I bet.
> Should it be possible to increase SOMAXCONN, or does that interact
> with/break the socket hash stuff? It looks like there are a few hardcoded
> 128 values in the pre-2.0.30 patch, but I'm not sure what they do.
Because the backlog queue is a signed char. If you change that (and test you
got all the cases), your machine should be happy with a 512 SOMAXCONN