We have session setup like below where we have a multi-node session where the Header and Trailer Uproc are going to run on the Solaris Node on MU LROPDB and the intermediate UPROC going to run on the Windows Node on MU LRTSI103.
The setup of session is like the UPROC on MU LRTSI103 has condition on *_Header UPROC and trailer had condition on UPROC on LRTSI103
We observed that on some time the session execute in sequence i.e. UPROC_HEADER (LROPDB) > UPROC (LRTSI103) > UPROC_TRAILER (LROPDB), but sometime the UPROC (LRTSI103) > UPROC_TRAILER (LROPDB) goes in Event Wait status even though the condition is fulfilled as shown below.
Can you let us know the cause and how to correct the same.
Release : 6.10.X
Component: DOLLAR UNIVERSE
From looking at the logs we identified that the cause of the behavior being observed is due to fact that the session execution is not happening in sequence as you expect. As highlighted in below screen shot the sequence of same "Session Execution" one for failed and one for passed the difference is that the passed one is executing in right sequence i.e. on date 08/01/2023 at 19:30:08 19:30:10 19:30:14 whereas the failed one all UPROC in session executed on same date 06/01/2023 at time 19:30:09.
We observe some wrong session design configuration in the session and we would like to recommend a change to the same.
1: Remove the dependency condition among the UPROC i.e. UPROC > UPROC_HEADER and UPROC_TRAILER > UPROC, as it is not required the Session hierarchy in the session take care of sequential execution.
2: As you want a sequential execution please change the session hierarchy in session like below i.e. UPROC_HEADER > UPROC > UPROC_TRAILER
Save the changes in the session and make sure the updated objects are available on all nodes and try to execute the same.