1

Web Server

  • AD with 2008 Domain Controllers
  • Windows Server 2003
  • IIS 6.0
  • Virtual Directory / ASP.NET 4.0 / MVC 3
  • Anonymous / Integrated allowed

Device

  • Motorola MK 3000 Price Checkers
  • Windows CE 6.0
  • Pocket Browser 3.0
  • Wired

The problem

In only one of our locations, maybe once every two weeks, all the devices become unable to connect to the web server. The lowest level of detail that I can see is that the requests going out from these devices stop getting logged in the IIS logs.

Questions

  • What can we do to troubleshoot where the failure is? Wireshark? Firewall settings?
  • A server reboot always fixes the issue - What changes? Is there a service that helps pass through non-domain users that is failing somehow?
  • (Speaking as a developer) Is there anything involving Kerberos that we should be looking at?

If you need more info just leave a comment and I will try and gather it.

Any and all ideas are welcome. Thank you in advance for your help.

Improve this question
4
  • 1
    Can anything else on the web server other than your app be accessed during this time?
    – squillman
    Jan 29, 2013 at 14:01
  • Yes, everything else functions as normal. Also - forgot to mention - authenticated users (e.g. me from my PC) can still access the web application.
    – IronicMuffin
    Jan 29, 2013 at 14:14
  • Does restarting the w3svc or the application pool clear the symptom?
    – Greg Askew
    Jan 29, 2013 at 14:15
  • I don't think they've tried that. I will have them give it a shot. Would a recycle suffice, or a full on stop/start?
    – IronicMuffin
    Jan 29, 2013 at 14:25

3 Answers 3

Reset to default
1

If IIS isn't even logging the HTTP requests, it broke and you'll need to add some error checking/logging to your application to help figure out what it is (if you can't find it from the Windows event viewer and the HTTPERR logs.)

Improve this answer
2
  • I'm sorry I forgot an important point above - domain devices such as my PC can still access the web application.
    – IronicMuffin
    Jan 29, 2013 at 14:14
  • More info: There is nothing that stands out in the event logs on the server. As far as the client side, it redirects to it's standard error page due to the failed request. We will have to contact the vendor to see if there is any valuable logging on the device.
    – IronicMuffin
    Jan 29, 2013 at 14:28
1

Since other things are working when your app is broken I would start by recycling the app pool that your app is in. If there are other apps sharing the app pool then you should create a new app pool to isolate your application and see where that takes you.

You could also try completely recreating the app pool (delete it and create a new one). I've experienced app pools getting toasted when there was nothing wrong with the app running in it. I had the similar symptoms you're showing and the new app pool worked.

Improve this answer
1
  • Spoke to one of our admins. They had tried an iisreset with no luck. Apparently, they've also reset the app pool on previous occasions with no luck. We are going to try and recreate it now to see if that helps.
    – IronicMuffin
    Jan 29, 2013 at 14:37
0

Turns out the problem was due to DNS being configured incorrectly for one thing, and the other thing being NETBIOS failing to resolve the correct web server for the devices.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .