QUERY RUN ON ONE SERVER TAKE LOT TIME

Question

I HAVE 2 SERVERS WITH O/S SOLARIS. ONE IS PRODUCTION AND SECOND ONE IS TESTING SERVER
I HAVE A PROCEDURE WHICH WHEN I RUN ON TESTING SERVER RUN FINE BUT WHEN I RUN THE SAME PROCEDURE ON PRODUCTION SERVER IT STILL RUNNING AND THERE IS NO ERROR IN ALERT LOG FILE AND ALSO IN LISTENER LOGS.
THERE IS ERROR IN ALERT LOG FILE LIKE ORA-03136 inbound connection timed out (ORA-3136).
BUT I DONE THE SQLNET MODIFICATION THAT IS SQLNET.INBOUND_CONNECT_TIMEOUT TO 0
MY DB VERSION IS 10.0.2.0
can any one will help
regards,
vikas kohli

Answer 1

Vikas Kohli wrote:
I HAVE 2 SERVERS WITH O/S SOLARIS. ONE IS PRODUCTION AND SECOND ONE IS TESTING SERVER
I HAVE A PROCEDURE WHICH WHEN I RUN ON TESTING SERVER RUN FINE BUT WHEN I RUN THE SAME PROCEDURE ON PRODUCTION SERVER IT STILL RUNNING AND THERE IS NO ERROR IN ALERT LOG FILE AND ALSO IN LISTENER LOGS.
THERE IS ERROR IN ALERT LOG FILE LIKE ORA-03136 inbound connection timed out (ORA-3136).
BUT I DONE THE SQLNET MODIFICATION THAT IS SQLNET.INBOUND_CONNECT_TIMEOUT TO 0
MY DB VERSION IS 10.0.2.0
can any one will help
regards,
vikas kohliThere are a number of possible reasons, but with such little information there's no real way for us to tell you.
http://coskan.wordpress.com/2011/06/24/plan-is-same-wait-profile-is-different/
Outlines one example where test/prod instances are "identical" yet produce different access paths for a query. If you read the post you'll get an appreciation for the myriad of different reasons you could have differences.
Some more common reasons would be
1) data volume differences between test/prod
2) hardware differences between test/prod
3) inaccurate statistics on objects involved in the query / procedure in question (gathered with the DBMS_STATS package) in either test or prod (or a difference between the two)
4) load differences between test and prod (test has 2 concurrent users, prod has 2000 for instance)
the list could go on for quite some time.