IIS 7 and Above
Browser Not Responding after w3wp crash
Last post Jan 02, 2018 07:24 AM by Yuk Ding
Oct 13, 2017 10:54 AM|sherwincool|LINK
We used a Session State saved to SQL Server to allow web garden/farm, but when w3wp crashed, the browser that triggered it stop responding. Even we created a new tab (Chrome) and access the website, its still not loading. The only solution to recover is
to close the browser and reopen it. We already see the bugs that causes the crashing but my concern is How are we going to respond to the browser if other crash encountered again.
We use previously InProc and it just end the session. Or maybe Im just missing some settings like response timeout etc.
Below is the section of web config:
<sessionState mode="SQLServer" customProvider="DefaultSessionProvider" sqlCommandTimeout="60" sqlConnectionString="data source=SQLSERVER;Password=PASSWORD;Persist Security Info=True;User ID=USER;">
Oct 16, 2017 05:49 AM|Yuk Ding|LINK
In fact, I think the session state section could be an asp.net section. So you could try to set the sql server session state following this link:
https://support.microsoft.com/en-us/help/317604/how-to-configure-sql-server-to-store-asp-net-session-state Or add the failed request
I'm a bit confused why don't you try to fix the worker process crash?
Oct 16, 2017 08:32 AM|sherwincool|LINK
Thank you Yuk!
We will fix the worker process crash, but my concern now is if we already fixed it and there will be another crash (somewhere along the way), we will experience again the non-responsive browser. What we are trying to solve here is the recovery of the browser
after the crash.
I also noticed in Chrome debugger tool that the request status is still pending.
Or add the failed request
Please enlighten me.
Oct 18, 2017 08:04 AM|Yuk Ding|LINK
I think maybe it has nothing to do with the session state. I remembered the session will get closed when you close the web browser. If you don't close the web browser, the client side should not close the session, then it should keep hanging. Maybe you need
to set the session time out or expire time.
Nov 07, 2017 11:37 AM|sherwincool|LINK
I have fixed the locking of session after appcrash by overriding the session state of action to Read Only. I just noticed that the IIS tries multiple time to process the request that causing the crash resulting to multiple appcrash. Is there any relationship
of that request being ASYNC?
If I set 3 worker processes, after 1 worker process crashed, other worker process tried to accommodate that request resulting to multiple appcrash. I just experience 6 continuous appcrash for that particular request.
Jan 02, 2018 07:24 AM|Yuk Ding|LINK
I think it should be related to the async. In addition, if the crash effect the other worker processes, you could try to leave the session state in process.