Re: [squid-users] swap.state, swap.state.new, and cache shrinking

On Thu, 30 Oct 2003, Noel Clarkson wrote:

> read and guessed, I think my problem was that squid was trying to rebuild
> the swap.state file by creating a swap.state.new file and this was running
> the partiton out of space, and then doing the rebuild again after failing
> (clearing the space, and rebuilding and running out of space over and
> over). So I've now cleared the cache (or at least the swap.state) and
> things have settled down, but I'm wondering if this is the likely
> explaination to my trouble.

If the partition where swap.state is kept is full then Squid will
repeately crash while trying to rebuild the cache index.

> The other question regarding this that I have is that I've now reduced the
> size that the cache can take up so that this won't become a problem again,
> but the actual space that the cache takes up is more than I have told it
> should be the max because the files from the cache before I told it to
> shrink are still around.

As long as you do not change the L1 / L2 parameters Squid will
automatically clean up the cache if you give it some time. Normally a
cache cleanup maintenance sweep takes up to 24 hours.

Alternatively you can do a complete clean of the cache.

1. Shut down Squid
2. Remove the cache directories
3. run "squid -z".
4. Start Squid again

While you are looking at these things, make sure you have log rotation
configured. If not the Squid log files including swap.state will grow
forever until it runs out of space.

Regards
Henrik
Received on Wed Oct 29 2003 - 18:38:34 MST