[Zope] Re: Squid headaches

Dieter Maurer dieter@handshake.de
Sat, 31 May 2003 19:42:31 +0200


ATT: cross post to ZODB-DEV.

Wankyu Choi wrote at 2003-5-31 01:20 +0900:
 > > Can you post a before and after of "Peer Cache Statistics"  page in squid 
 > > cachemanager CGI
 > 
 > About 10 minutes ago, I caught Squid getting restarted. Every night at 1 AM,
 > the same thing happens: the packing script starts running; Squid declares
 > peers dead; my cron job restarts Squid; the packing script keeps running;
 > everything's back to normal, just a couple of minutes of a service blackout
 > (the squid watch script runs every 10 minutes).

Zope becomes very slow during packing.

I do not yet fully understand why, but the complete host seems to
be affected by the high disc bandwidth required during packing.

In the second packing phase (the so called copying phase),
the packing thread helds the storage lock most of the time.
This can make accessing the ZODB impossible by normal operations,
especially because the "lock_release", "lock_acquire" probably
does not succeed in giving the lock to a waiting thread.
I have a patch that restricts this locking to only a minor
fraction of time.
However, I know that this is not responsible alone for the
sluggish responses during packing.


Dieter