"Stats primary election" failure or unable to manage Stats DB object in RVC
search cancel

"Stats primary election" failure or unable to manage Stats DB object in RVC

book

Article ID: 408360

calendar_today

Updated On:

Products

VMware vCenter Server VMware vSAN

Issue/Introduction

Symptoms :

  • When attempting to manage or troubleshoot the vSAN Performance Service, you encounter the following symptoms:
  • The vSAN Skyline Health Service displays a Red error for "Stats primary election".
  • The .vsan.stats folder or vSAN performance statistics object shows a health status of Inaccessible or All Paths Down (APD).
  • Attempts to remove or recreate the statistics database object via the Ruby vSphere Console (RVC) using commands like vsan.perf.stats_object_delete or vsan.perf.cluster_object_create fail with a generic timeout or task
  • Can not remove or recreate stat object in RVC with out error

Environment

VMware vSAN 8.x

Cause

  • This issue occurs because the advanced configuration parameter /VSAN/DOMPauseAllCCPs is set to 1 (True) on one or more ESXi hosts in the cluster
  • DOMPauseAllCCPs is set 1 on one or more hosts.
  • Run the following to check
    esxcfg-advcfg -g /VSAN/DOMPauseAllCCPs     

 

Resolution

Step 1: Revert DOMPauseAllCCPs to Default

1. Log in to each ESXi host in the affected vSAN cluster via SSH using root credentials.

2. Run the following command to check the current value of the parameter:

esxcfg-advcfg -g /VSAN/DOMPauseAllCCPs

3. If the returned value is 1, reset it to the default value of 0 by running below command :

esxcfg-advcfg -s 0 /VSAN/DOMPauseAllCCPs

4. (Optional but Recommended) Verify that the accompanying parameter /VSAN/IgnoreClusterMemberListUpdates is also set to 0:

esxcfg-advcfg -g /VSAN/IgnoreClusterMemberListUpdates

If it returns 1, reset it using below command :

esxcfg-advcfg -s 0 /VSAN/IgnoreClusterMemberListUpdates

Step 2: Clear and Recreate the Performance Stats Object via RVC

1. Connect to your vCenter Server Appliance via SSH and log into the RVC console:

rvc [email protected]@localhost

2. Navigate to your vSAN cluster path within RVC:

cd /localhost/<datacenter_name>/computers/<cluster_name>

3. Forcefully delete the problematic or stale performance statistics object: 

vsan.perf.stats_object_delete .

4. Recreate a new, healthy statistics object and enable the Performance Service across the cluster: 

vsan.perf.cluster_object_create .

5. Return to the vSphere Client, navigate to Cluster > Monitor > vSAN > Skyline Health, and click Retest to confirm the alarm clears .

. remove or recreate stat object in RVC

Additional Information

Powered by Wolken Software