Master instance encountered an error (TD Connector)

bcarlson · Post by **bcarlson** » Tue Jul 21, 2009 2:57 pm

We are getting a strange error in a simple DataStage v8 job. We are on DataStage 8.0.1 and Teradata V2R6.2. The job reads from a dataset and uses the TD Connector to load to a target table (bulk load).

work_dly_loc_stmt1,12: [IIS-CONN-TERA-005027] Unable to issue a checkpoint at 133,167 rows because the master instance encountered an error (CC_TeraAdapter::doCheckpoint, file CC_TeraAdapter.cpp, line 5,869) [pxbridge.C:2945]

Has anyone seen this before? I searched around DSXChange and couldn't find any references anywhere.

Brad.

bcarlson · Post by **bcarlson** » Mon Mar 26, 2012 3:23 pm

Hmm, I am now running into this exact same issue with DS8.1. I see no one had any ideas about my original posting. Maybe someone knows now what to do? (DS8.1 FP1, TD13.0)

bcarlson · Post by **bcarlson** » Mon Mar 26, 2012 3:26 pm

By the way, i know we ahve run into this when the Record Count parameter is set to something greater than 0 and the load operation is set to bulk. In that context, DS uses the Record Count to indicate how often the load should be checkpointed. If you leave it at teh default of 0, then checkpointing is done once at the end.

In the current job, Record Count is already set to 0.

Brad.

bcarlson · Post by **bcarlson** » Mon Mar 26, 2012 3:31 pm

More details. I searched online and found one reference in DSXChange (my own, with no posted responses except my own). I found another from IBM (here) that references the same thing in DS8.0.1 and a potential APAR that should have been included in the DS8.1FP1. It recommended changing the sync poll option from 0 to 2, 3 or 4. The job was already running with 2, so I tried 0 and 3 and no change.

The issue also appears to be intermittent. I ran the same job this AM and it ran okay, now it will not run successfully at all.

Brad.