Blogpost is updated for CE2.1
This little blogpost will explain how-to change the CVM memory size in Nutanix Community Edition. I assume you are here because you got an error similar to this:
As you can see the CVM has to much memory assigned and can’t start on the node. To change this we need to edit CVM parameter file. Type: sudo vi /etc/nutanix/config/cvm_config.json
When you are in the editor type the “i” button to switch to insert mode. Now you can change the file. Change the values in red to 20480 (CVM will get 20GB memory).
When ready type ESCAPE and then “:wq” (without the “”). This will save the file.
Reboot AHV with: shutdown -h now
Type: virsh dominfo <cvm_machine_name>
The new setting is active. If all is correct the CVM will autostart.
psst…the change does not persist. something about a AVH host side flag is needed (but can’t find it anywhere).
Correct. This blogpost is not valid for CE 2.1. I will make a banner on top.
In the meantime let me also try to look for the flag as well.
Did some research. 😉 Post is updated.
yep, that worked. Thanks!
This didn’t seem to do anything for me on a CE I just built – I wanted to drop the CVM RAM usage down to 10GB. The old setmaxmem virsh command didn’t work either as it was telling me
[root@NTNX-b7a0d5f0-A ~]# virsh setmaxmem NTNX-b7a0d5f0-A-CVM –config –size 10GiB
error: Unable to change MaxMemorySize
error: Requested operation is not valid: initial memory size of a domain with NUMA nodes cannot be modified with this API
So I did a virsh edit CVMNAME and just manually changed where it said 16777216 to 10485760 and that seems to have worked.
Cheers,
Steve
What version of ce are you using? As that is the method to do prior ce 2,1.
The latest – just downloaded it (phoenix.x86_64-fnd_5.6.1_patch-aos_6.8.1_ga.iso)
Did you reboot the host after changing the memory as described in the blogpost? (Just trying to get why it isn’t working so I can update the blogpost. As it was working for me and Dr. Nutanix).
No I didn’t reboot the host, just the CVM! I’m rebuilding to a different machine at the moment, so I’ll try it again when that’s done.
Yes rebooting the node is required to get it active.
Hi, I did a reboot and it worked fine on the new node I just built, thank you!
Your welcome. Have fun with the lab 👍🏻👍🏻