Forum Stats

  • 3,840,360 Users
  • 2,262,592 Discussions
  • 7,901,243 Comments

Discussions

AQ queue to queue propagation stopped working - Oracle 12c

user12212618
user12212618 Member Posts: 1
edited Mar 9, 2016 12:49PM in Advanced Queueing

I have a queue that is setup with dblink propagation that stopped propagating its messages. We are running this on several of our systems and this has been running successfully in this particular environment for almost year.  One thing to note is approx a month ago this occurred and I dropped and rebuilt the queue given its not production. To troubleshoot - I created a 2nd queue and that too stopped working after approximately 30 minutes (I enQ right after it was created, about 5 minutes later and then again after another 20 minutes) the last request did not DeQ as the first two did.    This is a low volume queue so processes only a handful of requests throughout the day on a normal basis.    I have verified the following items:

dblink for the destination queue is valid;

in DBA_QUEUE_SCHEDULES.schedule_disabled = N and there are no errors being reported. 

we are running Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production   on exadata.

When it stops working I have used  dbms_aqadm.unschedule_propagation followed by dbms_aqadm.schedule_propagation and it works for a few minutes.

Has anyone else seen this happen and / or have suggestion for further trouble-shooting?  thanks in advance.

Answers

  • kev22257
    kev22257 Member Posts: 38 Blue Ribbon
    edited Mar 9, 2016 12:49PM

    Would you believe that I came here today to ask the same question?  This issue has been plaguing us for a while now, but it's not as frequent as you describe.

    This morning we had a node on one database crash due to a known bug in the version of 11g that we are on (11.2.0.4).  This causes messages from a different database to stop getting propagated to the database that crashed (even after being back up for hours).  To correct the issue we disable and enable the propagation schedule on the source side and messages start to flow again.  The difference between your issue and ours is that we only have the issue in the event a database actually goes down whereas yours seems to happen intermittently, but maybe that is the same thing?

    I wonder if the network between your two databases is a bit questionable/ unstable?  If that's the case then perhaps the issue we are experiencing IS the same (although different means to the same result).  Perhaps a loss in network connectivity (through a bad connection or a server failure) causes the AQ propagation to halt.  The only issue is why doesn't it recover?

    Kevin

This discussion has been closed.