cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Want the oppurtunity to discuss enhancements to PTC products? Join a working group! X

Windchill System MethodServers crash and restart due to "java.lang.StackOverflowError"

GaryMansell
12-Amethyst

Windchill System MethodServers crash and restart due to "java.lang.StackOverflowError"

I am using Windchill PDMLink Release 11.0 and Datecode with CPS M030-CPS07

This has happened twice in the last week, and it is causing outages of my PROD Windchill System for about 5 mins whilst the FG MS's all restart (it takes out all four FG MethodServers, but not BGMS).

How do I discover what is causing this - because the logs don't give any detail, just loads and loads of the of the errors?

Thanks

Gary

ACCEPTED SOLUTION
8 REPLIES 8

Attached is the error from the log

 

And the thread dump log

Hi @GaryMansell 

 

If you know exact time, when the error appear you could check Apache log and check what user actions where run and then you could just guess what was that. What part of Windchill it could case.. 

From that error it is hard to say what is a root case. 

 for me very suspicious is org.springframework.web.servlet.view.InternalResourceView.renderMergedOutputModel

some work with visualization and model in a structure which case fatal error,

 

Sometimes it is memory or some never ending loop in application and so on. 

 

PetrH 

Thanks for getting back to me, I have checked the Apache logs as per your (sensible) suggestion - but there is nothing there that looks relevant.

 

I am no Java expert, so not sure how to read the stack traces to see what the cause was, but my understanding with stack overflows is to look for the repeated functions (as it is likely stuck in a recursive loop that then consumes all heap) - so I presume this is the loop it is stuck in:

 

GaryMansell_0-1661346663929.png

 

@GaryMansell 

it is hard to say also for experts. You can just guess what could case this loop especially if it happens one time and you can not debug the process.

Backwards investigation is not easy without debug log trace.

PetrH

Many thanks Jeremy - I will look into this and "Accept as a Solution" if it turns out to be this.

 

Any ideas what the "HTTP OPTIONS method" referred to in the article is all about? I do see some of these requests in my Apache logs and, apparently, Windchill does not use this method OOTB (but we do have a customised system that might - I will look into this further).

 

I wondered if these "HTTP OPTIONS method" calls were an indication of someone attacking our Windchill System too...

Prior to this SPR I had never heard of of the OPTIONS method before. Searching the web I see the web client can send a OPTIONS request to the server for a URL, then server will respond with the permitted methods for that URL. Some common methods are GET, HEAD, POST, and PUT,

 

Not sure why you have OPTIONS method calls. I reviewed the internal notes of the SPR, but didn't see any indication what generates an OPTIONS request for an invalid URL. Does seem like a common problem, there's 28 customers linked to the SPR. 

That's good enough for me, I am going to accept as solution - I have had another PTC contact suggest that you are most likely correct with the cause, as it was a common issue at 11.0 M030 CPS07.

 

Thanks for your help.

 

Gary

Announcements


Top Tags