RE: [squid-users] Problem with squid... Dirty Cache constant restart loop....[SOLVED]

From: Khawar Nehal <khawar@dont-contact.us>
Date: Tue, 26 Jul 2005 12:31:07 +0500

Could take a guess.

Maybe some viruses on a windows machine is requesting something which
squid keeps giving to to in less than 1 millisecond.

Then squid mentiones the mem hit and release from mem in the store log.

Can the guru's of squid analyze situation and see if the thinking is in
the right direction ?

I On Mon, 2005-07-25 at 16:15 +0100, Ross McInnes wrote:
> Hey all again...
>
> Ive sorted this now. What cleared it was when I ran a default config
> pointing back to the default locations.
>
> I then changed the cache dir back to where it should be and it was still
> ok...
>
> Changed everything else one by one back to where it should be and it failed
> on the logs directory...
>
> My store.log file was 2GB ++... Funny thing was it got rotated at 4am this
> morning (does it every week)
>
> So had a look through it and there are 1000's of lines such as....
>
> 1122286149.325 RELEASE -1 FFFFFFFF 2DC3245064EB354E24E2A91F3AB185B4 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF 216DB1B58751EE8D10965063EF2581D4 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF AA135BE9264086B83BC91DA34755DF42 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF CC6D2D7C3C4C026DFA8B79C5EA1CF548 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF DE84219FE2466C35C6F7A97D2884AA02 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF D7A9D8327A71DCD1E8BB681E33DD1005 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF 05C4E760CA6F74198E6AC587AB57A93F ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF A9FAF6B5D35654D62AC88009E09F01E0 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF 292329206E1C38E653A2AE53789C44F2 ?
> ? ? ? ?/? ?/? ? ?
> 1122286149.325 RELEASE -1 FFFFFFFF 7E901F327EEF9E1C5B640324807BE474 ?
> ? ? ? ?/? ?/? ? ?
>
> Any ideas what this is??
>
> But anyways the web is working again :)
>
> Cheers
>
> Ross
>
> -----Original Message-----
> From: Ross McInnes [mailto:sysrm@stvincent.ac.uk]
> Sent: 25 July 2005 14:56
> To: squid-users@squid-cache.org
> Subject: RE: [squid-users] Problem with squid... Dirty Cache constant
> restart loop....
>
>
> Heya gert,
>
> df -h shows
>
> Filesystem Size Used Avail Use% Mounted on
> /dev/sda5 22G 6.3G 15G 31% /
> /dev/sda1 99M 36M 58M 39% /boot
> none 501M 0 501M 0% /dev/shm
> /dev/sda2 9.7G 2.9G 6.3G 32% /var/log/squid
> /dev/sdb1 68G 65M 64G 1% /var/squid/cache <-- cache dir
>
> I can ping/traceroute/nslookup/host everywhere
>
> Ive done that, ive even set it back to defaults on /usr/local/squid/cache
> (still says its dirty)
>
> Done all of that im afraid...
>
> Ross
>
> -----Original Message-----
> From: Gert Brits [mailto:gbrits@techconcepts.co.za]
> Sent: 25 July 2005 14:43
> To: squid-users@squid-cache.org
> Subject: RE: [squid-users] Problem with squid... Dirty Cache constant
> restart loop....
>
> Hi There
>
> 1.) Check your disk space ( your cache partition if it is on its own ).
> 2.) Check if you have a proper DNS server that can resolve DNS names under
> /etc/resolve.conf
> 3.) Remove your old cache under your cache directory.
> 4.) squid -z
> 5.) restart squid.
>
>
>
>
>
> -----Original Message-----
> From: Ross McInnes [mailto:sysrm@stvincent.ac.uk]
> Sent: 25 July 2005 03:27 PM
> To: squid-users@squid-cache.org
> Subject: [squid-users] Problem with squid... Dirty Cache constant restart
> loop....
>
>
> Hi all.
>
> Been running squid for years, never really had any problems till today.
>
> Everything seems to have been running fine till just before 12pm..
>
> Then someone comes in and asks if the net is down.... Sure enough squid has
> died and is no longer running on the server.
>
> I do a service squid start
>
> Try to connect to the internet again... Nothing. No username/password
> request etc.
>
> Tail -f /var/log/squid/cache.log I get...
>
> 2005/07/25 15:10:14| Starting Squid Cache version 2.4.STABLE6 for
> i686-pc-linux-gnu... 2005/07/25 15:10:14| Process ID 5379 2005/07/25
> 15:10:14| With 1024 file descriptors available 2005/07/25 15:10:14|
> Performing DNS Tests... 2005/07/25 15:10:14| Successful DNS name lookup
> tests... 2005/07/25 15:10:14| DNS Socket created on FD 4 2005/07/25
> 15:10:14| Adding nameserver 127.0.0.1 from /etc/resolv.conf 2005/07/25
> 15:10:14| helperOpenServers: Starting 20 'squirm' processes 2005/07/25
> 15:10:14| helperOpenServers: Starting 15 'ncsa_auth' processes
> 2005/07/25 15:10:17| Unlinkd pipe opened on FD 44 2005/07/25 15:10:17| Swap
> maxSize 61440000 KB, estimated 4726153 objects 2005/07/25 15:10:17| Target
> number of buckets: 236307 2005/07/25 15:10:17| Using 262144 Store buckets
> 2005/07/25 15:10:17| Max Mem size: 8192 KB 2005/07/25 15:10:17| Max Swap
> size: 61440000 KB 2005/07/25 15:10:17| Rebuilding storage in
> /var/squid/cache (DIRTY) 2005/07/25 15:10:17| Using Least Load store dir
> selection 2005/07/25 15:10:17| Set Current Directory to /var/squid/cache
>
> Repeated over and over and over etc.
>
> Ok, I figure that the cached files are not happy,
>
> So I remove the cache files, then set them back up using -z still the same
> problem.
>
> Rebooted the box, removed squirm and ncsa_auth, still the same... Im looking
> to remove the cache next to see if that helps it.
>
> Looked on google etc for this problem and no one seems to have an answer....
>
> Any ideas?
>
> Many Thanks
>
> Ross
Received on Tue Jul 26 2005 - 02:07:55 MDT

This archive was generated by hypermail pre-2.1.9 : Mon Aug 01 2005 - 12:00:03 MDT