On 2004-06-07 at 21:29:18 Brian Feldman wrote: >> Here are two excerpts from kdump output that basically repeat all the >> time: > I see a lot of accept(2) there... I think it's a good possibility > Robert accidentally broke accept[1]()'s error cleanup. Hmm, I'm starting to think that this is why I ran into a similar problem with Squid yesterday; it ran out of file descriptors, started spamming the console, and made the machine quite inaccessible. :) (Luckily there's a serial console.) I've run Squid on FreeBSD for years now, without ever encountering something like this! I then unthinkingly bumped kern.maxfiles to 8192, but it's probably better if I try Robert's patch too.
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:56 UTC