On Tue, 21 Apr 2009 12:54:53 +1200 (NZST), "Amos Jeffries" <squid3@xxxxxxxxxxxxx> wrote: [cut] >> 2009/04/20 18:05:20| Store rebuilding is 822.48% complete > what type of file system is in use? ext3 > with what settings? 4x 300GB SAS disks cache_swap_low 80 cache_swap_high 85 cache_dir aufs /var/cache/proxy/cache1 256000 256 256 cache_dir aufs /var/cache/proxy/cache2 256000 256 256 cache_dir aufs /var/cache/proxy/cache3 256000 256 256 cache_dir aufs /var/cache/proxy/cache4 256000 256 256 maximum_object_size 256 MB minimum_object_size 1 KB cache_replacement_policy heap LFUDA Now (that's why does not work anymore): /dev/sdb1 276G 262G 0 100% /var/cache/proxy/cache1 /dev/sdc1 276G 262G 0 100% /var/cache/proxy/cache2 /dev/sdd1 276G 262G 0 100% /var/cache/proxy/cache3 /dev/sde1 276G 262G 0 100% /var/cache/proxy/cache4 Before crash was 20 or 35% free disk space on each disk, and was like that for 6 weeks (then reached the limits and was not growing anymore until the upgrade to STABLE14 crashed the box). > with what disk available? > on what operating system? BlueWhite64 (an unofficial 64 bits Slackware port). > is it rebuilding after saying DIRTY or CLEAN cache? CLEAN after upgrade, DIRTY after crashes. > does deleting the swap.state file(s) when squid is stopped fix things? I will try. The stranger think was store rebuild reporting > 100%. Thanks, Herbert.
