
inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action send_back(bit<32> result) {
bit<48> tmp;
/* Put the result back in */
hdr.p4calc.res = result;
/* Swap the MAC addresses */
tmp = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = hdr.ethernet.srcAddr;
hdr.ethernet.srcAddr = tmp;
/* Send the packet back to the port it came from */
standard_metadata.egress_spec = standard_metadata.ingress_port;
}
action operation_add() {
send_back(hdr.p4calc.operand_a + hdr.p4calc.operand_b);
}
action operation_sub() {
send_back(hdr.p4calc.operand_a - hdr.p4calc.operand_b);
}
action operation_and() {
send_back(hdr.p4calc.operand_a & hdr.p4calc.operand_b);
}
action operation_or() {
send_back(hdr.p4calc.operand_a | hdr.p4calc.operand_b);
}
action operation_xor() {
send_back(hdr.p4calc.operand_a ^ hdr.p4calc.operand_b);
}
action operation_drop() {
mark_to_drop();
}
table calculate {
key = {
hdr.p4calc.op : exact;
}
actions = {
operation_add;
operation_sub;
operation_and;
operation_or;
operation_xor;
operation_drop;
}
const default_action = operation_drop();
const entries = {
P4CALC_PLUS : operation_add();
P4CALC_MINUS: operation_sub();
P4CALC_AND : operation_and();
P4CALC_OR : operation_or();
P4CALC_CARET: operation_xor();
}
}
apply {
if (hdr.p4calc.isValid()) {
calculate.apply();
} else {
operation_drop();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.p4calc);
}
}
/*************************************************************************
*********************** S W I T T C H **********************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,12 @@
{
"hosts": [
"h1",
"h2"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" }
},
"links": [
["h1", "s1"], ["h2", "s1"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,198 @@
# Implementing ECN
## Introduction
The objective of this tutorial is to extend basic L3 forwarding with
an implementation of Explicit Congestion Notification (ECN).
ECN allows end-to-end notification of network congestion without
dropping packets. If an end-host supports ECN, it puts the value of 1
or 2 in the `ipv4.ecn` field. For such packets, each switch may
change the value to 3 if the queue size is larger than a threshold.
The receiver copies the value to sender, and the sender can lower the
rate.
As before, we have already defined the control plane rules for
routing, so you only need to implement the data plane logic of your P4
program.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the reference.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`ecn.p4`, which initially implements L3 forwarding. Your job (in the
next step) will be to extend it to properly append set the ECN bits
Before that, let's compile the incomplete `ecn.p4` and bring up a
network in Mininet to test its behavior.
1. In your shell, run:
```bash
make
```
This will:
* compile `ecn.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`) configured
in a triangle. There are 5 hosts. `h1` and `h11` are connected to `s1`.
`h2` and `h22` are connected to `s2` and `h3` is connected to `s3`.
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, etc
(`10.0.<Switchid>.<hostID>`).
* The control plane programs the P4 tables in each switch based on
`sx-commands.txt`
2. We want to send a low rate traffic from `h1` to `h2` and a high
rate iperf traffic from `h11` to `h22`. The link between `s1` and
`s2` is common between the flows and is a bottleneck because we
reduced its bandwidth to 512kbps in topology.json. Therefore, if we
capture packets at `h2`, we should see the right ECN value.
![Setup](setup.png)
3. You should now see a Mininet command prompt. Open four terminals
for `h1`, `h11`, `h2`, `h22`, respectively:
```bash
mininet> xterm h1 h11 h2 h22
```
3. In `h2`'s XTerm, start the server that captures packets:
```bash
./receive.py
```
4. in `h22`'s XTerm, start the iperf UDP server:
```bash
iperf -s -u
```
5. In `h1`'s XTerm, send one packet per second to `h2` using send.py
say for 30 seconds:
```bash
./send.py 10.0.2.2 "P4 is cool" 30
```
The message "P4 is cool" should be received in `h2`'s xterm,
6. In `h11`'s XTerm, start iperf client sending for 15 seconds
```bash
iperf -c 10.0.2.22 -t 15 -u
```
7. At `h2`, the `ipv4.tos` field (DiffServ+ECN) is always 1
8. type `exit` to close each XTerm window
Your job is to extend the code in `ecn.p4` to implement the ECN logic
for setting the ECN flag.
## Step 2: Implement ECN
The `ecn.p4` file contains a skeleton P4 program with key pieces of
logic replaced by `TODO` comments. These should guide your
implementation---replace each `TODO` with logic implementing the
missing piece.
First we have to change the ipv4_t header by splitting the TOS field
into DiffServ and ECN fields. Remember to update the checksum block
accordingly. Then, in the egress control block we must compare the
queue length with ECN_THRESHOLD. If the queue length is larger than
the threshold, the ECN flag will be set. Note that this logic should
happen only if the end-host declared supporting ECN by setting the
original ECN to 1 or 2.
A complete `ecn.p4` will contain the following components:
1. Header type definitions for Ethernet (`ethernet_t`) and IPv4 (`ipv4_t`).
2. Parsers for Ethernet, IPv4,
3. An action to drop a packet, using `mark_to_drop()`.
4. An action (called `ipv4_forward`), which will:
1. Set the egress port for the next hop.
2. Update the ethernet destination address with the address of
the next hop.
3. Update the ethernet source address with the address of the switch.
4. Decrement the TTL.
5. An egress control block that checks the ECN and
`standard_metadata.enq_qdepth` and sets the ipv4.ecn.
6. A deparser that selects the order in which fields inserted into the outgoing
packet.
7. A `package` instantiation supplied with the parser, control,
checksum verification and recomputation and deparser.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, when your message from
`h1` is delivered to `h2`, you should see `tos` values change from 1
to 3 as the queue builds up. `tos` may change back to 1 when iperf
finishes and the queue depletes.
To easily track the `tos` values you may want to redirect the output
of `h2` to a file by running the following for `h2`
```bash
./receive.py > h2.log
```
and just print the `tos` values `grep tos h2.log` in a separate window
```
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x3
tos = 0x3
tos = 0x3
tos = 0x3
tos = 0x3
tos = 0x3
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
tos = 0x1
```
### Food for thought
How can we let the user configure the threshold?
### Troubleshooting
There are several ways that problems might manifest:
1. `ecn.p4` fails to compile. In this case, `make` will report the
error emitted from the compiler and stop.
2. `ecn.p4` compiles but does not support the control plane rules in
the `sX-commands.txt` files that `make` tries to install using
the BMv2 CLI. In this case, `make` will log the CLI tool output
in the `logs` directory. Use these error messages to fix your `ecn.p4`
implementation.
3. `ecn.p4` compiles, and the control plane rules are installed, but
the switch does not process packets in the desired way. The
`/tmp/p4s.<switch-name>.log` files contain trace messages
describing how each switch processes each packet. The output is
detailed and can help pinpoint logic errors in your implementation.
The `build/<switch-name>-<interface-name>.pcap` also contains the
pcap of packets on each interface. Use `tcpdump -r <filename> -xxx`
to print the hexdump of the packets.
4. `ecn.p4` compiles and all rules are installed. Packets go through
and the logs show that the queue length was not high enough to set
the ECN bit. Then either lower the threshold in the p4 code or
reduce the link bandwidth in `topology.json`
#### Cleaning up Mininet
In the latter two cases above, `make` may leave a Mininet instance
running in the background. Use the following command to clean up
these instances:
```bash
make stop
```
## Next Steps
Congratulations, your implementation works! Move on to the next
exercise: [Multi-Hop Route Inspection](../mri), which identifies which
link is the source of congestion.

View File

@@ -0,0 +1,188 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<8> TCP_PROTOCOL = 0x06;
const bit<16> TYPE_IPV4 = 0x800;
const bit<19> ECN_THRESHOLD = 10;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
/*
* TODO: split tos to two fields 6 bit diffserv and 2 bit ecn
*/
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> tos;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply {
/*
* TODO:
* - if ecn is 1 or 2
* - compare standard_metadata.enq_qdepth with threshold
* and set hdr.ipv4.ecn to 3 if larger
*/
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
/* TODO: replace tos with diffserve and ecn */
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.tos,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,37 @@
#!/usr/bin/env python
import sys
import struct
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet
from scapy.all import IP, UDP, Raw
from scapy.layers.inet import _IPOption_HDR
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def handle_pkt(pkt):
print "got a packet"
pkt.show2()
# hexdump(pkt)
sys.stdout.flush()
def main():
iface = 'h2-eth0'
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(filter="udp and port 4321", iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,5 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:00:01:01 2
table_add ipv4_lpm ipv4_forward 10.0.1.11/32 => 00:00:00:00:01:0b 1
table_add ipv4_lpm ipv4_forward 10.0.2.0/24 => 00:00:00:02:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.0/24 => 00:00:00:03:02:00 4

View File

@@ -0,0 +1,5 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:00:02:02 2
table_add ipv4_lpm ipv4_forward 10.0.2.22/32 => 00:00:00:00:02:16 1
table_add ipv4_lpm ipv4_forward 10.0.1.0/24 => 00:00:00:01:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.0/24 => 00:00:00:03:03:00 4

View File

@@ -0,0 +1,4 @@
table_set_default ipv4_lpm drop
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:00:03:03 1
table_add ipv4_lpm ipv4_forward 10.0.1.10/24 => 00:00:00:01:04:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.10/24 => 00:00:00:02:04:00 3

View File

@@ -0,0 +1,50 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
from scapy.all import sendp, send, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import Ether, IP, UDP
from scapy.all import IntField, FieldListField, FieldLenField, ShortField
from scapy.layers.inet import _IPOption_HDR
from time import sleep
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def main():
if len(sys.argv)<4:
print 'pass 2 arguments: <destination> "<message>" <duration>'
exit(1)
addr = socket.gethostbyname(sys.argv[1])
iface = get_if()
pkt = Ether(src=get_if_hwaddr(iface), dst="ff:ff:ff:ff:ff:ff") / IP(dst=addr, tos=1) / UDP(dport=4321, sport=1234) / sys.argv[2]
pkt.show2()
#hexdump(pkt)
try:
for i in range(int(sys.argv[3])):
sendp(pkt, iface=iface)
sleep(1)
except KeyboardInterrupt:
raise
if __name__ == '__main__':
main()

Binary file not shown.

After

Width:  |  Height:  |  Size: 221 KiB

View File

@@ -0,0 +1,188 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<8> TCP_PROTOCOL = 0x06;
const bit<16> TYPE_IPV4 = 0x800;
const bit<19> ECN_THRESHOLD = 10;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<6> diffserv;
bit<2> ecn;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action mark_ecn() {
hdr.ipv4.ecn = 3;
}
apply {
if (hdr.ipv4.ecn == 1 || hdr.ipv4.ecn == 2){
if (standard_metadata.enq_qdepth >= ECN_THRESHOLD){
mark_ecn();
}
}
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.ecn,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,18 @@
{
"hosts": [
"h1",
"h2",
"h3",
"h11",
"h22"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["h11", "s1"], ["s1", "s2", "0", 0.5], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s2", "h22"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,142 @@
# Load Balancing
In this exercise, you will implement a form of load balancing based on
a simple version of Equal-Cost Multipath Forwarding. The switch you
will implement will use two tables to forward packets to one of two
destination hosts at random. The first table will use a hash function
(applied to a 5-tuple consisting of the source and destination
IP addresses, IP protocol, and source and destination TCP ports)
to select one of two hosts. The second table will use the
computed hash value to forward the packet to the selected host.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the
> reference.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`load_balance.p4`, which initially drops all packets. Your job (in
the next step) will be to extend it to properly forward packets.
Before that, let's compile the incomplete `load_balance.p4` and bring
up a switch in Mininet to test its behavior.
1. In your shell, run:
```bash
make
```
This will:
* compile `load_balance.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`) configured
in a triangle, each connected to one host (`h1`, `h2`, `h3`).
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, etc.
* We use the IP address 10.0.0.1 to indicate traffic that should be
load balanced between `h2` and `h3`.
2. You should now see a Mininet command prompt. Open three terminals
for `h1`, `h2` and `h3`, respectively:
```bash
mininet> xterm h1 h2 h3
```
3. Each host includes a small Python-based messaging client and
server. In `h2` and `h3`'s XTerms, start the servers:
```bash
./receive.py
```
4. In `h1`'s XTerm, send a message from the client:
```bash
./send.py 10.0.0.1 "P4 is cool"
```
The message will not be received.
5. Type `exit` to leave each XTerm and the Mininet command line.
The message was not received because each switch is programmed with
`load_balance.p4`, which drops all packets on arrival. Your job is to
extend this file.
### A note about the control plane
P4 programs define a packet-processing pipeline, but the rules
governing packet processing are inserted into the pipeline by the
control plane. When a rule matches a packet, its action is invoked
with parameters supplied by the control plane as part of the rule.
In this exercise, the control plane logic has already been
implemented. As part of bringing up the Mininet instance, the
`make` script will install packet-processing rules in the tables of
each switch. These are defined in the `s1-commands.txt` file.
**Important:** A P4 program also defines the interface between the
switch pipeline and control plane. The `s1-commands.txt` file contains
a list of commands for the BMv2 switch API. These commands refer to
specific tables, keys, and actions by name, and any changes in the P4
program that add or rename tables, keys, or actions will need to be
reflected in these command files.
## Step 2: Implement Load Balancing
The `load_balance.p4` file contains a skeleton P4 program with key
pieces of logic replaced by `TODO` comments. These should guide your
implementation---replace each `TODO` with logic implementing the
missing piece.
A complete `load_balance.p4` will contain the following components:
1. Header type definitions for Ethernet (`ethernet_t`) and IPv4 (`ipv4_t`).
2. Parsers for Ethernet and IPv4 that populate `ethernet_t` and `ipv4_t` fields.
3. An action to drop a packet, using `mark_to_drop()`.
4. **TODO:** An action (called `set_ecmp_select`), which will:
1. Hashes the 5-tuple specified above using the `hash` extern
2. Stores the result in the `meta.ecmp_select` field
5. **TODO:** A control that:
1. Applies the `ecmp_group` table.
2. Applies the `ecmp_nhop` table.
6. A deparser that selects the order in which fields inserted into the outgoing
packet.
7. A `package` instantiation supplied with the parser, control, and deparser.
> In general, a package also requires instances of checksum verification
> and recomputation controls. These are not necessary for this tutorial
> and are replaced with instantiations of empty controls.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, your message from
`h1` should be delivered to `h2` or `h3`. If you send several
messages, some should be received by each server.
### Food for thought
### Troubleshooting
There are several ways that problems might manifest:
1. `load_balance.p4` fails to compile. In this case, `make` will
report the error emitted from the compiler and stop.
2. `load_balance.p4` compiles but does not support the control plane
rules in the `sX-commands.txt` files that `make` tries to install
using the BMv2 CLI. In this case, `make` will log the CLI tool output
in the `logs` directory. Use these error messages to fix your `load_balance.p4`
implementation.
3. `load_balance.p4` compiles, and the control plane rules are
installed, but the switch does not process packets in the desired way.
The `/tmp/p4s.<switch-name>.log` files contain trace messages
describing how each switch processes each packet. The output is
detailed and can help pinpoint logic errors in your implementation.
#### Cleaning up Mininet
In the latter two cases above, `make` may leave a Mininet instance
running in the background. Use the following command to clean up
these instances:
```bash
mn -c
```
## Next Steps
Congratulations, your implementation works!

View File

@@ -0,0 +1,218 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
header ethernet_t {
bit<48> dstAddr;
bit<48> srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
bit<32> srcAddr;
bit<32> dstAddr;
}
header tcp_t {
bit<16> srcPort;
bit<16> dstPort;
bit<32> seqNo;
bit<32> ackNo;
bit<4> dataOffset;
bit<3> res;
bit<3> ecn;
bit<6> ctrl;
bit<16> window;
bit<16> checksum;
bit<16> urgentPtr;
}
struct metadata {
bit<14> ecmp_select;
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
tcp_t tcp;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
0x800: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition select(hdr.ipv4.protocol) {
6: parse_tcp;
default: accept;
}
}
state parse_tcp {
packet.extract(hdr.tcp);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action set_ecmp_select(bit<16> ecmp_base, bit<32> ecmp_count) {
/* TODO: hash on 5-tuple and save the hash result in meta.ecmp_select
so that the ecmp_nhop table can use it to make a forwarding decision accordingly */
}
action set_nhop(bit<48> nhop_dmac, bit<32> nhop_ipv4, bit<9> port) {
hdr.ethernet.dstAddr = nhop_dmac;
hdr.ipv4.dstAddr = nhop_ipv4;
standard_metadata.egress_spec = port;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ecmp_group {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
drop;
set_ecmp_select;
}
size = 1024;
}
table ecmp_nhop {
key = {
meta.ecmp_select: exact;
}
actions = {
drop;
set_nhop;
}
size = 2;
}
apply {
if (hdr.ipv4.isValid() && hdr.ipv4.ttl > 0) {
ecmp_group.apply();
ecmp_nhop.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action rewrite_mac(bit<48> smac) {
hdr.ethernet.srcAddr = smac;
}
action drop() {
mark_to_drop();
}
table send_frame {
key = {
standard_metadata.egress_port: exact;
}
actions = {
rewrite_mac;
drop;
}
size = 256;
}
apply {
send_frame.apply();
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
packet.emit(hdr.tcp);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,52 @@
#!/usr/bin/env python
import sys
import struct
import os
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import ShortField, IntField, LongField, BitField, FieldListField, FieldLenField
from scapy.all import IP, UDP, Raw
from scapy.layers.inet import _IPOption_HDR
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class IPOption_MRI(IPOption):
name = "MRI"
option = 31
fields_desc = [ _IPOption_HDR,
FieldLenField("length", None, fmt="B",
length_of="swids",
adjust=lambda pkt,l:l+4),
ShortField("count", 0),
FieldListField("swids",
[],
IntField("", 0),
length_from=lambda pkt:pkt.count*4) ]
def handle_pkt(pkt):
print "got a packet"
pkt.show2()
# hexdump(pkt)
sys.stdout.flush()
def main():
ifaces = filter(lambda i: 'eth' in i, os.listdir('/sys/class/net/'))
iface = ifaces[0]
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(filter="tcp", iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,6 @@
table_set_default ecmp_group drop
table_add ecmp_group set_ecmp_select 10.0.0.1/32 => 0 2
table_add ecmp_nhop set_nhop 0 => 00:00:00:00:01:02 10.0.2.2 2
table_add ecmp_nhop set_nhop 1 => 00:00:00:00:01:03 10.0.3.3 3
table_add send_frame rewrite_mac 2 => 00:00:00:01:02:00
table_add send_frame rewrite_mac 3 => 00:00:00:01:03:00

View File

@@ -0,0 +1,4 @@
table_set_default ecmp_group drop
table_add ecmp_group set_ecmp_select 10.0.2.2/32 => 0 1
table_add ecmp_nhop set_nhop 0 => 00:00:00:00:02:02 10.0.2.2 1
table_add send_frame rewrite_mac 1 => 00:00:00:02:01:00

View File

@@ -0,0 +1,4 @@
table_set_default ecmp_group drop
table_add ecmp_group set_ecmp_select 10.0.3.3/32 => 0 1
table_add ecmp_nhop set_nhop 0 => 00:00:00:00:03:03 10.0.3.3 1
table_add send_frame rewrite_mac 1 => 00:00:00:03:01:00

View File

@@ -0,0 +1,41 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
from scapy.all import sendp, send, get_if_list, get_if_hwaddr
from scapy.all import Packet
from scapy.all import Ether, IP, UDP, TCP
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
def main():
if len(sys.argv)<3:
print 'pass 2 arguments: <destination> "<message>"'
exit(1)
addr = socket.gethostbyname(sys.argv[1])
iface = get_if()
print "sending on interface %s to %s" % (iface, str(addr))
pkt = Ether(src=get_if_hwaddr(iface), dst='ff:ff:ff:ff:ff:ff')
pkt = pkt /IP(dst=addr) / TCP(dport=1234, sport=random.randint(49152,65535)) / sys.argv[2]
pkt.show2()
sendp(pkt, iface=iface, verbose=False)
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,225 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
header ethernet_t {
bit<48> dstAddr;
bit<48> srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
bit<32> srcAddr;
bit<32> dstAddr;
}
header tcp_t {
bit<16> srcPort;
bit<16> dstPort;
bit<32> seqNo;
bit<32> ackNo;
bit<4> dataOffset;
bit<3> res;
bit<3> ecn;
bit<6> ctrl;
bit<16> window;
bit<16> checksum;
bit<16> urgentPtr;
}
struct metadata {
bit<14> ecmp_select;
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
tcp_t tcp;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
0x800: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition select(hdr.ipv4.protocol) {
6: parse_tcp;
default: accept;
}
}
state parse_tcp {
packet.extract(hdr.tcp);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action set_ecmp_select(bit<16> ecmp_base, bit<32> ecmp_count) {
hash(meta.ecmp_select,
HashAlgorithm.crc16,
ecmp_base,
{ hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr,
hdr.ipv4.protocol,
hdr.tcp.srcPort,
hdr.tcp.dstPort },
ecmp_count);
}
action set_nhop(bit<48> nhop_dmac, bit<32> nhop_ipv4, bit<9> port) {
hdr.ethernet.dstAddr = nhop_dmac;
hdr.ipv4.dstAddr = nhop_ipv4;
standard_metadata.egress_spec = port;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ecmp_group {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
drop;
set_ecmp_select;
}
size = 1024;
}
table ecmp_nhop {
key = {
meta.ecmp_select: exact;
}
actions = {
drop;
set_nhop;
}
size = 2;
}
apply {
if (hdr.ipv4.isValid() && hdr.ipv4.ttl > 0) {
ecmp_group.apply();
ecmp_nhop.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action rewrite_mac(bit<48> smac) {
hdr.ethernet.srcAddr = smac;
}
action drop() {
mark_to_drop();
}
table send_frame {
key = {
standard_metadata.egress_port: exact;
}
actions = {
rewrite_mac;
drop;
}
size = 256;
}
apply {
send_frame.apply();
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
packet.emit(hdr.tcp);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,16 @@
{
"hosts": [
"h1",
"h2",
"h3"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["s1", "s2"], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,244 @@
# Implementing MRI
## Introduction
The objective of this tutorial is to extend basic L3 forwarding with a
scaled-down version of In-Band Network Telemetry (INT), which we call
Multi-Hop Route Inspection (MRI).
MRI allows users to track the path and the length of queues that every
packet travels through. To support this functionality, you will need
to write a P4 program that appends an ID and queue length to the
header stack of every packet. At the destination, the sequence of
switch IDs correspond to the path, and each ID is followed by the
queue length of the port at switch.
As before, we have already defined the control plane rules, so you
only need to implement the data plane logic of your P4 program.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the reference.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`mri.p4`, which initially implements L3 forwarding. Your job (in the
next step) will be to extend it to properly prepend the MRI custom
headers.
Before that, let's compile the incomplete `mri.p4` and bring up a
switch in Mininet to test its behavior.
1. In your shell, run:
```bash
make
```
This will:
* compile `mri.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`) configured
in a triangle. There are 5 hosts. `h1` and `h11` are connected to `s1`.
`h2` and `h22` are connected to `s2` and `h3` is connected to `s3`.
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, etc
(`10.0.<Switchid>.<hostID>`).
* The control plane programs the P4 tables in each switch based on
`sx-commands.txt`
2. We want to send a low rate traffic from `h1` to `h2` and a high
rate iperf traffic from `h11` to `h22`. The link between `s1` and
`s2` is common between the flows and is a bottleneck because we
reduced its bandwidth to 512kbps in topology.json. Therefore, if we
capture packets at `h2`, we should see high queue size for that
link.
![Setup](setup.png)
3. You should now see a Mininet command prompt. Open four terminals
for `h1`, `h11`, `h2`, `h22`, respectively:
```bash
mininet> xterm h1 h11 h2 h22
```
3. In `h2`'s xterm, start the server that captures packets:
```bash
./receive.py
```
4. in `h22`'s xterm, start the iperf UDP server:
```bash
iperf -s -u
```
5. In `h1`'s xterm, send one packet per second to `h2` using send.py
say for 30 seconds:
```bash
./send.py 10.0.2.2 "P4 is cool" 30
```
The message "P4 is cool" should be received in `h2`'s xterm,
6. In `h11`'s xterm, start iperf client sending for 15 seconds
```bash
iperf -c 10.0.2.22 -t 15 -u
```
7. At `h2`, the MRI header has no hop info (`count=0`)
8. type `exit` to close each xterm window
You should see the message received at host `h2`, but without any
information about the path the message took. Your job is to extend
the code in `mri.p4` to implement the MRI logic to record the path.
### A note about the control plane
P4 programs define a packet-processing pipeline, but the rules
governing packet processing are inserted into the pipeline by the
control plane. When a rule matches a packet, its action is invoked
with parameters supplied by the control plane as part of the rule.
In this exercise, the control plane logic has already been
implemented. As part of bringing up the Mininet instance, the
`make` script will install packet-processing rules in the tables of
each switch. These are defined in the `sX-commands.txt` files, where
`X` corresponds to the switch number.
## Step 2: Implement MRI
The `mri.p4` file contains a skeleton P4 program with key pieces of
logic replaced by `TODO` comments. These should guide your
implementation---replace each `TODO` with logic implementing the
missing piece.
MRI will require two custom headers. The first header, `mri_t`,
contains a single field `count`, which indicates the number of switch
IDs that follow. The second header, `switch_t`, contains switch ID and
Queue depth fields of each switch hop the packet goes through.
One of the biggest challenges in implementing MRI is handling the
recursive logic for parsing these two headers. We will use a
`parser_metadata` field, `remaining`, to keep track of how many
`switch_t` headers we need to parse. In the `parse_mri` state, this
field should be set to `hdr.mri.count`. In the `parse_swtrace` state,
this field should be decremented. The `parse_swtrace` state will
transition to itself until `remaining` is 0.
The MRI custom headers will be carried inside an IP Options
header. The IP Options header contains a field, `option`, which
indicates the type of the option. We will use a special type 31 to
indicate the presence of the MRI headers.
Beyond the parser logic, you will add a table in egress, `swtrace` to
store the switch ID and queue depth, and actions that increment the
`count` field, and append a `switch_t` header.
A complete `mri.p4` will contain the following components:
1. Header type definitions for Ethernet (`ethernet_t`), IPv4 (`ipv4_t`),
IP Options (`ipv4_option_t`), MRI (`mri_t`), and Switch (`switch_t`).
2. Parsers for Ethernet, IPv4, IP Options, MRI, and Switch that will
populate `ethernet_t`, `ipv4_t`, `ipv4_option_t`, `mri_t`, and
`switch_t`.
3. An action to drop a packet, using `mark_to_drop()`.
4. An action (called `ipv4_forward`), which will:
1. Set the egress port for the next hop.
2. Update the ethernet destination address with the address of
the next hop.
3. Update the ethernet source address with the address of the switch.
4. Decrement the TTL.
5. An ingress control that:
1. Defines a table that will read an IPv4 destination address, and
invoke either `drop` or `ipv4_forward`.
2. An `apply` block that applies the table.
6. At egress, an action (called `add_swtrace`) that will add the
switch ID and queue depth.
8. An egress control that applies a table (`swtrace`) to store the
switch ID and queue depth, and calls `add_swtrace`.
9. A deparser that selects the order in which fields inserted into the outgoing
packet.
10. A `package` instantiation supplied with the parser, control,
checksum verification and recomputation and deparser.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, when your message
from `h1` is delivered to `h2`, you should see the sequence of
switches through which the packet traveled plus the corresponding
queue depths. The expected output will look like the following,
which shows the MRI header, with a `count` of 2, and switch ids
(`swids`) 2 and 1. The queue depth at the common link (from s1 to
s2) is high.
```
got a packet
###[ Ethernet ]###
dst = 00:04:00:02:00:02
src = f2:ed:e6:df:4e:fa
type = 0x800
###[ IP ]###
version = 4L
ihl = 10L
tos = 0x0
len = 42
id = 1
flags =
frag = 0L
ttl = 62
proto = udp
chksum = 0x60c0
src = 10.0.1.1
dst = 10.0.2.2
\options \
|###[ MRI ]###
| copy_flag = 0L
| optclass = control
| option = 31L
| length = 20
| count = 2
| \swtraces \
| |###[ SwitchTrace ]###
| | swid = 2
| | qdepth = 0
| |###[ SwitchTrace ]###
| | swid = 1
| | qdepth = 17
###[ UDP ]###
sport = 1234
dport = 4321
len = 18
chksum = 0x1c7b
###[ Raw ]###
load = 'P4 is cool'
```
### Troubleshooting
There are several ways that problems might manifest:
1. `mri.p4` fails to compile. In this case, `make` will report the
error emitted from the compiler and stop.
2. `mri.p4` compiles but does not support the control plane rules in
the `sX-commands.txt` files that `make` tries to install using the BMv2 CLI.
In this case, `make` will log the CLI tool output in the `logs` directory.
Use these error messages to fix your `mri.p4` implementation.
3. `mri.p4` compiles, and the control plane rules are installed, but
the switch does not process packets in the desired way. The
`/tmp/p4s.<switch-name>.log` files contain trace messages describing
how each switch processes each packet. The output is detailed and can
help pinpoint logic errors in your implementation. The
`build/<switch-name>-<interface-name>.pcap` also contains the pcap of
packets on each interface. Use `tcpdump -r <filename> -xxx` to print
the hexdump of the packets.
4. `mri.p4` compiles and all rules are installed. Packets go through
and the logs show that the queue length is always 0. Then either
reduce the link bandwidth in `topology.json`.
#### Cleaning up Mininet
In the latter two cases above, `make` may leave a Mininet instance
running in the background. Use the following command to clean up
these instances:
```bash
make stop
```
## Next Steps
Congratulations, your implementation works! Move on to [Source
Routing](../source_routing).

View File

@@ -0,0 +1,281 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<8> UDP_PROTOCOL = 0x11;
const bit<16> TYPE_IPV4 = 0x800;
const bit<5> IPV4_OPTION_MRI = 31;
#define MAX_HOPS 9
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
typedef bit<32> switchID_t;
typedef bit<32> qdepth_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
header ipv4_option_t {
bit<1> copyFlag;
bit<2> optClass;
bit<5> option;
bit<8> optionLength;
}
header mri_t {
bit<16> count;
}
header switch_t {
switchID_t swid;
qdepth_t qdepth;
}
struct ingress_metadata_t {
bit<16> count;
}
struct parser_metadata_t {
bit<16> remaining;
}
struct metadata {
ingress_metadata_t ingress_metadata;
parser_metadata_t parser_metadata;
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
ipv4_option_t ipv4_option;
mri_t mri;
switch_t[MAX_HOPS] swtraces;
}
error { IPHeaderTooShort }
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
verify(hdr.ipv4.ihl >= 5, error.IPHeaderTooShort);
transition select(hdr.ipv4.ihl) {
5 : accept;
default : parse_ipv4_option;
}
}
state parse_ipv4_option {
/*
* TODO: Add logic to:
* - Extract the ipv4_option header.
* - If value is equal to IPV4_OPTION_MRI, transition to parse_mri.
* - Otherwise, accept.
*/
transition accept;
}
state parse_mri {
/*
* TODO: Add logic to:
* - Extract hdr.mri.
* - Set meta.parser_metadata.remaining to hdr.mri.count
* - Select on the value of meta.parser_metadata.remaining
* - If the value is equal to 0, accept.
* - Otherwise, transition to parse_swtrace.
*/
transition accept;
}
state parse_swtrace {
/*
* TODO: Add logic to:
* - Extract hdr.swtraces.next.
* - Decrement meta.parser_metadata.remaining by 1
* - Select on the value of meta.parser_metadata.remaining
* - If the value is equal to 0, accept.
* - Otherwise, transition to parse_swtrace.
*/
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action add_swtrace(switchID_t swid) {
/*
* TODO: add logic to:
- Increment hdr.mri.count by 1
- Add a new swtrace header by calling push_front(1) on hdr.swtraces.
- Set hdr.swtraces[0].swid to the id parameter
- Set hdr.swtraces[0].qdepth to (qdepth_t)standard_metadata.deq_qdepth
- Increment hdr.ipv4.ihl by 2
- Increment hdr.ipv4.totalLen by 8
- Increment hdr.ipv4_option.optionLength by 8
*/
}
table swtrace {
actions = {
/* TODO: add the correct action */
NoAction;
}
default_action = NoAction();
}
apply {
/*
* TODO: add logic to:
* - If hdr.mri is valid:
* - Apply table swtrace
*/
swtrace.apply();
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
/* TODO: emit ipv4_option, mri and swtraces headers */
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,57 @@
#!/usr/bin/env python
import sys
import struct
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import PacketListField, ShortField, IntField, LongField, BitField, FieldListField, FieldLenField
from scapy.all import IP, UDP, Raw
from scapy.layers.inet import _IPOption_HDR
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class SwitchTrace(Packet):
fields_desc = [ IntField("swid", 0),
IntField("qdepth", 0)]
def extract_padding(self, p):
return "", p
class IPOption_MRI(IPOption):
name = "MRI"
option = 31
fields_desc = [ _IPOption_HDR,
FieldLenField("length", None, fmt="B",
length_of="swtraces",
adjust=lambda pkt,l:l*2+4),
ShortField("count", 0),
PacketListField("swtraces",
[],
SwitchTrace,
count_from=lambda pkt:(pkt.count*1)) ]
def handle_pkt(pkt):
print "got a packet"
pkt.show2()
# hexdump(pkt)
sys.stdout.flush()
def main():
iface = 'h2-eth0'
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(filter="udp and port 4321", iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,6 @@
table_set_default ipv4_lpm drop
table_set_default swtrace add_swtrace 1
table_add ipv4_lpm ipv4_forward 10.0.1.1/32 => 00:00:00:00:01:01 2
table_add ipv4_lpm ipv4_forward 10.0.1.11/32 => 00:00:00:00:01:0b 1
table_add ipv4_lpm ipv4_forward 10.0.2.0/24 => 00:00:00:02:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.0/24 => 00:00:00:03:02:00 4

View File

@@ -0,0 +1,6 @@
table_set_default ipv4_lpm drop
table_set_default swtrace add_swtrace 2
table_add ipv4_lpm ipv4_forward 10.0.2.2/32 => 00:00:00:00:02:02 2
table_add ipv4_lpm ipv4_forward 10.0.2.22/32 => 00:00:00:00:02:16 1
table_add ipv4_lpm ipv4_forward 10.0.1.0/24 => 00:00:00:01:03:00 3
table_add ipv4_lpm ipv4_forward 10.0.3.0/24 => 00:00:00:03:03:00 4

View File

@@ -0,0 +1,5 @@
table_set_default ipv4_lpm drop
table_set_default swtrace add_swtrace 3
table_add ipv4_lpm ipv4_forward 10.0.3.3/32 => 00:00:00:00:03:01 1
table_add ipv4_lpm ipv4_forward 10.0.1.0/24 => 00:00:00:01:04:00 2
table_add ipv4_lpm ipv4_forward 10.0.2.0/24 => 00:00:00:02:04:00 3

View File

@@ -0,0 +1,78 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
from scapy.all import sendp, send, hexdump, get_if_list, get_if_hwaddr
from scapy.all import Packet, IPOption
from scapy.all import Ether, IP, UDP
from scapy.all import IntField, FieldListField, FieldLenField, ShortField, PacketListField
from scapy.layers.inet import _IPOption_HDR
from time import sleep
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class SwitchTrace(Packet):
fields_desc = [ IntField("swid", 0),
IntField("qdepth", 0)]
def extract_padding(self, p):
return "", p
class IPOption_MRI(IPOption):
name = "MRI"
option = 31
fields_desc = [ _IPOption_HDR,
FieldLenField("length", None, fmt="B",
length_of="swtraces",
adjust=lambda pkt,l:l*2+4),
ShortField("count", 0),
PacketListField("swtraces",
[],
SwitchTrace,
count_from=lambda pkt:(pkt.count*1)) ]
def main():
if len(sys.argv)<3:
print 'pass 2 arguments: <destination> "<message>"'
exit(1)
addr = socket.gethostbyname(sys.argv[1])
iface = get_if()
pkt = Ether(src=get_if_hwaddr(iface), dst="ff:ff:ff:ff:ff:ff") / IP(
dst=addr, options = IPOption_MRI(count=0,
swtraces=[])) / UDP(
dport=4321, sport=1234) / sys.argv[2]
# pkt = Ether(src=get_if_hwaddr(iface), dst="ff:ff:ff:ff:ff:ff") / IP(
# dst=addr, options = IPOption_MRI(count=2,
# swtraces=[SwitchTrace(swid=0,qdepth=0), SwitchTrace(swid=1,qdepth=0)])) / UDP(
# dport=4321, sport=1234) / sys.argv[2]
pkt.show2()
#hexdump(pkt)
try:
for i in range(int(sys.argv[3])):
sendp(pkt, iface=iface)
sleep(1)
except KeyboardInterrupt:
raise
if __name__ == '__main__':
main()

Binary file not shown.

After

Width:  |  Height:  |  Size: 221 KiB

View File

@@ -0,0 +1,268 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<8> UDP_PROTOCOL = 0x11;
const bit<16> TYPE_IPV4 = 0x800;
const bit<5> IPV4_OPTION_MRI = 31;
#define MAX_HOPS 9
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
typedef bit<32> switchID_t;
typedef bit<32> qdepth_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
header ipv4_option_t {
bit<1> copyFlag;
bit<2> optClass;
bit<5> option;
bit<8> optionLength;
}
header mri_t {
bit<16> count;
}
header switch_t {
switchID_t swid;
qdepth_t qdepth;
}
struct ingress_metadata_t {
bit<16> count;
}
struct parser_metadata_t {
bit<16> remaining;
}
struct metadata {
ingress_metadata_t ingress_metadata;
parser_metadata_t parser_metadata;
}
struct headers {
ethernet_t ethernet;
ipv4_t ipv4;
ipv4_option_t ipv4_option;
mri_t mri;
switch_t[MAX_HOPS] swtraces;
}
error { IPHeaderTooShort }
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
verify(hdr.ipv4.ihl >= 5, error.IPHeaderTooShort);
transition select(hdr.ipv4.ihl) {
5 : accept;
default : parse_ipv4_option;
}
}
state parse_ipv4_option {
packet.extract(hdr.ipv4_option);
transition select(hdr.ipv4_option.option) {
IPV4_OPTION_MRI: parse_mri;
default: accept;
}
}
state parse_mri {
packet.extract(hdr.mri);
meta.parser_metadata.remaining = hdr.mri.count;
transition select(meta.parser_metadata.remaining) {
0 : accept;
default: parse_swtrace;
}
}
state parse_swtrace {
packet.extract(hdr.swtraces.next);
meta.parser_metadata.remaining = meta.parser_metadata.remaining - 1;
transition select(meta.parser_metadata.remaining) {
0 : accept;
default: parse_swtrace;
}
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
apply {
if (hdr.ipv4.isValid()) {
ipv4_lpm.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action add_swtrace(switchID_t swid) {
hdr.mri.count = hdr.mri.count + 1;
hdr.swtraces.push_front(1);
hdr.swtraces[0].swid = swid;
hdr.swtraces[0].qdepth = (qdepth_t)standard_metadata.deq_qdepth;
hdr.ipv4.ihl = hdr.ipv4.ihl + 2;
hdr.ipv4_option.optionLength = hdr.ipv4_option.optionLength + 8;
hdr.ipv4.totalLen = hdr.ipv4.totalLen + 8;
}
table swtrace {
actions = {
add_swtrace;
NoAction;
}
default_action = NoAction();
}
apply {
if (hdr.mri.isValid()) {
swtrace.apply();
}
}
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.ipv4);
packet.emit(hdr.ipv4_option);
packet.emit(hdr.mri);
packet.emit(hdr.swtraces);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,18 @@
{
"hosts": [
"h1",
"h2",
"h3",
"h11",
"h22"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["h11", "s1"], ["s1", "s2", "0", 0.5], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s2", "h22"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1,5 @@
BMV2_SWITCH_EXE = simple_switch_grpc
NO_P4 = true
P4C_ARGS = --p4runtime-file $(basename $@).p4info --p4runtime-format text
include ../../utils/Makefile

After

Width:  |  Height:  |  Size: 154 B

View File

@@ -0,0 +1,188 @@
# Implementing a Control Plane using P4 Runtime
## Introduction
In this exercise, we will be using P4 Runtime to send flow entries to the
switch instead of using the switch's CLI. We will be building on the same P4
program that you used in the [basic_tunnel](../basic_tunnel) exercise. The
P4 program has been renamed to `advanced_tunnel.py` and has been augmented
with two counters (`ingressTunnelCounter`, `egressTunnelCounter`) and
two new actions (`myTunnel_ingress`, `myTunnel_egress`).
You will use the starter program, `mycontroller.py`, and a few helper
libraries in the `p4runtime_lib` directory to create the table entries
necessary to tunnel traffic between host 1 and 2.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the
> reference.
## Step 1: Run the (incomplete) starter code
The starter code for this assignment is in a file called `mycontroller.py`,
and it will install only some of the rules that you need to tunnel traffic between
two hosts.
Let's first compile the new P4 program, start the network, use `mycontroller.py`
to install a few rules, and look at the `ingressTunnelCounter` to see that things
are working as expected.
1. In your shell, run:
```bash
make
```
This will:
* compile `advanced_tunnel.p4`,
* start a Mininet instance with three switches (`s1`, `s2`, `s3`)
configured in a triangle, each connected to one host (`h1`, `h2`, `h3`), and
* assign IPs of `10.0.1.1`, `10.0.2.2`, `10.0.3.3` to the respective hosts.
2. You should now see a Mininet command prompt. Start a ping between h1 and h2:
```bash
mininet> h1 ping h2
```
Because there are no rules on the switches, you should **not** receive any
replies yet. You should leave the ping running in this shell.
3. Open another shell and run the starter code:
```bash
cd ~/tutorials/P4D2_2017_Fall/exercises/p4runtime
./mycontroller.py
```
This will install the `advanced_tunnel.p4` program on the switches and push the
tunnel ingress rules.
The program prints the tunnel ingress and egress counters every 2 seconds.
You should see the ingress tunnel counter for s1 increasing:
```
s1 ingressTunnelCounter 100: 2 packets
```
The other counters should remain at zero.
4. Press `Ctrl-C` to the second shell to stop `mycontroller.py`
Each switch is currently mapping traffic into tunnels based on the destination IP
address. Your job is to write the rules that forward the traffic between the switches
based on the tunnel ID.
### Potential Issues
If you see the following error message when running `mycontroller.py`, then
the gRPC server is not running on one or more switches.
```
p4@p4:~/tutorials/P4D2_2017_Fall/exercises/p4runtime$ ./mycontroller.py
...
grpc._channel._Rendezvous: <_Rendezvous of RPC that terminated with (StatusCode.UNAVAILABLE, Connect Failed)>
```
You can check to see which of gRPC ports are listening on the machine by running:
```bash
sudo netstat -lpnt
```
The easiest solution is to enter `Ctrl-D` or `exit` in the `mininet>` prompt,
and re-run `make`.
### A note about the control plane
A P4 program defines a packet-processing pipeline, but the rules
within each table are inserted by the control plane. In this case,
`mycontroller.py` implements our control plane, instead of installing static
table entries like we have in the previous exercises.
**Important:** A P4 program also defines the interface between the
switch pipeline and control plane. This interface is defined in the
`advanced_tunnel.p4info` file. The table entries that you build in `mycontroller.py`
refer to specific tables, keys, and actions by name, and we use a P4Info helper
to convert the names into the IDs that are required for P4 Runtime. Any changes
in the P4 program that add or rename tables, keys, or actions will need to be
reflected in your table entries.
## Step 2: Implement Tunnel Forwarding
The `mycontroller.py` file is a basic controller plane that does the following:
1. Establishes a gRPC connection to the switches for the P4 Runtime service.
2. Pushes the P4 program to each switch.
3. Writes tunnel ingress and tunnel egress rules for two tunnels between h1 and h2.
4. Reads tunnel ingress and egress counters every 2 seconds.
It also contains comments marked with `TODO` which indicate the functionality
that you need to implement.
Your job will be to write the tunnel transit rule in the `writeTunnelRules` function
that will match on tunnel ID and forward packets to the next hop.
![topology](../basic_tunnel/topo.png)
In this exercise, you will be interacting with some of the classes and methods in
the `p4runtime_lib` directory. Here is a summary of each of the files in the directory:
- `helper.py`
- Contains the `P4InfoHelper` class which is used to parse the `p4info` files.
- Provides translation methods from entity name to and from ID number.
- Builds P4 program-dependent sections of P4 Runtime table entries.
- `switch.py`
- Contains the `SwitchConnection` class which grabs the gRPC client stub, and
establishes connections to the switches.
- Provides helper methods that construct the P4 Runtime protocol buffer messages
and makes the P4 Runtime gRPC service calls.
- `bmv2.py`
- Contains `Bmv2SwitchConnection` which extends `SwitchConnections` and provides
the BMv2-specific device payload to load the P4 program.
- `convert.py`
- Provides convenience methods to encode and decode from friendly strings and
numbers to the byte strings required for the protocol buffer messages.
- Used by `helper.py`
## Step 3: Run your solution
Follow the instructions from Step 1. If your Mininet network is still running,
you will just need to run the following in your second shell:
```bash
./my_controller.py
```
You should start to see ICMP replies in your Mininet prompt, and you should start to
see the values for all counters start to increment.
### Extra Credit and Food for Thought
You might notice that the rules that are printed by `mycontroller.py` contain the entity
IDs rather than the table names. You can use the P4Info helper to translate these IDs
into entry names.
Also, you may want to think about the following:
- What assumptions about the topology are baked into your implementation? How would you
need to change it for a more realistic network?
- Why are the byte counters different between the ingress and egress counters?
- What is the TTL in the ICMP replies? Why is it the value that it is?
Hint: The default TTL is 64 for packets sent by the hosts.
If you are interested, you can find the protocol buffer and gRPC definitions here:
- [P4 Runtime](https://github.com/p4lang/PI/blob/master/proto/p4/p4runtime.proto)
- [P4 Info](https://github.com/p4lang/PI/blob/master/proto/p4/config/p4info.proto)
#### Cleaning up Mininet
If the Mininet shell crashes, it may leave a Mininet instance
running in the background. Use the following command to clean up:
```bash
make clean
```
#### Running the reference solution
To run the reference solution, you should run the following command from the
`~/tutorials/P4D2_2017_Fall/exercises/p4runtime` directory:
```bash
solution/my_controller.py
```
## Next Steps
Congratulations, your implementation works! Move onto the next assignment
[ecn](../ecn)!

View File

@@ -0,0 +1,238 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_MYTUNNEL = 0x1212;
const bit<16> TYPE_IPV4 = 0x800;
const bit<32> MAX_TUNNEL_ID = 1 << 16;
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header myTunnel_t {
bit<16> proto_id;
bit<16> dst_id;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
myTunnel_t myTunnel;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_MYTUNNEL: parse_myTunnel;
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_myTunnel {
packet.extract(hdr.myTunnel);
transition select(hdr.myTunnel.proto_id) {
TYPE_IPV4: parse_ipv4;
default: accept;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
counter(MAX_TUNNEL_ID, CounterType.packets_and_bytes) ingressTunnelCounter;
counter(MAX_TUNNEL_ID, CounterType.packets_and_bytes) egressTunnelCounter;
action drop() {
mark_to_drop();
}
action ipv4_forward(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.srcAddr = hdr.ethernet.dstAddr;
hdr.ethernet.dstAddr = dstAddr;
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
action myTunnel_ingress(bit<16> dst_id) {
hdr.myTunnel.setValid();
hdr.myTunnel.dst_id = dst_id;
hdr.myTunnel.proto_id = hdr.ethernet.etherType;
hdr.ethernet.etherType = TYPE_MYTUNNEL;
ingressTunnelCounter.count((bit<32>) hdr.myTunnel.dst_id);
}
action myTunnel_forward(egressSpec_t port) {
standard_metadata.egress_spec = port;
}
action myTunnel_egress(macAddr_t dstAddr, egressSpec_t port) {
standard_metadata.egress_spec = port;
hdr.ethernet.dstAddr = dstAddr;
hdr.ethernet.etherType = hdr.myTunnel.proto_id;
hdr.myTunnel.setInvalid();
egressTunnelCounter.count((bit<32>) hdr.myTunnel.dst_id);
}
table ipv4_lpm {
key = {
hdr.ipv4.dstAddr: lpm;
}
actions = {
ipv4_forward;
myTunnel_ingress;
drop;
NoAction;
}
size = 1024;
default_action = NoAction();
}
table myTunnel_exact {
key = {
hdr.myTunnel.dst_id: exact;
}
actions = {
myTunnel_forward;
myTunnel_egress;
drop;
}
size = 1024;
default_action = drop();
}
apply {
if (hdr.ipv4.isValid() && !hdr.myTunnel.isValid()) {
// Process only non-tunneled IPv4 packets.
ipv4_lpm.apply();
}
if (hdr.myTunnel.isValid()) {
// Process all tunneled packets.
myTunnel_exact.apply();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply {
update_checksum(
hdr.ipv4.isValid(),
{ hdr.ipv4.version,
hdr.ipv4.ihl,
hdr.ipv4.diffserv,
hdr.ipv4.totalLen,
hdr.ipv4.identification,
hdr.ipv4.flags,
hdr.ipv4.fragOffset,
hdr.ipv4.ttl,
hdr.ipv4.protocol,
hdr.ipv4.srcAddr,
hdr.ipv4.dstAddr },
hdr.ipv4.hdrChecksum,
HashAlgorithm.csum16);
}
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.myTunnel);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,183 @@
#!/usr/bin/env python2
import argparse
import os
from time import sleep
import p4runtime_lib.bmv2
import p4runtime_lib.helper
SWITCH_TO_HOST_PORT = 1
SWITCH_TO_SWITCH_PORT = 2
def writeTunnelRules(p4info_helper, ingress_sw, egress_sw, tunnel_id,
dst_eth_addr, dst_ip_addr):
'''
Installs three rules:
1) An tunnel ingress rule on the ingress switch in the ipv4_lpm table that
encapsulates traffic into a tunnel with the specified ID
2) A transit rule on the ingress switch that forwards traffic based on
the specified ID
3) An tunnel egress rule on the egress switch that decapsulates traffic
with the specified ID and sends it to the host
:param p4info_helper: the P4Info helper
:param ingress_sw: the ingress switch connection
:param egress_sw: the egress switch connection
:param tunnel_id: the specified tunnel ID
:param dst_eth_addr: the destination IP to match in the ingress rule
:param dst_ip_addr: the destination Ethernet address to write in the
egress rule
'''
# 1) Tunnel Ingress Rule
table_entry = p4info_helper.buildTableEntry(
table_name="MyIngress.ipv4_lpm",
match_fields={
"hdr.ipv4.dstAddr": (dst_ip_addr, 32)
},
action_name="MyIngress.myTunnel_ingress",
action_params={
"dst_id": tunnel_id,
})
ingress_sw.WriteTableEntry(table_entry)
print "Installed ingress tunnel rule on %s" % ingress_sw.name
# 2) Tunnel Transit Rule
# The rule will need to be added to the myTunnel_exact table and match on
# the tunnel ID (hdr.myTunnel.dst_id). Traffic will need to be forwarded
# using the myTunnel_forward action on the port connected to the next switch.
#
# For our simple topology, switch 1 and switch 2 are connected using a
# link attached to port 2 on both switches. We have defined a variable at
# the top of the file, SWITCH_TO_SWITCH_PORT, that you can use as the output
# port for this action.
#
# We will only need a transit rule on the ingress switch because we are
# using a simple topology. In general, you'll need on transit rule for
# each switch in the path (except the last switch, which has the egress rule),
# and you will need to select the port dynamically for each switch based on
# your topology.
# TODO build the transit rule
# TODO install the transit rule on the ingress switch
print "TODO Install transit tunnel rule"
# 3) Tunnel Egress Rule
# For our simple topology, the host will always be located on the
# SWITCH_TO_HOST_PORT (port 1).
# In general, you will need to keep track of which port the host is
# connected to.
table_entry = p4info_helper.buildTableEntry(
table_name="MyIngress.myTunnel_exact",
match_fields={
"hdr.myTunnel.dst_id": tunnel_id
},
action_name="MyIngress.myTunnel_egress",
action_params={
"dstAddr": dst_eth_addr,
"port": SWITCH_TO_HOST_PORT
})
egress_sw.WriteTableEntry(table_entry)
print "Installed egress tunnel rule on %s" % egress_sw.name
def readTableRules(p4info_helper, sw):
'''
Reads the table entries from all tables on the switch.
:param p4info_helper: the P4Info helper
:param sw: the switch connection
'''
print '\n----- Reading tables rules for %s -----' % sw.name
for response in sw.ReadTableEntries():
for entity in response.entities:
entry = entity.table_entry
# TODO For extra credit, you can use the p4info_helper to translate
# the IDs the entry to names
print entry
print '-----'
def printCounter(p4info_helper, sw, counter_name, index):
'''
Reads the specified counter at the specified index from the switch. In our
program, the index is the tunnel ID. If the index is 0, it will return all
values from the counter.
:param p4info_helper: the P4Info helper
:param sw: the switch connection
:param counter_name: the name of the counter from the P4 program
:param index: the counter index (in our case, the tunnel ID)
'''
for response in sw.ReadCounters(p4info_helper.get_counters_id(counter_name), index):
for entity in response.entities:
counter = entity.counter_entry
print "%s %s %d: %d packets (%d bytes)" % (
sw.name, counter_name, index,
counter.data.packet_count, counter.data.byte_count
)
def main(p4info_file_path, bmv2_file_path):
# Instantiate a P4 Runtime helper from the p4info file
p4info_helper = p4runtime_lib.helper.P4InfoHelper(p4info_file_path)
# Create a switch connection object for s1 and s2;
# this is backed by a P4 Runtime gRPC connection
s1 = p4runtime_lib.bmv2.Bmv2SwitchConnection('s1',
address='127.0.0.1:50051',
device_id=0)
s2 = p4runtime_lib.bmv2.Bmv2SwitchConnection('s2',
address='127.0.0.1:50052',
device_id=1)
# Install the P4 program on the switches
s1.SetForwardingPipelineConfig(p4info=p4info_helper.p4info,
bmv2_json_file_path=bmv2_file_path)
print "Installed P4 Program using SetForwardingPipelineConfig on %s" % s1.name
s2.SetForwardingPipelineConfig(p4info=p4info_helper.p4info,
bmv2_json_file_path=bmv2_file_path)
print "Installed P4 Program using SetForwardingPipelineConfig on %s" % s2.name
# Write the rules that tunnel traffic from h1 to h2
writeTunnelRules(p4info_helper, ingress_sw=s1, egress_sw=s2, tunnel_id=100,
dst_eth_addr="00:00:00:00:02:02", dst_ip_addr="10.0.2.2")
# Write the rules that tunnel traffic from h2 to h1
writeTunnelRules(p4info_helper, ingress_sw=s2, egress_sw=s1, tunnel_id=200,
dst_eth_addr="00:00:00:00:01:01", dst_ip_addr="10.0.1.1")
# TODO Uncomment the following two lines to read table entries from s1 and s2
#readTableRules(p4info_helper, s1)
#readTableRules(p4info_helper, s2)
# Print the tunnel counters every 2 seconds
try:
while True:
sleep(2)
print '\n----- Reading tunnel counters -----'
printCounter(p4info_helper, s1, "MyIngress.ingressTunnelCounter", 100)
printCounter(p4info_helper, s2, "MyIngress.egressTunnelCounter", 100)
printCounter(p4info_helper, s2, "MyIngress.ingressTunnelCounter", 200)
printCounter(p4info_helper, s1, "MyIngress.egressTunnelCounter", 200)
except KeyboardInterrupt:
print " Shutting down."
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='P4Runtime Controller')
parser.add_argument('--p4info', help='p4info proto in text format from p4c',
type=str, action="store", required=False,
default='./build/advanced_tunnel.p4info')
parser.add_argument('--bmv2-json', help='BMv2 JSON file from p4c',
type=str, action="store", required=False,
default='./build/advanced_tunnel.json')
args = parser.parse_args()
if not os.path.exists(args.p4info):
parser.print_help()
print "\np4info file not found: %s\nHave you run 'make'?" % args.p4info
parser.exit(1)
if not os.path.exists(args.bmv2_json):
parser.print_help()
print "\nBMv2 JSON file not found: %s\nHave you run 'make'?" % args.bmv2_json
parser.exit(1)
main(args.p4info, args.bmv2_json)

View File

@@ -0,0 +1,30 @@
# Copyright 2017-present Open Networking Foundation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
from switch import SwitchConnection
from p4.tmp import p4config_pb2
def buildDeviceConfig(bmv2_json_file_path=None):
"Builds the device config for BMv2"
device_config = p4config_pb2.P4DeviceConfig()
device_config.reassign = True
with open(bmv2_json_file_path) as f:
device_config.device_data = f.read()
return device_config
class Bmv2SwitchConnection(SwitchConnection):
def buildDeviceConfig(self, **kwargs):
return buildDeviceConfig(**kwargs)

View File

@@ -0,0 +1,119 @@
# Copyright 2017-present Open Networking Foundation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import re
import socket
import math
'''
This package contains several helper functions for encoding to and decoding from byte strings:
- integers
- IPv4 address strings
- Ethernet address strings
'''
mac_pattern = re.compile('^([\da-fA-F]{2}:){5}([\da-fA-F]{2})$')
def matchesMac(mac_addr_string):
return mac_pattern.match(mac_addr_string) is not None
def encodeMac(mac_addr_string):
return mac_addr_string.replace(':', '').decode('hex')
def decodeMac(encoded_mac_addr):
return ':'.join(s.encode('hex') for s in encoded_mac_addr)
ip_pattern = re.compile('^(\d{1,3}\.){3}(\d{1,3})$')
def matchesIPv4(ip_addr_string):
return ip_pattern.match(ip_addr_string) is not None
def encodeIPv4(ip_addr_string):
return socket.inet_aton(ip_addr_string)
def decodeIPv4(encoded_ip_addr):
return socket.inet_ntoa(encoded_ip_addr)
def bitwidthToBytes(bitwidth):
return int(math.ceil(bitwidth / 8.0))
def encodeNum(number, bitwidth):
byte_len = bitwidthToBytes(bitwidth)
num_str = '%x' % number
if number >= 2 ** bitwidth:
raise Exception("Number, %d, does not fit in %d bits" % (number, bitwidth))
return ('0' * (byte_len * 2 - len(num_str)) + num_str).decode('hex')
def decodeNum(encoded_number):
return int(encoded_number.encode('hex'), 16)
def encode(x, bitwidth):
'Tries to infer the type of `x` and encode it'
byte_len = bitwidthToBytes(bitwidth)
if (type(x) == list or type(x) == tuple) and len(x) == 1:
x = x[0]
encoded_bytes = None
if type(x) == str:
if matchesMac(x):
encoded_bytes = encodeMac(x)
elif matchesIPv4(x):
encoded_bytes = encodeIPv4(x)
else:
# Assume that the string is already encoded
encoded_bytes = x
elif type(x) == int:
encoded_bytes = encodeNum(x, bitwidth)
else:
raise Exception("Encoding objects of %r is not supported" % type(x))
assert(len(encoded_bytes) == byte_len)
return encoded_bytes
if __name__ == '__main__':
# TODO These tests should be moved out of main eventually
mac = "aa:bb:cc:dd:ee:ff"
enc_mac = encodeMac(mac)
assert(enc_mac == '\xaa\xbb\xcc\xdd\xee\xff')
dec_mac = decodeMac(enc_mac)
assert(mac == dec_mac)
ip = "10.0.0.1"
enc_ip = encodeIPv4(ip)
assert(enc_ip == '\x0a\x00\x00\x01')
dec_ip = decodeIPv4(enc_ip)
assert(ip == dec_ip)
num = 1337
byte_len = 5
enc_num = encodeNum(num, byte_len * 8)
assert(enc_num == '\x00\x00\x00\x05\x39')
dec_num = decodeNum(enc_num)
assert(num == dec_num)
assert(matchesIPv4('10.0.0.1'))
assert(not matchesIPv4('10.0.0.1.5'))
assert(not matchesIPv4('1000.0.0.1'))
assert(not matchesIPv4('10001'))
assert(encode(mac, 6 * 8) == enc_mac)
assert(encode(ip, 4 * 8) == enc_ip)
assert(encode(num, 5 * 8) == enc_num)
assert(encode((num,), 5 * 8) == enc_num)
assert(encode([num], 5 * 8) == enc_num)
num = 256
byte_len = 2
try:
enc_num = encodeNum(num, 8)
raise Exception("expected exception")
except Exception as e:
print e

View File

@@ -0,0 +1,183 @@
# Copyright 2017-present Open Networking Foundation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import re
import google.protobuf.text_format
from p4 import p4runtime_pb2
from p4.config import p4info_pb2
from p4runtime_lib.convert import encode
class P4InfoHelper(object):
def __init__(self, p4_info_filepath):
p4info = p4info_pb2.P4Info()
# Load the p4info file into a skeleton P4Info object
with open(p4_info_filepath) as p4info_f:
google.protobuf.text_format.Merge(p4info_f.read(), p4info)
self.p4info = p4info
def get(self, entity_type, name=None, id=None):
if name is not None and id is not None:
raise AssertionError("name or id must be None")
for o in getattr(self.p4info, entity_type):
pre = o.preamble
if name:
if (pre.name == name or pre.alias == name):
return o
else:
if pre.id == id:
return o
if name:
raise AttributeError("Could not find %r of type %s" % (name, entity_type))
else:
raise AttributeError("Could not find id %r of type %s" % (id, entity_type))
def get_id(self, entity_type, name):
return self.get(entity_type, name=name).preamble.id
def get_name(self, entity_type, id):
return self.get(entity_type, id=id).preamble.name
def get_alias(self, entity_type, id):
return self.get(entity_type, id=id).preamble.alias
def __getattr__(self, attr):
# Synthesize convenience functions for name to id lookups for top-level entities
# e.g. get_tables_id(name_string) or get_actions_id(name_string)
m = re.search("^get_(\w+)_id$", attr)
if m:
primitive = m.group(1)
return lambda name: self.get_id(primitive, name)
# Synthesize convenience functions for id to name lookups
# e.g. get_tables_name(id) or get_actions_name(id)
m = re.search("^get_(\w+)_name$", attr)
if m:
primitive = m.group(1)
return lambda id: self.get_name(primitive, id)
raise AttributeError("%r object has no attribute %r" % (self.__class__, attr))
def get_match_field(self, table_name, name=None, id=None):
for t in self.p4info.tables:
pre = t.preamble
if pre.name == table_name:
for mf in t.match_fields:
if name is not None:
if mf.name == name:
return mf
elif id is not None:
if mf.id == id:
return mf
raise AttributeError("%r has no attribute %r" % (table_name, name if name is not None else id))
def get_match_field_id(self, table_name, match_field_name):
return self.get_match_field(table_name, name=match_field_name).id
def get_match_field_name(self, table_name, match_field_id):
return self.get_match_field(table_name, id=match_field_id).name
def get_match_field_pb(self, table_name, match_field_name, value):
p4info_match = self.get_match_field(table_name, match_field_name)
bitwidth = p4info_match.bitwidth
p4runtime_match = p4runtime_pb2.FieldMatch()
p4runtime_match.field_id = p4info_match.id
match_type = p4info_match.match_type
if match_type == p4info_pb2.MatchField.VALID:
valid = p4runtime_match.valid
valid.value = bool(value)
elif match_type == p4info_pb2.MatchField.EXACT:
exact = p4runtime_match.exact
exact.value = encode(value, bitwidth)
elif match_type == p4info_pb2.MatchField.LPM:
lpm = p4runtime_match.lpm
lpm.value = encode(value[0], bitwidth)
lpm.prefix_len = value[1]
elif match_type == p4info_pb2.MatchField.TERNARY:
lpm = p4runtime_match.ternary
lpm.value = encode(value[0], bitwidth)
lpm.mask = encode(value[1], bitwidth)
elif match_type == p4info_pb2.MatchField.RANGE:
lpm = p4runtime_match.range
lpm.low = encode(value[0], bitwidth)
lpm.high = encode(value[1], bitwidth)
else:
raise Exception("Unsupported match type with type %r" % match_type)
return p4runtime_match
def get_match_field_value(self, match_field):
match_type = match_field.WhichOneof("field_match_type")
if match_type == 'valid':
return match_field.valid.value
elif match_type == 'exact':
return match_field.exact.value
elif match_type == 'lpm':
return (match_field.lpm.value, match_field.lpm.prefix_len)
elif match_type == 'ternary':
return (match_field.ternary.value, match_field.ternary.mask)
elif match_type == 'range':
return (match_field.range.low, match_field.range.high)
else:
raise Exception("Unsupported match type with type %r" % match_type)
def get_action_param(self, action_name, name=None, id=None):
for a in self.p4info.actions:
pre = a.preamble
if pre.name == action_name:
for p in a.params:
if name is not None:
if p.name == name:
return p
elif id is not None:
if p.id == id:
return p
raise AttributeError("action %r has no param %r" % (action_name, name if name is not None else id))
def get_action_param_id(self, action_name, param_name):
return self.get_action_param(action_name, name=param_name).id
def get_action_param_name(self, action_name, param_id):
return self.get_action_param(action_name, id=param_id).name
def get_action_param_pb(self, action_name, param_name, value):
p4info_param = self.get_action_param(action_name, param_name)
p4runtime_param = p4runtime_pb2.Action.Param()
p4runtime_param.param_id = p4info_param.id
p4runtime_param.value = encode(value, p4info_param.bitwidth)
return p4runtime_param
def buildTableEntry(self,
table_name,
match_fields={},
action_name=None,
action_params={}):
table_entry = p4runtime_pb2.TableEntry()
table_entry.table_id = self.get_tables_id(table_name)
if match_fields:
table_entry.match.extend([
self.get_match_field_pb(table_name, match_field_name, value)
for match_field_name, value in match_fields.iteritems()
])
if action_name:
action = table_entry.action.action
action.action_id = self.get_actions_id(action_name)
if action_params:
action.params.extend([
self.get_action_param_pb(action_name, field_name, value)
for field_name, value in action_params.iteritems()
])
return table_entry

View File

@@ -0,0 +1,88 @@
# Copyright 2017-present Open Networking Foundation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
from abc import abstractmethod
import grpc
from p4 import p4runtime_pb2
from p4.tmp import p4config_pb2
class SwitchConnection(object):
def __init__(self, name, address='127.0.0.1:50051', device_id=0):
self.name = name
self.address = address
self.device_id = device_id
self.p4info = None
self.channel = grpc.insecure_channel(self.address)
self.client_stub = p4runtime_pb2.P4RuntimeStub(self.channel)
@abstractmethod
def buildDeviceConfig(self, **kwargs):
return p4config_pb2.P4DeviceConfig()
def SetForwardingPipelineConfig(self, p4info, dry_run=False, **kwargs):
device_config = self.buildDeviceConfig(**kwargs)
request = p4runtime_pb2.SetForwardingPipelineConfigRequest()
request.device_id = self.device_id
config = request.config
config.p4info.CopyFrom(p4info)
config.p4_device_config = device_config.SerializeToString()
request.action = p4runtime_pb2.SetForwardingPipelineConfigRequest.VERIFY_AND_COMMIT
if dry_run:
print "P4 Runtime SetForwardingPipelineConfig:", request
else:
self.client_stub.SetForwardingPipelineConfig(request)
def WriteTableEntry(self, table_entry, dry_run=False):
request = p4runtime_pb2.WriteRequest()
request.device_id = self.device_id
update = request.updates.add()
update.type = p4runtime_pb2.Update.INSERT
update.entity.table_entry.CopyFrom(table_entry)
if dry_run:
print "P4 Runtime Write:", request
else:
self.client_stub.Write(request)
def ReadTableEntries(self, table_id=None, dry_run=False):
request = p4runtime_pb2.ReadRequest()
request.device_id = self.device_id
entity = request.entities.add()
table_entry = entity.table_entry
if table_id is not None:
table_entry.table_id = table_id
else:
table_entry.table_id = 0
if dry_run:
print "P4 Runtime Read:", request
else:
for response in self.client_stub.Read(request):
yield response
def ReadCounters(self, counter_id=None, index=None, dry_run=False):
request = p4runtime_pb2.ReadRequest()
request.device_id = self.device_id
entity = request.entities.add()
counter_entry = entity.counter_entry
if counter_id is not None:
counter_entry.counter_id = counter_id
else:
counter_entry.counter_id = 0
if index is not None:
counter_entry.index = index
if dry_run:
print "P4 Runtime Read:", request
else:
for response in self.client_stub.Read(request):
yield response

View File

@@ -0,0 +1,206 @@
#!/usr/bin/env python2
import argparse
import os
from time import sleep
# NOTE: Appending to the PYTHON_PATH is only required in the `solution` directory.
# It is not required for mycontroller.py in the top-level directory.
import sys
sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
import p4runtime_lib.bmv2
import p4runtime_lib.helper
SWITCH_TO_HOST_PORT = 1
SWITCH_TO_SWITCH_PORT = 2
def writeTunnelRules(p4info_helper, ingress_sw, egress_sw, tunnel_id,
dst_eth_addr, dst_ip_addr):
'''
Installs three rules:
1) An tunnel ingress rule on the ingress switch in the ipv4_lpm table that
encapsulates traffic into a tunnel with the specified ID
2) A transit rule on the ingress switch that forwards traffic based on
the specified ID
3) An tunnel egress rule on the egress switch that decapsulates traffic
with the specified ID and sends it to the host
:param p4info_helper: the P4Info helper
:param ingress_sw: the ingress switch connection
:param egress_sw: the egress switch connection
:param tunnel_id: the specified tunnel ID
:param dst_eth_addr: the destination IP to match in the ingress rule
:param dst_ip_addr: the destination Ethernet address to write in the
egress rule
'''
# 1) Tunnel Ingress Rule
table_entry = p4info_helper.buildTableEntry(
table_name="MyIngress.ipv4_lpm",
match_fields={
"hdr.ipv4.dstAddr": (dst_ip_addr, 32)
},
action_name="MyIngress.myTunnel_ingress",
action_params={
"dst_id": tunnel_id,
})
ingress_sw.WriteTableEntry(table_entry)
print "Installed ingress tunnel rule on %s" % ingress_sw.name
# 2) Tunnel Transit Rule
# The rule will need to be added to the myTunnel_exact table and match on
# the tunnel ID (hdr.myTunnel.dst_id). Traffic will need to be forwarded
# using the myTunnel_forward action on the port connected to the next switch.
#
# For our simple topology, switch 1 and switch 2 are connected using a
# link attached to port 2 on both switches. We have defined a variable at
# the top of the file, SWITCH_TO_SWITCH_PORT, that you can use as the output
# port for this action.
#
# We will only need a transit rule on the ingress switch because we are
# using a simple topology. In general, you'll need on transit rule for
# each switch in the path (except the last switch, which has the egress rule),
# and you will need to select the port dynamically for each switch based on
# your topology.
table_entry = p4info_helper.buildTableEntry(
table_name="MyIngress.myTunnel_exact",
match_fields={
"hdr.myTunnel.dst_id": tunnel_id
},
action_name="MyIngress.myTunnel_forward",
action_params={
"port": SWITCH_TO_SWITCH_PORT
})
ingress_sw.WriteTableEntry(table_entry)
print "Installed transit tunnel rule on %s" % ingress_sw.name
# 3) Tunnel Egress Rule
# For our simple topology, the host will always be located on the
# SWITCH_TO_HOST_PORT (port 1).
# In general, you will need to keep track of which port the host is
# connected to.
table_entry = p4info_helper.buildTableEntry(
table_name="MyIngress.myTunnel_exact",
match_fields={
"hdr.myTunnel.dst_id": tunnel_id
},
action_name="MyIngress.myTunnel_egress",
action_params={
"dstAddr": dst_eth_addr,
"port": SWITCH_TO_HOST_PORT
})
egress_sw.WriteTableEntry(table_entry)
print "Installed egress tunnel rule on %s" % egress_sw.name
def readTableRules(p4info_helper, sw):
'''
Reads the table entries from all tables on the switch.
:param p4info_helper: the P4Info helper
:param sw: the switch connection
'''
print '\n----- Reading tables rules for %s -----' % sw.name
for response in sw.ReadTableEntries():
for entity in response.entities:
entry = entity.table_entry
# TODO For extra credit, you can use the p4info_helper to translate
# the IDs the entry to names
table_name = p4info_helper.get_tables_name(entry.table_id)
print '%s: ' % table_name,
for m in entry.match:
print p4info_helper.get_match_field_name(table_name, m.field_id),
print '%r' % (p4info_helper.get_match_field_value(m),),
action = entry.action.action
action_name = p4info_helper.get_actions_name(action.action_id)
print '->', action_name,
for p in action.params:
print p4info_helper.get_action_param_name(action_name, p.param_id),
print '%r' % p.value,
print
def printCounter(p4info_helper, sw, counter_name, index):
'''
Reads the specified counter at the specified index from the switch. In our
program, the index is the tunnel ID. If the index is 0, it will return all
values from the counter.
:param p4info_helper: the P4Info helper
:param sw: the switch connection
:param counter_name: the name of the counter from the P4 program
:param index: the counter index (in our case, the tunnel ID)
'''
for response in sw.ReadCounters(p4info_helper.get_counters_id(counter_name), index):
for entity in response.entities:
counter = entity.counter_entry
print "%s %s %d: %d packets (%d bytes)" % (
sw.name, counter_name, index,
counter.data.packet_count, counter.data.byte_count
)
def main(p4info_file_path, bmv2_file_path):
# Instantiate a P4 Runtime helper from the p4info file
p4info_helper = p4runtime_lib.helper.P4InfoHelper(p4info_file_path)
# Create a switch connection object for s1 and s2;
# this is backed by a P4 Runtime gRPC connection
s1 = p4runtime_lib.bmv2.Bmv2SwitchConnection('s1',
address='127.0.0.1:50051',
device_id=0)
s2 = p4runtime_lib.bmv2.Bmv2SwitchConnection('s2',
address='127.0.0.1:50052',
device_id=1)
# Install the P4 program on the switches
s1.SetForwardingPipelineConfig(p4info=p4info_helper.p4info,
bmv2_json_file_path=bmv2_file_path)
print "Installed P4 Program using SetForwardingPipelineConfig on %s" % s1.name
s2.SetForwardingPipelineConfig(p4info=p4info_helper.p4info,
bmv2_json_file_path=bmv2_file_path)
print "Installed P4 Program using SetForwardingPipelineConfig on %s" % s2.name
# Write the rules that tunnel traffic from h1 to h2
writeTunnelRules(p4info_helper, ingress_sw=s1, egress_sw=s2, tunnel_id=100,
dst_eth_addr="00:00:00:00:02:02", dst_ip_addr="10.0.2.2")
# Write the rules that tunnel traffic from h2 to h1
writeTunnelRules(p4info_helper, ingress_sw=s2, egress_sw=s1, tunnel_id=200,
dst_eth_addr="00:00:00:00:01:01", dst_ip_addr="10.0.1.1")
# TODO Uncomment the following two lines to read table entries from s1 and s2
readTableRules(p4info_helper, s1)
readTableRules(p4info_helper, s2)
# Print the tunnel counters every 2 seconds
try:
while True:
sleep(2)
print '\n----- Reading tunnel counters -----'
printCounter(p4info_helper, s1, "MyIngress.ingressTunnelCounter", 100)
printCounter(p4info_helper, s2, "MyIngress.egressTunnelCounter", 100)
printCounter(p4info_helper, s2, "MyIngress.ingressTunnelCounter", 200)
printCounter(p4info_helper, s1, "MyIngress.egressTunnelCounter", 200)
except KeyboardInterrupt:
print " Shutting down."
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='P4Runtime Controller')
parser.add_argument('--p4info', help='p4info proto in text format from p4c',
type=str, action="store", required=False,
default='./build/advanced_tunnel.p4info')
parser.add_argument('--bmv2-json', help='BMv2 JSON file from p4c',
type=str, action="store", required=False,
default='./build/advanced_tunnel.json')
args = parser.parse_args()
if not os.path.exists(args.p4info):
parser.print_help()
print "\np4info file not found: %s\nHave you run 'make'?" % args.p4info
parser.exit(1)
if not os.path.exists(args.bmv2_json):
parser.print_help()
print "\nBMv2 JSON file not found: %s\nHave you run 'make'?" % args.bmv2_json
parser.exit(1)
main(args.p4info, args.bmv2_json)

View File

@@ -0,0 +1,16 @@
{
"hosts": [
"h1",
"h2",
"h3"
],
"switches": {
"s1": {},
"s2": {},
"s3": {}
},
"links": [
["h1", "s1"], ["s1", "s2"], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s3", "h3"]
]
}

View File

@@ -0,0 +1 @@
include ../../utils/Makefile

View File

@@ -0,0 +1,147 @@
# Implementing Source Routing
## Introduction
The objective of this exercise is to implement source routing. With
source routing, the source host guides each switch in the network to
send the packet to a specific port. The host puts a stack of output
ports in the packet. In this example, we just put the stack after
Ethernet header and select a special etherType to indicate that. Each
switch pops an item from the stack and forwards the packet according
to the specified port number.
Your switch must parse the source routing stack. Each item has a bos
(bottom of stack) bit and a port number. The bos bit is 1 only for the
last entry of stack. Then at ingress, it should pop an entry from the
stack and set the egress port accordingly. Note that the last hop can
also revert back the etherType to `TYPE_IPV4`.
> **Spoiler alert:** There is a reference solution in the `solution`
> sub-directory. Feel free to compare your implementation to the
> reference.
## Step 1: Run the (incomplete) starter code
The directory with this README also contains a skeleton P4 program,
`source_routing.p4`, which initially drops all packets. Your job (in
the next step) will be to extend it to properly to route packets.
Before that, let's compile the incomplete `source_routing.p4` and
bring up a network in Mininet to test its behavior.
1. In your shell, run:
```bash
make
```
This will:
* compile `source_routing.p4`, and
* start a Mininet instance with three switches (`s1`, `s2`, `s3`) configured
in a triangle, each connected to one host (`h1`, `h2`, `h3`).
Check the network topology using the `net` command in mininet.
You can also change the topology in topology.json
* The hosts are assigned IPs of `10.0.1.1`, `10.0.2.2`, etc
(`10.0.<Switchid>.<hostID>`).
2. You should now see a Mininet command prompt. Open two terminals for
`h1` and `h2`, respectively:
```bash
mininet> xterm h1 h2
```
3. Each host includes a small Python-based messaging client and
server. In `h2`'s xterm, start the server:
```bash
./receive.py
```
4. In `h1`'s xterm, send a message from the client:
```bash
./send.py 10.0.2.2
```
5. Type a list of port numbers. say `2 3 2 2 1`. This should send the
packet through `h1`, `s1`, `s2`, `s3`, `s1`, `s2`, and
`h2`. However, `h2` will not receive the message.
6. Type `q` to exit send.py and type `exit` to leave each xterm and
the Mininet command line.
The message was not received because each switch is programmed with
`source_routing.p4`, which drops all packets on arrival. You can
verify this by looking at `build/logs/s1.log`. Your job is to extend
the P4 code so packets are delivered to their destination.
## Step 2: Implement source routing
The `source_routing.p4` file contains a skeleton P4 program with key
pieces of logic replaced by `TODO` comments. These should guide your
implementation---replace each `TODO` with logic implementing the
missing piece.
A complete `source_routing.p4` will contain the following components:
1. Header type definitions for Ethernet (`ethernet_t`) and IPv4
(`ipv4_t`) and Source Route (`srcRoute_t`).
2. **TODO:** Parsers for Ethernet and Source Route that populate
`ethernet` and `srcRoutes` fields.
3. An action to drop a packet, using `mark_to_drop()`.
4. **TODO:** An action (called `srcRoute_nhop`), which will:
1. Set the egress port for the next hop.
2. remove the first entry of srcRoutes
5. A control with an `apply` block that:
1. checks the existence of source routes.
2. **TODO:** if statement to change etherent.etherType if it is the last hop
3. **TODO:** call srcRoute_nhop action
6. A deparser that selects the order in which fields inserted into the outgoing
packet.
7. A `package` instantiation supplied with the parser, control, and deparser.
> In general, a package also requires instances of checksum verification
> and recomputation controls. These are not necessary for this tutorial
> and are replaced with instantiations of empty controls.
## Step 3: Run your solution
Follow the instructions from Step 1. This time, your message from `h1`
should be delivered to `h2`.
Check the `ttl` of the IP header. Each hop decrements `ttl`. The port
sequence `2 3 2 2 1`, forces the packet to have a loop, so the `ttl`
should be 59 at `h2`. Can you find the port sequence for the shortest
path?
### Food for thought
* Can we change the program to handle both IPv4 forwarding and source
routing at the same time?
* How would you enhance your program to let the first switch add the
path, so that source routing would be transparent to end-hosts?
### Troubleshooting
There are several ways that problems might manifest:
1. `source_routing.p4` fails to compile. In this case, `make` will
report the error emitted from the compiler and stop.
2. `source_routing.p4` compiles but switches or mininet do not start.
Do you have another instance of mininet running? Did the previous
run of mininet crash? if yes, check "Cleaning up Mininet" bellow.
3. `source_routing.p4` compiles but the switch does not process
packets in the desired way. The `/tmp/p4s.<switch-name>.log`
files contain trace messages describing how each switch processes
each packet. The output is detailed and can help pinpoint logic
errors in your implementation. The
`build/<switch-name>-<interface-name>.pcap` also contains the pcap
of packets on each interface. Use `tcpdump -r <filename> -xxx` to
print the hexdump of the packets.
#### Cleaning up Mininet
In the cases above, `make` may leave a Mininet instance running in
the background. Use the following command to clean up these
instances:
```bash
mn -c
```
## Next Steps
Congratulations, your implementation works! Move on to
[Calculator](../calc).

View File

@@ -0,0 +1,59 @@
#!/usr/bin/env python
import sys
import struct
from scapy.all import sniff, sendp, hexdump, get_if_list, get_if_hwaddr, bind_layers
from scapy.all import Packet, IPOption
from scapy.all import IP, UDP, Raw, Ether
from scapy.layers.inet import _IPOption_HDR
from scapy.fields import *
def get_if():
ifs=get_if_list()
iface=None
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class IPOption_MRI(IPOption):
name = "MRI"
option = 31
fields_desc = [ _IPOption_HDR,
FieldLenField("length", None, fmt="B",
length_of="swids",
adjust=lambda pkt,l:l+4),
ShortField("count", 0),
FieldListField("swids",
[],
IntField("", 0),
length_from=lambda pkt:pkt.count*4) ]
def handle_pkt(pkt):
print "got a packet"
pkt.show2()
# hexdump(pkt)
sys.stdout.flush()
class SourceRoute(Packet):
fields_desc = [ BitField("bos", 0, 1),
BitField("port", 0, 15)]
class SourceRoutingTail(Packet):
fields_desc = [ XShortField("etherType", 0x800)]
bind_layers(Ether, SourceRoute, type=0x1234)
bind_layers(SourceRoute, SourceRoute, bos=0)
bind_layers(SourceRoute, SourceRoutingTail, bos=1)
def main():
iface = 'h2-eth0'
print "sniffing on %s" % iface
sys.stdout.flush()
sniff(filter="udp and port 4321", iface = iface,
prn = lambda x: handle_pkt(x))
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,73 @@
#!/usr/bin/env python
import argparse
import sys
import socket
import random
import struct
from scapy.all import sendp, send, get_if_list, get_if_hwaddr, bind_layers
from scapy.all import Packet
from scapy.all import Ether, IP, UDP
from scapy.fields import *
import readline
def get_if():
ifs=get_if_list()
iface=None # "h1-eth0"
for i in get_if_list():
if "eth0" in i:
iface=i
break;
if not iface:
print "Cannot find eth0 interface"
exit(1)
return iface
class SourceRoute(Packet):
fields_desc = [ BitField("bos", 0, 1),
BitField("port", 0, 15)]
bind_layers(Ether, SourceRoute, type=0x1234)
bind_layers(SourceRoute, SourceRoute, bos=0)
bind_layers(SourceRoute, IP, bos=1)
def main():
if len(sys.argv)<2:
print 'pass 2 arguments: <destination>'
exit(1)
addr = socket.gethostbyname(sys.argv[1])
iface = get_if()
print "sending on interface %s to %s" % (iface, str(addr))
while True:
print
s = str(raw_input('Type space separated port nums '
'(example: "2 3 2 2 1") or "q" to quit: '))
if s == "q":
break;
print
i = 0
pkt = Ether(src=get_if_hwaddr(iface), dst='ff:ff:ff:ff:ff:ff');
for p in s.split(" "):
try:
pkt = pkt / SourceRoute(bos=0, port=int(p))
i = i+1
except ValueError:
pass
if pkt.haslayer(SourceRoute):
pkt.getlayer(SourceRoute, i).bos = 1
pkt = pkt / IP(dst=addr) / UDP(dport=4321, sport=1234)
pkt.show2()
sendp(pkt, iface=iface, verbose=False)
#pkt = pkt / SourceRoute(bos=0, port=2) / SourceRoute(bos=0, port=3);
#pkt = pkt / SourceRoute(bos=0, port=2) / SourceRoute(bos=0, port=2);
#pkt = pkt / SourceRoute(bos=1, port=1)
if __name__ == '__main__':
main()

View File

@@ -0,0 +1,181 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_IPV4 = 0x800;
const bit<16> TYPE_SRCROUTING = 0x1234;
#define MAX_HOPS 9
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header srcRoute_t {
bit<1> bos;
bit<15> port;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
srcRoute_t[MAX_HOPS] srcRoutes;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
transition select(hdr.ethernet.etherType) {
TYPE_SRCROUTING: parse_srcRouting;
default: accept;
}
}
state parse_srcRouting {
packet.extract(hdr.srcRoutes.next);
transition select(hdr.srcRoutes.last.bos) {
1: parse_ipv4;
default: parse_srcRouting;
}
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action srcRoute_nhop() {
standard_metadata.egress_spec = (bit<9>)hdr.srcRoutes[0].port;
hdr.srcRoutes.pop_front(1);
}
action srcRoute_finish() {
hdr.ethernet.etherType = TYPE_IPV4;
}
action update_ttl(){
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
apply {
if (hdr.srcRoutes[0].isValid()){
if (hdr.srcRoutes[0].bos == 1){
srcRoute_finish();
}
srcRoute_nhop();
if (hdr.ipv4.isValid()){
update_ttl();
}
}else{
drop();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.srcRoutes);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,191 @@
/* -*- P4_16 -*- */
#include <core.p4>
#include <v1model.p4>
const bit<16> TYPE_IPV4 = 0x800;
const bit<16> TYPE_SRCROUTING = 0x1234;
#define MAX_HOPS 9
/*************************************************************************
*********************** H E A D E R S ***********************************
*************************************************************************/
typedef bit<9> egressSpec_t;
typedef bit<48> macAddr_t;
typedef bit<32> ip4Addr_t;
header ethernet_t {
macAddr_t dstAddr;
macAddr_t srcAddr;
bit<16> etherType;
}
header srcRoute_t {
bit<1> bos;
bit<15> port;
}
header ipv4_t {
bit<4> version;
bit<4> ihl;
bit<8> diffserv;
bit<16> totalLen;
bit<16> identification;
bit<3> flags;
bit<13> fragOffset;
bit<8> ttl;
bit<8> protocol;
bit<16> hdrChecksum;
ip4Addr_t srcAddr;
ip4Addr_t dstAddr;
}
struct metadata {
/* empty */
}
struct headers {
ethernet_t ethernet;
srcRoute_t[MAX_HOPS] srcRoutes;
ipv4_t ipv4;
}
/*************************************************************************
*********************** P A R S E R ***********************************
*************************************************************************/
parser MyParser(packet_in packet,
out headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
state start {
transition parse_ethernet;
}
state parse_ethernet {
packet.extract(hdr.ethernet);
/*
* TODO: Modify the next line to select on hdr.ethernet.etherType
* If the value is TYPE_SRCROUTING transition to parse_srcRouting
* otherwise transition to accept.
*/
transition accept;
}
state parse_srcRouting {
/*
* TODO: extract the next entry of hdr.srcRoutes
* while hdr.srcRoutes.last.bos is 0 transition to this state
* otherwise parse ipv4
*/
transition accept;
}
state parse_ipv4 {
packet.extract(hdr.ipv4);
transition accept;
}
}
/*************************************************************************
************ C H E C K S U M V E R I F I C A T I O N *************
*************************************************************************/
control MyVerifyChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
************** I N G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyIngress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
action drop() {
mark_to_drop();
}
action srcRoute_nhop() {
/*
* TODO: set standard_metadata.egress_spec
* to the port in hdr.srcRoutes[0] and
* pop an entry from hdr.srcRoutes
*/
}
action srcRoute_finish() {
hdr.ethernet.etherType = TYPE_IPV4;
}
action update_ttl(){
hdr.ipv4.ttl = hdr.ipv4.ttl - 1;
}
apply {
if (hdr.srcRoutes[0].isValid()){
/*
* TODO: add logic to:
* - If final srcRoutes (top of stack has bos==1):
* - change etherType to IP
* - choose next hop and remove top of srcRoutes stack
*/
if (hdr.ipv4.isValid()){
update_ttl();
}
}else{
drop();
}
}
}
/*************************************************************************
**************** E G R E S S P R O C E S S I N G *******************
*************************************************************************/
control MyEgress(inout headers hdr,
inout metadata meta,
inout standard_metadata_t standard_metadata) {
apply { }
}
/*************************************************************************
************* C H E C K S U M C O M P U T A T I O N **************
*************************************************************************/
control MyComputeChecksum(inout headers hdr, inout metadata meta) {
apply { }
}
/*************************************************************************
*********************** D E P A R S E R *******************************
*************************************************************************/
control MyDeparser(packet_out packet, in headers hdr) {
apply {
packet.emit(hdr.ethernet);
packet.emit(hdr.srcRoutes);
packet.emit(hdr.ipv4);
}
}
/*************************************************************************
*********************** S W I T C H *******************************
*************************************************************************/
V1Switch(
MyParser(),
MyVerifyChecksum(),
MyIngress(),
MyEgress(),
MyComputeChecksum(),
MyDeparser()
) main;

View File

@@ -0,0 +1,16 @@
{
"hosts": [
"h1",
"h2",
"h3"
],
"switches": {
"s1": { "cli_input" : "s1-commands.txt" },
"s2": { "cli_input" : "s2-commands.txt" },
"s3": { "cli_input" : "s3-commands.txt" }
},
"links": [
["h1", "s1"], ["s1", "s2"], ["s1", "s3"],
["s3", "s2"], ["s2", "h2"], ["s3", "h3"]
]
}