Re: [SQU] Page faults with Solaris 8

Solaris pagefaults for all disk I/O IIRC, so high pagefaults numbers are
probably to be expected.

Recommendation: Read up a bit on Sun Solaris performance tuning. SUN has
even published a good book on the subject, covering most aspects you
need to know (and a bit more).

For Squid you need to have the box tuned for an application which both
uses lots of memory and disk I/O.

--
Henrik Nordstrom
J.D. Bronson wrote:
> 
> I know this has been talked about *but* I am wondering if anyone on the
> list has experience with 'Page faults' AND Solaris...
> 
> As you can see, I only ran squid for a short time and saw only 43 requests.
> But 28 page faults happened according to the log when squid was exited.
> 
> I have enough ram and the machine is quite quiet especially at this time.
> 
> Can anyone shed some light on this? I do indeed understand paging, but
> wanted to see if this is normal or if there is a tweak I can do <?>
> 
> Cache_mem_size = 8MB
> and I do use memory_pools on
> 
> Thanx in advance!
> 
> 2001/01/28 07:35:00| Preparing for shutdown after 43 requests
> 2001/01/28 07:35:00| Waiting 30 seconds for active connections to finish
> 2001/01/28 07:35:00| FD 14 Closing HTTP connection
> 2001/01/28 07:35:02| Shutting down...
> 2001/01/28 07:35:02| FD 15 Closing ICP connection
> 2001/01/28 07:35:02| Closing unlinkd pipe on FD 12
> 2001/01/28 07:35:02| storeDirWriteCleanLogs: Starting...
> 2001/01/28 07:35:02|   Finished.  Wrote 557 entries.
> 2001/01/28 07:35:02|   Took 0.0 seconds (48799.7 entries/sec).
> CPU Usage: 0.670 seconds = 0.400 user + 0.270 sys
> Maximum Resident Size: 0 KB
> Page faults with physical i/o: 28
> 
>   -JDB
> 
> --
> To unsubscribe, see http://www.squid-cache.org/mailing-lists.html
--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html