Discussion created by sherriekubis on Aug 23, 2016
Latest reply on Aug 23, 2016 by sherriekubis

Oracle RDBMS 

SDE 10.2.2


We have many nightly update jobs that use arcpy, others use FME.  Sporadically, some of the arcpy jobs fail.  Since they run in batch mode, the developer is having trouble seeing what the returned error is.  Sometimes the jobs all work without issue, other times maybe 5 to 10 of them fail.  


To determine the failures I created a login trigger for the data owner between the hours of the nightly updates.  I found 

enq:  TX - row lock contention on GDB_TABLES_LAST_MODIFIED.  It was previously suggested to me to increase initrans from 4 to 8, but that made no difference.  In the development environment, I increased it to 50 and it still made no difference, so I set it back to 4. 


Has anyone had similar issues?  


The failures happen with arcpy jobs, but not with FME jobs, both from the same Windows server as the client to the database.