RE: System Analysis Suggestions

Hi Jim,

Thanks for the input. Our 400 runs through our Foundry Networks Layer-3 gigabit switch on which we have our fibre network take the bulk of the transactions. The web server is also on the same switch, so I think I can rule out a poor / low speed connection.

/b;

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Jim Franz
Sent: Monday, August 10, 2009 9:33 PM
To: Midrange Systems Technical Discussion
Subject: Re: System Analysis Suggestions

Since this seems focused on web traffic, check your ethernet
config is as open as the rest of your network. The slowest part of
your ethernet > switch>router, etc will limit your throughput - I've seen GB
Ethernet limited by a 10M switch, or a older ethernet still set to 10M half
duplex when the card supported more.
Jim Franz

----- Original Message -----
From: "Nathan Andelin" <nandelin@xxxxxxxxx>
To: "Midrange Systems Technical Discussion" <midrange-l@xxxxxxxxxxxx>
Sent: Monday, August 10, 2009 9:23 PM
Subject: Re: System Analysis Suggestions

From: Trevor Perry
The CPU% being high is not a measure of good or bad performance.

I agree. You can still get good response, even though CPU utilization may
be high. I performed a stress test against some of our Web applications
recently, using HP Loadrunner to measure and test performance. It was
enlightening to see that most requests completed in under 15 milliseconds,
even though our single-core server was pounded with about 400 requests per
second and CPU was at 80%. However, our Web applications run under IBM
i - not Windows.

the numbers of the prestart QZDASOINIT jobs may be configured very
low

Mike Cunningham made that point too.

and when a new job is needed, the system has to spend some
resources creating that new job.

But if the number of prestart DB servers is too low, that may only explain
a temporary performance lag, until more jobs are started, which is
automatic.

Brian indicated that CPU utilization gradually crept up over a couple year
period as new applications were added. That sounds normal for servers -
reflecting additional workload. A continuous high-rate of page faults
could indicate a memory problem, but he didn't say that. Maybe this is a
Windows problem. When workloads are distributed across multiple server
tiers, it's harder to get a handle on the problem.

-Nathan.




--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.