How do I increase the heap size of the java heap for the vme manager

[mikejdunphy]

So I was uploading a 200+gb image file and it was going to the nfs image  datastore but somehow the vme manager
acting as a middleman I guess consumed all its memory and would not even allow logins. I killed the upload and things
got all better. I also increased the size of the vm from 2cpu/12gb to 4cpu/24gb but a 'ps aux | grep morpheus | grep java'
on the manager shows that the size is currently 3gb so it didnt increase.

How can I increase the resources for the vme manager? I imagine I just need to find the config file and add  "-Xms4g -Xmx8g"
to the java opts to set the start and max size. Appears to be nothing in the current gui to configure these settings

After I increase it I probably need to stop/start the manager ... how do I do that ? Is there a systemctl I can run ? I found these

systemctl list-unit-files |grep -i morp
morpheus-node-runsvdir.service enabled enabled
morpheus-runsvdir.service enabled enabled

The next time I do a install if I do one I will select at least a medium sized vme manager.

[Ranveer]

Hello,
Once Memeory increased from Manager GUI, it is immediately reflected on Manager VM
No other Manual action needed from VM Manager's OS.
When GUI was showing 16GB, below was the memory seen on VME OS.

root@hpevmemgr2:~# free
total used free shared buff/cache available
Mem: 16302620 12269724 570144 18860 3462752 3667740
Swap: 4194300 1374420 2819880

Increased Memeory to 24 GB in GUI:

Immediately detected on the OS. 
root@hpevmemgr2:~# free
total used free shared buff/cache available
Mem: 24691228 12378040 9103780 18872 3209408 11884060
Swap: 4194300 1381332 2812968

We dont need to take any actions on VME OS, the update should be populated from GUI.

Regards

Ranveer

[mikejdunphy]

i dont think you understand the question

Yes the memory immediatley increased when I did to

root@hp-vmemanager:~# free
total used free shared buff/cache available
Mem: 24738012 9027884 14365388 4280 1763676 15710128
Swap: 4194300 2485520 1708780

The issue is the vm never ran out of memory but the java process  that runs the manager and all the backend stuff which is famous 
for memory leaks and not releasing memory filled up and basically rendered the manager useless., folks could not even login ... it was locked
up tighter then a 2 dollar watch.

I killed the upload and instantly it came back to life

So I am asking how to increase the resources particularly memory of this nasty tomcat java engine so that I can live with these
java bugs or features

[Ranveer]

Hello 
Thanks for making it more clear, if you have a valid support contract, i will request to opne VME case with HPE Support
We will be happy tio take it up to Engineering.
Regards

[KoreyG · Accepted solution]

Hello @mikejdunphy,

After changing the memory for the VM, login to the manager via the terminal and use the following command:

morpheus-ctl reconfigure

Changing the memory is not enough, as you mentioned, but the reconfigure command will look at the configured memory and change the heap and other services memory configuration accordingly based on the new value.  You may need to restart the service afterwards using morpheus-ctl restart morpheus-ui, or just reboot the VM entirely, but usually is not required.

Hope that helps!