On Jul 6, 2009, at 6:11 PM, Danek Duvall wrote:
We're still trying to figure out what's going on. We fixed one
configuration error earlier, but are continuing to get timeout
errors. We
thought we had it narrowed down to a flood attack from a single
host, but
that load has now disappeared, but the problem hasn't. Still probing.
I adjusted the tcp stack which has helped unstuff the queue a bit. I
sent an email to Stephen and Dan about that and some other errors that
I noticed. There is an overall high volume, but likely not enough to
grind things to slow crawl.
Why don't you send email over to tonic-ops with what you changed as we
can possibly be of help, but I don't like to poke around too much with
stuff that isn't ours to poke.
e.
_______________________________________________
website-discuss mailing list
[email protected]
