Hi
I suspect you have not updated your Frontier squid configuration in a little while. There have been a bunch of fixes for problems with log rotate. For more details:
http://frontier.cern.ch/dist/rpms/frontier-squidRELEASE_NOTES
Checking the monitoring:
http://frontier.cern.ch/squidstats/mrtgatlas/UKI-LT2-QMUL/index.html
In the last few days you have indeed had a bit of a spike of hits and from checking the panda monitor you are running lots of jobs testing new software so I don't think the load you are observing is unexpected.
Alastair
On 2 Oct 2012, at 15:58, D.Traynor wrote:
> Hi
>
> In the last few days our frontier-squid, which is also used for CVMFS,
> has started to fill up its access.log file at the rate of 1 gig an hour.
> I've hopefully now set up compressed logrotate but this will still fill
> up the disk a the rate of few gig per day.
>
> We've not changed our config. Does any one else see this issue?
> I guess it's coming from the rampup from atlas mentioned in Re: squid on
> sl6 thread (all logs I've looked at point to frontierATLAS). Is there
> anything else I can do to reduce the log load? I will need to put in
> place a more longtime strategy when I deploy my new pair of
> (frontier-)squids in the near future.
>
> thnaks
>
> dan
>
>
> relevant offending part of the squid.conf file is
>
> ...
> logformat awstats %>a %ui %un [%{%d/%b/%Y:%H:%M:%S +0000}tl] "%rm %ru
> HTTP/%rv" %Hs %<st %Ss:%Sh %tr "%{X-Frontier-Id}>h"
> ...
> access_log /var/log/squid/access.log awstats
> ...
>
>
>
> --
> * Dr Daniel Traynor, Grid system administrator
> * Physics, QMUL, London, Tel +44(0)2078825059
|