RuntimeError: workspace already in transaction mode

[TomSellsted]

Greetings,

I have a python script that I run continuously to update and insert records in a SQL Server Express database.  This has been working properly for many months until this weekend.  When I try running the script, I get a "RuntimeError: workspace already in transaction mode" error and won't execute any further. 

I have tried various solutions from posts I have seen, but none have solved this problem.  None of the feature classes are registered as versioned.  I am running SQL Server 2008 R2. 

I am a bit mystifed to know what has changed that would cause this problem.  I appreciate any help!  Thanks very much!

Regards,

Tom

[XanderBakker]

Are the two cursors that update/insert features running at the same moment on the same workspace? If so, did that work before? Is the script running as a scheduled task and is it possible that there are two instances of the script running at the same time?

[XanderBakker]

Or better yet: ArcGIS Help (10.2, 10.2.1, and 10.2.2)

Note:

Opening simultaneous insert or update operations on the same workspace using different cursors requires the start of an edit session.

So, use an edit session to avoid the problem

[TomSellsted]

Xander,

Thanks again.  I have never had to set these up as registered or use an edit session.  They have been running without those for a really long time.  That is why I am wondering what has changed about the environment that would create this problem.

I would really  prefer not to register them and use an edit session.

Regards,

Tom

[TomSellsted]

Xander,

Thanks very much for the quick response!  The two cursors are running consecutively.  The update occurs first followed by the insert.

This script has been running for over 3 years without any modificaton.

I actually have the script running as a service using the srvany.exe.  This has also worked very well.  Since it is not functioning, I stop the service and I run the script manually to see what is going on.

I have looked to see if multiple instance are running, but I only see one instance in the processes in my Task Manager.

Regards,

Tom

[XanderBakker]

Interesting... Would it be a lot of work to adapt the script and see if using an edit session resolves the problem?

[TomSellsted]

I could certainly copy my script and try it.  I am wondering if the performance would slow down greatly by using an edit session. 

The python script is listening on a UDP port for vehicle message and updating a vehicles feature class and inserting a record into a historical position feature class.

Without an edit session it takes about .02 seconds to process a vehicle position.  I am assuming that a edit session might take a bit longer.

Regards,

Tom

[XanderBakker]

mmm, I see your point... Not really sure how much the process would loose in performance when using an edit session. If you call your script very frequently, I can imagine that this will not be an option.

[TomSellsted]

Xander,

It looks like I may have reached a finite limit of the geodatabase size.  This may be causing my problem.  I am removing some older position data and see if that is what is causing my problem.

Thanks very much!

Regards,

Tom

[XanderBakker]

OK, great. Please post back in case that is the reason, so others can benefit from it.