Re: some linux tuning

On Tue, 29 Jun 1999, Stephen R. van den Berg wrote:

> Jeff Barrow wrote:
> >Could you look at something for me:
> >In client_side.c around line 1900...
>
> >There seem to be 3 cases where inbuf is allocated but not freed (error
> >cases)... is this correct? With transparent proxies, this could be what's
>
> Even more so, apparently.
> It looked serious enough to patch myself, in case you didn't yet, try this:
>
> diff -U 2 -b -B -p -r -d --horizon-lines=2 -X /usr/local/etc/xdiff squid-2.2.STABLE3/src/client_side.c squid-B2/src/client_side.c
> --- squid-2.2.STABLE3/src/client_side.c Mon May 10 18:00:40 1999
> +++ squid-B2/src/client_side.c Tue Jun 29 11:39:33 1999

I applied this patch last night and it seems to have helped noticably.
In the past 8 hours, I've only leaked 7mb on the big server that generally
has to have squid killed every 24-48 hours.

OTOH, I'm not sure if this is related to the patch, but I just saw squid
(2.2S3 + the 2 leak patches from nlanr, + the patch above) crash for the
first time.

1999/06/30 09:18:14| clientReadRequest: FD 129 Invalid Request
FATAL: Received Segment Violation...dying.

I don't know if the segv was a result of that error, but it happened the
same minute. I've gotten the same message numerous times without crashes
before and after last night's patch.

----don't waste your cpu, crack rc5...www.distributed.net team enzo---
 Jon Lewis *jlewis@lewis.org*| Spammers will be winnuked or
 System Administrator | nestea'd...whatever it takes
 Atlantic Net | to get the job done.
_________http://www.lewis.org/~jlewis/pgp for PGP public key__________
Received on Wed Jun 30 1999 - 07:19:21 MDT