Re: [SQU] Ratio of phyiscal I/O to http requests is high and rising

From: Duane Wessels <wessels@dont-contact.us>
Date: Fri, 12 Jan 2001 16:19:43 -0700 (MST)

On Fri, 12 Jan 2001, Drash, Jim [EESUS] wrote:

> I have followed all the suggestions to lower memory use in the FAQ.
> Any ideas?

I don't think you need to worry about it. Your response time looks
quite good. You probably want to read up on page faults in the
FAQ. In particular, note:

  The ``page faults with physical IO'' comes from the OS via
  getrusage(). It's highly OS dependent on what it means.

So, page faults doesn't necessarily imply swapping and running out
of memory.

> Median Service Times (seconds) 5 min 60 min:
> HTTP Requests (All): 0.16775 0.20843
> Cache Misses: 0.68577 0.64968
> Cache Hits: 0.05046 0.02451
> Near Hits: 0.37825 0.33943
> Not-Modified Replies: 0.01035 0.01035
> DNS Lookups: 0.00278 0.00278
> ICP Queries: 0.00000 0.00000
> Resource usage for squid:
> UP Time: 36916.425 seconds
> CPU Time: 11309.910 seconds
> CPU Usage: 30.64%
> CPU Usage, 5 minute avg: 31.57%
> CPU Usage, 60 minute avg: 42.22%
> Maximum Resident Size: 0 KB
> Page faults with physical i/o: 120730
> Memory usage for squid via mallinfo():
> Total space in arena: 48592 KB
> Ordinary blocks: 43525 KB 14857 blks
> Small blocks: 0 KB 0 blks
> Holding blocks: 1644 KB 3 blks
> Free Small blocks: 0 KB
> Free Ordinary blocks: 5066 KB
> Total in use: 45169 KB 93%
> Total free: 5066 KB 10%
> Memory accounted for:
> Total accounted: 26005 KB

--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html
Received on Fri Jan 12 2001 - 16:22:33 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:57:27 MST