We are currently running V9.3.5.
Sunday Feb 26th we migrated our production Appman from Oracle VM 12c to 19c.
The migrated database is on the Oracle Exadata appliance which supports SCAN RAC. We know
Appman does not not do well with this, and was advised by Appman support we need to point to
a single node. We did this for our upgd, and test instances and have been running very well.
Well this is production, and you never know what to expect. Well since that Sunday only a
little more than a week ago we had some problems show up. A few days into the migration
we started seeing jobs get stuck in launched state, and eventually change to launch error.
When this happens it usually happens to most of the jobs that try to start. If you reset
a job in launch error state it just gets hung again.
We then have to stop Appman, and start it. Most of the times it won't stop, and we need
to kill the Appman processes. Then we start it, and reset the jobs. The jobs then seem to
run well for a while.
Just today this happened multiple times. We decided the last time this happened to also
recycle the database. Things have been running ok for the last few hours. We will see how
I reviewed the dstabase connection information in the logs where you see the procedure
being called, and the ms times. I tried to look for the ms times when the problems
started and that did show some higher than normal values. When the problems did occur
I asked the DBA to look at the database for us, and was told things look very quite.
I am at a loss for what else to check. We did question the ojdbc driver we have in place.
The database is at 19.17. The ojbc8.jar I am using is the 19.15 java 8 build. It supports
12c - 21c. I had this on before we migrated to 19c since it supports 12c. I was thinking
of looking at what newer ojdbc driver packages are available in case it fixes issues
in the one we are on.
Thank you everyone,