[squid-users] How to reduce cache_swap after cache disk is full

Hi, all
     I encountered the following question:
/etc/rc.d/init.d/squid -N -d1
...
...

2005/04/02 16:27:21| Ready to serve requests.
2005/04/02 16:27:22| Store rebuilding is 1.1% complete
2005/04/02 16:27:24| diskHandleWrite: FD 39: disk write error: (28) No space lef
t on device
FATAL: Write failure -- check your disk space and cache.log Aborted

df
/dev/hda1 147766 9321 130816 7% /boot
/dev/hda6 10080488 9515868 52552 100% /cache1
/dev/hda5 10080488 9568420 0 100% /cache2
none 514688 0 514688 0% /dev/shm
/dev/hda7 12104380 3549532 7939964 31% /var

squid2.5 stable7
   #./configure --prefix=/usr/local/squid --sysconfdir=/etc/squid
   --enable-auth=ntlm disable-internal-dns --enable-storeio=aufs,ufs
   --with-aufs-threads=32 --enable-cache-digests --enable-underscores
   --enable-removal-policies

   squid.conf is:
cache_dir aufs /cache1 10000 16 123
cache_dir aufs /cache2 10000 16 123
cache_access_log /var/log/squid/access.log
cache_log /var/log/squid/cache.log
cache_store_log none
log_ip_on_direct off
cache_replacement_policy heap LFUDA
cache_swap_low 93
cache_swap_high 95

question: Why the cache swap don't reduce to 93% after reached 95%? What can I
do under those case?

Thanks
Seewo chen
Received on Sat Apr 02 2005 - 01:41:48 MST