FLTABSZ setting has any relation with DSJE_TIMEOUT?

videsh77 · Post by **videsh77** » Tue Nov 13, 2007 11:11 am

Hi

We are getting following error

Status code = -14 DSJE_TIMEOUT
Error Return from ascential job with error code 255

consistently on random jobs when tried to run at same time. Those jobs which got aborted have one thing in common, they all read files from disk. The number of jobs aborted at the same time were always 11.

I tried further investigation with DSEngine & found in uvconfig have a setting named 'FLTABSZ' is set to 11. It says - it sets number of file lock entries in a file lock semaphore set.

I have a Qn - The setting for 'FLTABSZ' has any obligation with job timeouts? Would increating value associated for 'FLTABSZ' will affect in I/O subsystem?

richdhan · Post by **richdhan** » Tue Nov 13, 2007 12:53 pm

Hi Videsh,

How long are these jobs running for?

Find out what is the Inactivity Timeout set in Datastage Administrator.

Check if the jobs are running for a longer time than that specified in the Inactivity Timeout.

Also do a search on DSJE. There are many error codes.

Increasing the Inactivity Timeout resolved the issue in our case.

HTH
--Rich

videsh77 · Post by **videsh77** » Tue Nov 13, 2007 1:40 pm

Hi Richdhan

We get this error mentioned when job is accessed to run by scheduling component. Interestingly in DS Director, we dont get the log of this message, as this was a timeout.

Its more I could see many number of jobs try to read files from the disk. Also which environment variable best defines inactivity timeout 'APT_????'?

ray.wurlod · Post by **ray.wurlod** » Tue Nov 13, 2007 4:43 pm

FLTABSZ sets the depth of the file (table) lock table in memory, FSEMNUM sets the breadth of the lock table in memory. Neither would cause a timeout error - if FLTABSZ were too small you might get a "lock table full" fatal error, but this would be quick, not a timeout.

videsh77 · Post by **videsh77** » Tue Nov 13, 2007 7:33 pm

This means FLTABSZ has no relation with timeout error.

As the sequential files which are read, are not Big. Some of them are empty. Hence for this DataStage job, which is being run in multiple instance reading these sequential files, which setting can solve this problem of timeouts?

As we have many other DataStage jobs which are running in parallel & have DB iteraction, never got this sort of an error.