High CVM Memory Usage in Prism


Badge +10
Getting some high memory alerts on one of our cvms. Can I just power it off and increase the memory on the cvm or is there something special that needs to be done?
Thanks,
Erik

10 replies

Userlevel 1
Badge +2

@Eudon Can you please share below details

Output for command

cat /proc/meminfo

 

AOS version

Current CVM memory

Hypervisor version

Number of VM in cluster

Check container properties to see what additional features (dedup, erasure coding etc.) are enabled

 

Also, please refer to current Memory requirements for per CVM for AOS 5.10

https://portal.nutanix.com/page/documents/details?targetId=AHV-Admin-Guide-v5_10:app-cvm-memory-config-feats-r.html

Has any one resolved this issue yet? Same thing is happening to me. 

Also got the same issue. I still got CVM memory high even it upgraded to 36GB of CVM for all CVM. 36GB RAM for CVM is too High.
Badge
What was the result? I'm having the same issue:
Main memory usage in Controller VM or Prism Central VM xxx.xxx.xxx.xxx is high. 666356 KB of memory is free. And yet the alert policy shows critical is set at 768000 KB, so why is this even triggering? I've updated the NCC to the latest version 3.5.2 on both clusters and PrismCentral. Any help is appreciated.
Badge +7
Appreciate your response below is in the input
MemTotal: 32948044 kB
MemFree: 1640184 kB
MemAvailable: 12659008 kB
Buffers: 2821172 kB
Cached: 9114980 kB
SwapCached: 0 kB
Active: 24823868 kB
Inactive: 5258840 kB
Active(anon): 18146708 kB
Inactive(anon): 1672 kB
Active(file): 6677160 kB
Inactive(file): 5257168 kB
Unevictable: 0 kB
Mlocked: 0 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 360 kB
Writeback: 0 kB
AnonPages: 18146948 kB
Mapped: 596584 kB
Shmem: 1816 kB
Slab: 848520 kB
SReclaimable: 728940 kB
SUnreclaim: 119580 kB
KernelStack: 18592 kB
PageTables: 73392 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 16474020 kB
Committed_AS: 21293832 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 187420 kB
VmallocChunk: 34359449500 kB
HardwareCorrupted: 0 kB
AnonHugePages: 15556608 kB
HugePages_Total: 0
HugePages_Free: 0
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 2048 kB
DirectMap4k: 100224 kB
DirectMap2M: 5142528 kB
DirectMap1G: 30408704 kB
Userlevel 3
Badge +10
What version of AOS and hypervisor are you running?

Please run the following on the CVM and post results.

# cat /proc/meminfo
Badge +7
Memory utilization is still high even with 32 GB Memory
Userlevel 4
Badge +17
elarson have you try reboot the affected CVM and see the memory utilization ?
Userlevel 3
Badge +19
Hi,
There is no specific procedure unless u r running AHV, but I will recommend to go through CVM memory best practise based on AOS version and features enabled.

F.P
Userlevel 4
Badge +20
Yep, after you boot the CVM up, wait till cluster status returns to ok on ALL the nodes before moving on. If it's AHV, here's a post that goes through the commands.

https://www.linkedin.com/pulse/how-increase-controller-vm-memory-nutanix-community-edition-d-andrea

Or if you want to do it via script...

http://vcdx56.com/2017/02/change-ahv-based-nutanix-controller-virtual-machine-cvm-ram/

Reply