[squid-users] cache filled up ... why?

Squid shut down the other day because the cache was full and I have
no idea why it was full.

I have 50Gb of disk SOLELY for squid cache (on its own partition)
Nothing else shares this partition. The squid cache is by itself.
It is split across three disks (raid0) and appears as one device
I cranked the inodes way up (lots of small files)

Yet, after running for 30 days (and moving on average 16Gb of
data each day), the box suddenly ran out of space. This happened
to me before, so I set /etc/squid.conf to well below the 50gb limit
of the disk (set to 42Gb) and changed the inodes at the same time.

Some particulars follow, I am completely confused as to why it filled
up. How did it handle being up for 30 days (30days * 16Gb = 480Gb)
and suddenly fill up?

I need some hints as I'm completely out of ideas. What needs tuning?

- John

details:
------------
squid.conf
cache_dir ufs /var/spool/squid 42000 128 256

cache.log
2001/09/15 16:29:24| diskHandleWrite: FD 19: disk write error: (28) No space left on
device
FATAL: Write failure -- check your disk space and cache.log
Squid Cache (Version 2.3.STABLE4): Terminated abnormally.

I did kill the swap state file which freed up the 245Mb space you see
below.

[root@munin squid]# df
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/sda1 521748 181344 313900 37% /
/dev/sda2 3028108 1489168 1385120 52% /usr
/dev/sde1 3028080 32300 2841960 1% /var
/dev/sde2 14626572 2807120 11076448 20% /var/log/squid
/dev/md0 50135549 49354969 245619 100% /var/spool/squid
[root@munin squid]# df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/sda1 66400 9586 56814 14% /
/dev/sda2 384768 84603 300165 22% /usr
/dev/sde1 384768 383 384385 0% /var
/dev/sde2 1860480 114 1860366 0% /var/log/squid
/dev/md0 26750976 3775247 22975729 14% /var/spool/squid
Received on Mon Sep 17 2001 - 08:36:56 MDT