Yes, I have seen this before when an error in appendToResponse of a
WOComponent prevented the session to get checked back into the
SessionStore at the end of the R-R loop.
To prevent the thread exiting and respawning and to get the error
stack trace, I used Wonder's
er.extensions.ERXApplication.useSessionStoreDeadlockDetection (which
requires concurrent request handling to be turned off for that to
work). Once you find the problem and fix it, you can turn back on
concurrent request handling. This worked a treat and helped me find
and fix the problem quickly.
For the mechanics of how it works see ERXApplication and the inner
SessionInfo class and see references to the method public boolean
useSessionStoreDeadlockDetection().
HTH, Kieran
On May 8, 2009, at 10:29 AM, Ricardo J. Parada wrote:
Hi All,
I've used eomodels that point to different databases without
problems before. For example, one points to a reporting database
where we run queries and then the report is stored in the live
production database. This is so that we don't impact the main
database because the volume of reports gets very heavy.
Anyways, recently we split an eomodel into two so we could point one
eomodel to another database that will hold some reference data.
The entities in one model have relationships with destination
entities in the second model.
The problem I'm seeing is that when I set a to-one relationship
which is the inverse of a to-many relationship I get this exception:
(com.webobjects.foundation.NSKeyValueCoding$UnknownKeyException)
com.sun.jdi.InvocationException occurred invoking method.
This seems to happen within EOF inside
com.webobjects.foundation.NSKeyValueCoding._KeyBinding when the
setValueInBinding() method is getting called.
I can't print the stack trace because attempting to do so creates
another exception and nothing prints to the console. I also see
this in the console in Eclipse while the application is running and
I attempt to set the to-one relationship in the EO :
[2009-5-7 23:18:25 EDT] <WorkerThread13> <WOWorkerThread id=13
socket=null> Workerthread exiting due to error, respawning with ID
10013...
I was wondering if anyone has seen anything like this before? Or
does anybody have a clue what would cause a worker thread to exit?
I reverted to a single eomodel and the problem goes away... I'm
gonna attempt the eomodel split again to see what happens.
Thanks,
Ricardo
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/kieran_lists%40mac.com
This email sent to kieran_li...@mac.com
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com