In cachemgr, you can see the "queue" in "ASYNC IO Functions Counters". What
number do you see during peak load, or when the error shows in Cache.Log?
Are you using 280GB cache_dir on one HDD? Or multiple HDDs? What's your
HDD interface type (IDE, SATA, SCSI, etc.)?
Regards
HASSAN
----- Original Message -----
From: "Shekhar Gupta" <shekharsaha...@gmail.com>
To: "Kinkie" <gkin...@gmail.com>
Cc: <squid-users@squid-cache.org>
Sent: Thursday, March 05, 2009 19:18
Subject: Re: [squid-users] squidaio_queue_request: WARNING - Queue
congestion
For Last 1 hr .
____________
cpu_time = 86.267886 seconds
wall_time = 3600.031220 seconds
cpu_usage = 2.396309%
On Thu, Mar 5, 2009 at 2:07 PM, Kinkie <gkin...@gmail.com> wrote:
On Wed, Mar 4, 2009 at 6:59 AM, Shekhar Gupta <shekharsaha...@gmail.com>
wrote:
it's showing
client_http.requests = 102.190863/sec
It's strange. The load doesn't seem to be that high.
What's the CPU load for squid, and how is iowait on your system? (both
can be found using the 'top' command)
--
/kinkie
