sipb / spike Goto Github PK

Test that spike behaves correctly on pcap files

Backends should be able to report three possible states:

• Down
  • No traffic
• Up
  • Send traffic
• Drained
  • Don't send additional traffic if possible, but existing connections are OK

We should enforce that the maglev table size is prime in the constructor.
https://github.com/sipb/spike/blob/master/maglev/maglev.go#L40-L44

We need to bound the amount of packet fragments that we store. Relatedly, we should have a way of disabling support for packet fragmentation altogether.

Spike should expose a method of reconfiguring itself live, remotely.

I propose that we get rid of the function Add:
https://github.com/sipb/spike/blob/master/maglev/maglev.go#L109-L111

I think it complicates the api. I feel things would be simpler if only SetWeight were used.

Right now all packets get forwarded to a single backend pool

Test that spike behaves correctly on real hardware.

Snabb is single-threaded by default, but is capable of multithreading. We should run multiple packet handling threads if possible.

Add connection tracking instead of looking up the connection with consistent hashing every time.

A guide containing the required dependencies as well as dev setup instructions would be nice.

We can't tell the difference between packets being dropped and taking a long time to process. We should find a way to eliminate the timing dependence from tests.

Spike needs to report its own health somehow.

So far, we have made progress on consistent hashing, connection tracking, and server health checks.
consistent hashing: https://github.com/jma100/spike (jingweim)
server health checks: https://github.com/CatherineZeng/spike (yczeng)
connection tracking: https://github.com/ed1d1a8d/spike/blob/lru-testing/lru-test/lru-test.go (twang6)
We are trying to put together the modules atm.

issues: writing tests for consistent hashing module: (dkogut, gh1)

Currently spike always sends the encapsulated packets as IPv4. We should be able to emit both IPv4 and IPv6 packets.

The health checker doesn't account for the health endpoint itself being down. It should treat the service as down if the endpoint is unreachable.

We need to be able to handle fragmented IP packets.

We should fuzz spike to increase our confidence that it can handle malformed packets.

We need a way of passing configuration options into spike.

Currently spike reads and writes pcap files. We should get it working with an actual NIC.

"consistent" should not be there.

We need to advertise routes to the router somehow.