Performance and Load Testing

[Chris]

We have an asp application some of which will be being ported over to
asp.net so this may be a little OT. It is having some bad performance
problems and we need to isolate which areas are causing the problems. The
main performance problems are database related i.e. timeouts and deadlocks

Some questions:
Microsoft's ACT. How is it better than the Web Stress Application Tool?
Is ACT purely .net or can be used to look at any web app.
What is a good way of assessing which pages are causing the problems.
How can I relate the performance of the database to the application. The
system has LOTS (this app has 1000's of pages) of sql passed as strings will
this affect how easy the profiler is to read.
I suspect there are lots of memory leaks, are there any tools to look at
what is being left in the memory.

I guess what I want is a tool which will identify which pages are hogging
the most resources and if they are not being released so we can look at
optimising the code.

[=?Utf-8?B?UGV0ZXIgQnJvbWJlcmcgW0MjIE1WUF0=?=]

Act makes pretty realtime charts, but when it comes to accurately throwing a
multi-client stress test at an app, it falls on its face. Homer (Web Stress
Tool) is much better. There are others put out by commercial vendors that are
even better.
Peter

--
Site: http://www.eggheadcafe.com
UnBlog: http://petesbloggerama.blogspot.com
Short urls & more: http://ittyurl.net


"Chris" wrote:

We have an asp application some of which will be being ported over to
asp.net so this may be a little OT. It is having some bad performance
problems and we need to isolate which areas are causing the problems. The
main performance problems are database related i.e. timeouts and deadlocks

Some questions:
Microsoft's ACT. How is it better than the Web Stress Application Tool?
Is ACT purely .net or can be used to look at any web app.
What is a good way of assessing which pages are causing the problems.
How can I relate the performance of the database to the application. The
system has LOTS (this app has 1000's of pages) of sql passed as strings will
this affect how easy the profiler is to read.
I suspect there are lots of memory leaks, are there any tools to look at
what is being left in the memory.

I guess what I want is a tool which will identify which pages are hogging
the most resources and if they are not being released so we can look at
optimising the code.

[Chris]

Is there anything out there to look for memory leaks, it's more of a classic
asp question. Is it a stupid question? I.e. if it were that easy to test for
memory leaks everyone would do it? :)

"Peter Bromberg [C# MVP]" <pb*******@yahoo.yabbadabbadoo.comwrote in
message news:E5**********************************@microsof t.com...

Act makes pretty realtime charts, but when it comes to accurately throwing
a
multi-client stress test at an app, it falls on its face. Homer (Web
Stress
Tool) is much better. There are others put out by commercial vendors that
are
even better.
Peter

--
Site: http://www.eggheadcafe.com
UnBlog: http://petesbloggerama.blogspot.com
Short urls & more: http://ittyurl.net


"Chris" wrote:

>We have an asp application some of which will be being ported over to
asp.net so this may be a little OT. It is having some bad performance
problems and we need to isolate which areas are causing the problems. The
main performance problems are database related i.e. timeouts and
deadlocks

Some questions:
Microsoft's ACT. How is it better than the Web Stress Application Tool?
Is ACT purely .net or can be used to look at any web app.
What is a good way of assessing which pages are causing the problems.
How can I relate the performance of the database to the application. The
system has LOTS (this app has 1000's of pages) of sql passed as strings
will
this affect how easy the profiler is to read.
I suspect there are lots of memory leaks, are there any tools to look at
what is being left in the memory.

I guess what I want is a tool which will identify which pages are hogging
the most resources and if they are not being released so we can look at
optimising the code.