Re: [squid-users] Squid 2.6 disk cache growth

From: Amos Jeffries <squid3@dont-contact.us>
Date: Tue, 15 Jan 2008 12:07:50 +1300 (NZDT)

>
> I'm currently running Squid 2.6 as a reverse proxy server to ease the
> load on my application web servers.
>
> The problem that I'm running in to is that Squid keeps using up all
> available hard-disk and then stops servicing requests.
>
> My config looks like this:
>
> ---------------
> http_port 80 accel defaultsite=app2.tveyes.com
>
> cache_peer app1.tveyes.com parent 8080 0 no-query originserver
> round-robin weight=1
> cache_peer app2.tveyes.com parent 8080 0 no-query originserver
> round-robin weight=1
>
> visible_hostname app1.tveyes.com
>
> cache_dir ufs /var/spool/squid 2048 16 256
>
> cache_mem 512 MB
>
> [snip]
>
> cache_log none
> cache_store_log none
>
> emulate_httpd_log on
> ----------------
>
> I'm not sure why squid should be taking any more than 2GB of disk space,
> but every time I look at the cache when it has crashed, the swap.state
> file is 25GB.
>
> I apologize if this topic has been covered before, I've looked around
> and not really found a satisfactory answer to why Squid is behaving this
> way.

You need to rotate your squid logs frequently, possibly daily.
swap.state is a append-only journal and needs squid -k rotate to clean it up.

Amos
Received on Mon Jan 14 2008 - 16:07:53 MST

This archive was generated by hypermail pre-2.1.9 : Fri Feb 01 2008 - 12:00:04 MST