Paul Cocker wrote:
I assume the in-memory index is in addition to the memory_cache? So if
you have a 100GB disk cache you would need 1GB RAM... but that would
only cover the index and so you would need more memory for squid itself
and the memory_cache?
I believe so. Plus memory for all the ACLs and active connections etc.
Squid is unfortunately very RAM hungry and nobody has sponsored much in
the way of memory optimisation in a long while.
Amos
Paul Cocker
-----Original Message-----
From: Amos Jeffries [mailto:squid3@xxxxxxxxxxxxx]
Sent: 05 November 2007 23:44
To: Paul Cocker
Cc: squid-users@xxxxxxxxxxxxxxx
Subject: Re: Optimal maximum cache size
Is there such a thing as too much disk cache? Presumably squid has to
have some way of checking this cache, and at some point it takes
longer to look for a cached page than to serve it direct. At what
point do you hit that sort of problem, or is it so large no human mind
should worry?
:)
Paul
IT Systems Admin
Disk cache is limited by access time and ironically RAM.
Squid holds an in-memory index of 10MB-ram per GB-disk. With large disk
caches this can fill RAM pretty fast, particularly if the cache is full
of small objects. Large objects use less index space more disk.
Some with smaller systems hit the limit at 20-100GB, others in cache
farms reach TB.
As for the speed of lookup vs DIRECT. If anyone has stats, please let us
know.
Amos
TNT Post is the trading name for TNT Post UK Ltd (company number: 04417047), TNT Post (Doordrop Media) Ltd (00613278), TNT Post Scotland Ltd (05695897),TNT Post North Ltd (05701709) and TNT Post South West Ltd (05983401). Emma's Diary and Lifecycle are trading names for Lifecycle Marketing (Mother and Baby) Ltd (02556692). All companies are registered in England and Wales; registered address: 1 Globeside Business Park, Fieldhouse Lane, Marlow, Buckinghamshire, SL7 1HY.
