[squid-users] cache_dir weirdness

[2.4 STABLE2 on BSD/OS 4.2]

I cannot figure this out. After I built this machine, all four
cache partitions came up to 70% and stayed there. A few weeks
ago, I took the machine down to replace the NIC. When I brought
it back up, the first partition dropped to 58%. Yesterday, I
stopped and restarted the Squid process so that I could look at
the logs. It seems normal to me...

Rebuilding storage in /usr/local/squid/cache_1 (CLEAN)
Rebuilding storage in /usr/local/squid/cache_2 (CLEAN)
Rebuilding storage in /usr/local/squid/cache_3 (CLEAN)
Rebuilding storage in /usr/local/squid/cache_4 (CLEAN)

Done reading /usr/local/squid/cache_1 swaplog (294475
Done reading /usr/local/squid/cache_4 swaplog (314629
Done reading /usr/local/squid/cache_2 swaplog (323156
Done reading /usr/local/squid/cache_3 swaplog (357838

Almost two days later, cache_1 is *still* at 58%. Here's my
`df', altered slightly to avoid word-wrap:

Filesystem 1K-blocks Used Avail Capacity Mounted on
/dev/sd0a 49551 15617 31456 33% /
/dev/sd0d 991727 413517 528623 44% /usr
/dev/sd0e 991727 71787 870353 8% /usr/local
/dev/sd0f 1983982 1169899 714883 62% /var
/dev/sd0g 991727 87 942053 0% /home
/dev/sd1d 8426876 4617079 3388453 58% [cache_1]
/dev/sd1e 8426876 5607600 2397932 70% [cache_2]
/dev/sd2d 8426876 5613359 2392173 70% [cache_3]
/dev/sd2e 8426876 5609161 2396371 70% [cache_4]

From `squid.conf':

cache_dir ufs /usr/local/squid/cache_1 6000 64 256
cache_dir ufs /usr/local/squid/cache_2 6000 64 256
cache_dir ufs /usr/local/squid/cache_3 6000 64 256
cache_dir ufs /usr/local/squid/cache_4 6000 64 256

Any ideas?

Thanks!
Scott

--
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
*                                                       *
*  Easley Internet Solutions              864.859.2400  *
*  Easley, SC  USA                   Fax: 864.855.5864  *
*  http://www.easley.net/            AIM:     ExCavSGT  *
*                                                       *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *