Re: Safe to use Sitemesh with over 10,000 daily views?

Joost Baaij wrote:
> Thank you very much for the quick reply. I had already looked into several profiling tools, but these must be purchased which unfortunately is not an option. Is there any one in particular you can recommend?
> 
> Java 1.5 supposedly has profiling tools built-in, should I now be using those instead of a tool such as JProbe or OptimizeIt?

You can get free trials with many profilers.  I'm using YourKit these 
days (www.yourkit.com).

10,000 page views is so small traffic, it should be pretty trivial to 
test this.  Get a tool like siege or openload or anything that will 
pound a URL (http://opensourcetesting.org/performance.php)  Pick a 
simple JSP (assuming it can handle at least 100 req/s) then in 100 
seconds you'll reached your daily rate.  If your test runs for 5+ 
minutes against that page, then neither that page nor sitemesh (in that 
situation) are the leak.  It's sounds like you've got a hard to track 
down leak in production.

As a datapoint, we've used sitemesh for years without any leaks.

-- 
Serge Knystautas
Lokitech >> software . strategy . design >> http://www.lokitech.com
p. 301.656.5501
e. [email protected]

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]