It was a perfectly fine evening in support, till we got a mail saying client connection to datawarehouse production database is not working..first instinct we thought issue at client end, even a test connection i did from my client went fine to the database. but later other users started reporting, there ETl reports were failing..now we realised we have a problem at our hand.on examination of the alert log we found that local_listener parameter value got set to a wrong value autmatically....puzzling...
we set it back to its original value and issue got resolved..
SR was raised, Oracle confirmed it is a internal bug
bug 11772838 .please refer the note note ID 1384707.1: ORA Agent modify or overwrite LOCAL_LISTENER
How to avoid this issue in further..the note gives a workaround.
set local_listener value in the below format
ALTER SYSTEM SET LOCAL_LISTENER='(ADDRESS=(PORT=<port#>)(PROTOCOL=TCP)(HOST=<host>))' SID='<instance>';
One thing that puzzled me more is the RAC concept of high availability,obviously we had issue with local_listener value in only 1 node, being a 6 node RAC how was this affecting the client directly. clients are in fact using the scan listener to connect to db.
we set it back to its original value and issue got resolved..
SR was raised, Oracle confirmed it is a internal bug
bug 11772838 .please refer the note note ID 1384707.1: ORA Agent modify or overwrite LOCAL_LISTENER
How to avoid this issue in further..the note gives a workaround.
set local_listener value in the below format
ALTER SYSTEM SET LOCAL_LISTENER='(ADDRESS=(PORT=<port#>)(PROTOCOL=TCP)(HOST=<host>))' SID='<instance>';
One thing that puzzled me more is the RAC concept of high availability,obviously we had issue with local_listener value in only 1 node, being a 6 node RAC how was this affecting the client directly. clients are in fact using the scan listener to connect to db.
