Rough ASP.Net Server efficiency question

[Denaes]

Yes, I'm still new to ASP.Net 1.1 & 2.0, but I've been using WinForms for a few years (since 2003 I think?), so maybe my perception on bottlenecks is skewed here.

 

We're running an enterprise level ASP.Net 1.1 application which also has many pages still in Classic ASP. The Server is "2003 Server" on a modern cutting edge server... core duo or core quad or whatever.

 

The database is SQL Server 2000.

 

This app has been around for a while, so this hasn't always been the OS or server computer.

 

In comming to this project, the developers keep on saying that we need to handle things clientside when possible. Not for ease of use, but to save the server from bottlenecking.

 

We have like 400 users max, spread throughout the day (if everyone had to use the app in a single day, which isn't always the case). How would you figure out how many concurrent users? Maybe a few hitting at relatively the same time.

 

I don't know, it just seems that we're stressing ways to prevent serverside processing and from my perspective, it doesn't even seem like we're comming close to maxing out the Web Server or SQLServer's capabilities, yet we're doing silly optimizations to avoid the overhead of a datatable.

 

Maybe I'm just spoiled being a winforms developer. Any thoughts?

[PlausiblyDamp]

As a very basic tool you can get MS's Application Center Test (it ships with some version of VS anyway) - it's basic but does allow you to record a user using your site and then replay it back (as multiple simultaneous users as well) and will allow you to record various statistics about the site's behaviour. Also windows itself include Performance Monitor which is also a good tool for identifying potential problems.

 

Classic ASP might be causing some problems due to it's interpreted rather than compiled nature - but unless the people who are saying there are performance issues can back this up with some evidence it would appear you are wasting your time...

[snarfblam]

It seems to me that running scripts on the client side instead of making more logical and better structured code, even if it is on the server side, is tantamount to micro-optimization. In other words, the extra effort and cost that goes into so-called optimizations is not made back by the negligible gain in performance. Even if the gain is measurable, it isn't necessarily a wise investment. Consider how much more work is being made for the programmers.