It all depends on how calc manager is clustered. When you add a new calc manager node to the cluster, you would need to restart all calc manager services.
Now it also depends on how it is clustered. Do you have a single load balancer that handles both servers? Are they both configured on the same Shared Services? Typical synch time is 10 seconds.
When an object gets modified on one server, it synch with other servers via a database table. One write the changes, others read the changes..
What version is this? 18.104.22.168 or 22.214.171.124.500?
Thanks for your response. After adding the second calc manager WLS instance on the same EPM 126.96.36.199 server (vertical clustering), we restarted both OHS and the two Calc Managers (CalcMgr0 and CalcMgr1). OHS switches between the two Calc Managers and uses the same Shared Services.
When a business rule gets updated or created, the only method for re-synchronizing is to restart the two Calc Manager instances again.
You indicate that the synch time is 10 seconds. Where is this configured and how can we verify that this synchronization is occurring? What else might be causing what looks like one instance updating but not the other?
Btw, we're running 188.8.131.52.0, with plans to update to 184.108.40.206.500.
I think you would need to file a support call for us to take a look. Something is amiss here. You have 2 OHS running. Are they on 2 different boxes using the same shared service? How are these 2 OHS controlled? Do you have a load balancer in front of it?
Run the epmsys_registry.bat and look at the output. Search for CALC_LWA (LOGICAL_WEB_APP). Look at the children of that. How many CALC_WEBAPP do you see?
Based on that I can tell if it is a problem with the configuration.
In the meantime,
If you can get to the exploded directory of CALC manager, (On my Windows box it is C:\Oracle\Middleware\user_projects\domains\EPMSystem\servers\CalcMgr0\tmp\servers\CalcMgr0\tmp\_WL_user\CALC_220.127.116.11\qxf61x\war), edit CalcMgrConfig.properties under WEB-INF\conf
Add the following line to the end:
Now save the file and REPEAT the same for the other instances of CALC manager.
Now restart all CALC servers and see if the issue is resolved (just a temporary check to see if it working without any issues)
We opened Oracle SR #3-9181641801. https://support.oracle.com/epmos/faces/SrHome?_adf.ctrl-state=qb1f8hy70_9&_afrLoop=236576817261359We have one Exalytics server running OHS with two Calc Mgr instances (vertical).
The EPMSYS Registry report (registry.html), shows:
- One CALC_LWA (LOGICAL_WEB_APP)
- Two CALC_WEBAPP ('CalcMgr0' running on port 8500 and 'CalcMgr1' running on port 8501)
Is this configuration correct?
I made the change to the CalcMgrConfig.properties that you indicated and restarted the Calc Manager instances.