Dave,
I assume that with a server getting that amount of traffic, you've got Application and Include file caching turned on. The behaviour you are seeing is similart to what I get on 065 when I clear the cache on the server after uploading new taf or tcf files.
/John
Dave Machin wrote:
Load - we just did a restart, the server has been up for 27 minutes, and has read 5821 files (according to serverstatus) - so that's about 215 per minute. Over longer periods we average much less, about 150 per minute.
There are FOR loops in code - if there was an infinite loop somewhere I would expect to see the same .taf files in the logs immediately prior to the event - or would the stuck process not get logged?
Good question on the CPU, WiTango uses 100% CPU when this happens.
Nothing unusual in the events log.
----- Original Message ----- From: "Peter Ternstr�m" <[EMAIL PROTECTED]> To: <[email protected]> Sent: Thursday, January 13, 2005 10:58 PM Subject: Re: Witango-Talk: Any Help Appreciated - WiTango Stops Responding
beingDave,
- What load are you experiencing? Requests a day.
- Do you have any FOR or WHILE loops in your code?
- When the server stops responding, what does the "task manager" CPU meter say? How much processor power is the witango service using?
- Check the events log. Any unusual events before the server stops responding?
Peter
----- Original Message ----- From: "Dave Machin" <[EMAIL PROTECTED]> To: <[email protected]> Cc: <[EMAIL PROTECTED]> Sent: Friday, January 14, 2005 2:15 AM Subject: Witango-Talk: Any Help Appreciated - WiTango Stops Responding
We're having intermittent problems, usually once per day, sometimes more often.
Witango (.065, running on W2K SP3) just suddenly stops responding. The connections to our IIS web server start climbing and Witango never responds until killed and restarted.
We thought it might have something to do with our database (MS SQL 2000 SP3) but we've examined the SQL profiler traces and stepped through the IIS logs and WiTango does not appear to be issuing any queries at the time it stops. In fact, all open WiTango connections eventually go away if we leave Witango unresponsive long enough.
There are no events in the witango events log, related events on the system, or witango log messages written after Witango stops responding. IIS starts writing 500 errors about a minute after Witango stops responding to it. When Witango stops responding we see a build up of "dllhost" port connections (that we normally see come and go) that never go away. I assume these are attempted connections between IIS and Witango that get stuck.
We so far can see no common events prior to Witango's demise - the
immediately prior application requests and queries are always different.
The last event we had happened between a form being displayed and it
asubmitted (in other words, the form was successfully requested and then
doesn'tfew seconds later the form is submitted but WiTango is no longer responding).
The only other clue is that it seems to be related to system load since
the
outages only happen during the day and only on week days. But, it
seem to be a particular application causing the problem.
We're getting desperate on this one - we'd be willing to discus paying someone to help us solve this problem.
Dave Machin
________________________________________________________________________
TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
________________________________________________________________________ TO UNSUBSCRIBE: Go to http://www.witango.com/developer/maillist.taf
