We are having a strange issue where MaxL commands are not completing on certain commands (reset, execute calculation, execute aggregation) randomly.
This past weekend an AIX 6.1 patch, TL07) was applied to all Essbase servers due to an emergency with the system paging , we seem to have the problem on one of the three only (so far). We are running Essbase 220.127.116.11.102, the support matrix shows TL2+ supported, My Oracle Support does not show any knowledge base articles about TL07.
We are pretty stumped here and have been able to complete the commands from EAS, then kill each of the MaxL scripts so the rest of processing runs from the shell scripts, but we can't continue to do this every day. We stopped and restarted Essbase yesterday, this appeared to work for a short time, two of the 4 automations we restarted. Last night, some worked and some hung up on the mentioned commands from the scripts.
I will open a ticket with Oracle, but hoping someone has some information they can share.
There are no messages in the logs, the MaxL command is not logged, for instance, for reset, there is nothing in the log and the data is not cleared. For the calculations, it is not executed.
Using essmsh from unix command line to run the maxl.
the commands don't even begin.
Try using startmaxl instead of essmsh.
startmaxl can be found under <MIDDEWARE_HOME>\user_projects\epmsystem1\EssbaseServer\essbaseserver1\bin on Essbase server.
I dont know how to append above path in Unix system variables. but should be similar to what we do in Windows i.e. adding path to system variable PATH
or From command line, go to above path & run startmaxl.
See if this helps to narrow down the issue.
I don't mean to be disagreeable, however we have been running this release for over a year without an issue. I'm thinking this AIX patch is affecting the system unexpectedly or the admins did something different on this server compared to the others. Last week everything was ok, then after the patch it breaks. I did test on our development server and was able to reproduce one time. restarting essbase services and reran several times to see if it would fail again and it did not fail to run.
We have incorrporated the startmaxl script settings with our environment so we do not have to run it that way due to extensive automation scripts would have to change.