Nasty problem with the .net site ...

[Guest]

Hi there,

I'm about to commit suicide Since a couple of weeks I have a persistent
problem with the performance of my sites (well, actually it's one particular
site). For some reason the site is very unresponsive.

What happens? Well, the site just becomes unresponsive after some time. It
hangs, without giving an error whatsoever, not to the user, no errors in the
Event Viewer either. The processor isn't going sky-high, no excessive
RAM-usage, etc. Everything looks normal.

The funny thing is, this site has worked good for very long time (+3 years),
and there is no significant change. Apparantly something must have changed,
but I can't recall/find what did change.

I'd think it has something to do with a database call that is not being
closed, but I double-checked everything and can't seem to find anything.

What do I do, to get it working again, each time it hangs: I just "recycle"
the AppPool in the IIS Mgmt Console, and then it works again for some time
(depending whether it's busy on the server on not, this can be 2 minutes or 4
to 5 hours).

Does anyone has some kind of advice where I should continue my search? I've
tried a lot of things, like caching data from the database, so I would reduce
my database-traffic (site and database run on separate servers, website is
IIS 6 with aps.net v2.0, database-server is sql2000).

I'm not looking for an exact fix (which should be cool, but with the
information given, I don't see that happening), but would be nice to hear
from you guys if someone had similar problems in the past and how you got it
fixed (or how you could locate the problem).

Tx in advance!

Greetingz,
Koen Hoorelbeke.

 

[Guest]

Have you considered that it may not be a problem with the site, but a problem
with the server. Have you tried loading the site onto another server for
testing purposes to see if it does the same thing.

 

[Chris Fulstow]

Hi Koen,

Sounds like a tricky one. I don't know what to suggest really, other
than turning on as much ASP.NET tracing as you can, and perhaps using
SQL Server Profiler to try and identify any queries, transactions or
connections that aren't terminating properly.

As a temporary measure, maybe you could configure the application pool
to recycle the ASP.NET worker process every hour (or maybe a certain
number of requests) and see if it makes any difference.

Good luck,

Chris

 

[wfairl]

Put away the knife, rope, or (if you're a wuss) pills. Use adplus and
the Win debugging tools to get a look into what's going on with the IIS
process when it "hangs". If you're unable to read the dumps yourself
then try posting to one of the windows debugging forums or contacting
Microsoft PSS.

http://support.microsoft.com/default.aspx?scid=kb;en-us;286350

 

[Guest]

Tx for the advice. I tried this one, but the outcome was rather catastrophic
Ran it on the webserver (in production) and as a result memory usage went
sky-high. Couldn't get the processes stopped, eventually the machine went
dead. And since I tried it at night (luckily), I had to wait till the
sys-admin went to the colocation to give it a hard reset in the morning.

But I think I found the reason of all my problems. Eventually I could
isolate a webservice which went crazy after a certain amount of time.
Actually the webservice didn't do anything wrong. But I copied it from
another server and on that server it was housed in it's own application. On
the new server it didn't, and for some reason it got mixed up with the parent
application after some load.

Weird that I didn't have this problem in my development environment.

Oh well, learned something again It's one of those things that I had to
do under deadline-pressure. Seemed ok to do at that time. Lesson learned:
never give in to mgmt when it comes to critical applications.

Tx all for the input!

Greetingz,
Koen Hoorelbeke.