Mailing-List: contact users-help@cocoon.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@cocoon.apache.org
Received-SPF: neutral (hermes.apache.org: local policy)
Mime-Version: 1.0 (Apple Message framework v622)
In-Reply-To: <efa46b28618759ee7d60f07ef52062dc@apache.org>
References: <Pine.LNX.4.44.0504161254250.12474-100000@zeus.osmosis.gr>
 <efa46b28618759ee7d60f07ef52062dc@apache.org>
Content-Type: text/plain; charset=US-ASCII; format=flowed
Message-Id: <2907816b1004d4a5435e0184174c3b58@thenot.org>
Content-Transfer-Encoding: 7bit
From: =?ISO-8859-1?Q?Andr=E9_Th=E9not?= <andre-lists@thenot.org>
Subject: Re: Scalability, Performance, Stability
Date: Mon, 18 Apr 2005 10:43:30 -0400
To: users@cocoon.apache.org

On Apr 16, 2005, at 6:15, Bertrand Delacretaz wrote:

> If your system crashes in normal use, there must be a way to crash it 
> in the lab as well, where it's much easier to debug and do post-mortem 
> analyses. Often, just running a few tens of wget loops in a shell will 
> show problems already.

And if I may add to this, when doing load testing, have the requests 
come from a different machine, and make sure you know that the tool you 
use actually works for the scenario you are testing. I saw a client who 
was wondering why their Cocoon site wouldn't scale when in fact it was 
the tool they were using (jMeter) which was having trouble parsing the 
results fast enough in certain cases...

A.


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@cocoon.apache.org
For additional commands, e-mail: users-help@cocoon.apache.org