>
> Hello
>
> the emptying of cache dirs we thought that happens after swap.state corruption
> may have another reason.
>
> This server has 2 cache_dirs with 7000MB each
> suddenly the disk usage goes up and squid logs it in messages
> then squid with the store_diskd.c:1930 error
>
> after restart it cleans the disks, crashes still sometimes with the same error
>
> within all this time squid still acuse disk over limit but definitly it is
> not!
>
> shutting down squid with -k shutdown and restarting it manually interrupts the
> error and squid comes back to normal work
>
> so I guess we got something usefull here?
>
>
> Dec 18 16:32:01 wco-alm squid[80627]: WARNING: Disk space over limit: 17677474
> KB > 14336000 KB
> Dec 18 16:32:12 wco-alm squid[80627]: WARNING: Disk space over limit: 17666396
> KB > 14336000 KB
> Dec 18 16:32:20 wco-alm kernel: pid 80627 (squid), uid 1004: exited on signal
> 6
> Dec 18 16:32:20 wco-alm squid[80627]: assertion failed:
> diskd/store_dir_diskd.c:1930: "buf"
> Dec 18 16:32:30 wco-alm squid[97328]: WARNING: Closing open FD 2
>
SQUID version ?
M.
Received on Mon Dec 19 2005 - 06:39:43 MST
This archive was generated by hypermail pre-2.1.9 : Sat Dec 31 2005 - 12:00:02 MST