Solved

Prism VM status

  • 3 January 2018
  • 2 replies
  • 1061 views
J
Badge +7
Hi all,
I have a 13 node AHV cluster (5.1.3 / 20160925.90) where all my VM's (around 70) - status is randomly changing from On (which they are) to unknown. The VM's are running just fine but it seems Prism is unable to get their correct status. Also the CPU usage is constantly reporting exactly 50/100% when they are seen as on - regardless of their actual use - and 0% when off.

Have anyone seen this before, or have an idea what could cause it? NCC is not reporting anything of interest so I'm very puzzled.
icon

Best answer by Jumario 9 January 2018, 09:53

View original
This topic has been closed for comments

2 replies

C
Rank
Userlevel 3
Badge +13
Jumario Symptoms look like the acropolis service is crashing on the nodes and leadership is changing quickly. I would highly recommend creating a support case with Nutanix to further troubleshoot the issue.
J
Badge +7
This was apparently caused by a bug being triggered by having a cluster of mixed GPU and non-GPU nodes.
Terms & Conditions