Re: [squid-users] any updates/info on using spread with squid for logging ?

On Fri, 12 Aug 2005, john allspaw wrote:

> this is all true, but am I wrong in thinking that logging to disk incurs
> yet more disk I/O, and for both refers *and* access, then that's 2 disk
> writes per object request ?

Unlike syslog the logs are not flushed immediately to disk, instead your
OS carefully pages the written data out by it's fs buffer flushing
algorithms as done on any normal file write, usually using an elevator
queue. As a result each logfile write to disk contains very many requests
and the higher the load the more requests per logfile disk write.

If the logs are on a separate drive then the limit is quite near the I/O
platter bandwidth of your drive, provided you do not run short of memory.
If there is other activity stealing I/O time then performance will be
lower.

Regards
Henrik
Received on Mon Aug 15 2005 - 12:41:55 MDT