Tuesday, 13 June 2017

Nutanix Cluster is under-replicated

I've recently come across an error when trying to upgrade Nutanix Acropolis from version 5.0.1 to 5.0.2.

The error that was returned within Prism was - "Cluster is currently under-replicated".

This message occurs when the health-state of the Nutanix Cluster is on Level "warning" or "error" or the data resiliency state is not "Ok".

To resolve this issue follow the steps outlined below:-

#1 - Check if you can see under replication from curator command line.
  1. SSH to one of the CVM's
  2. Log in with the appropriate credentials (defaults are User name: nutanix Password: nutanix/4u)
  3. curator_cli get_under_replication_info summary=true


#2 - Find the ID for the last successful curator scan.
  1. curator_cli get_last_successful_scans 2>/dev/null


#3 - Check which CVM the curator master is running on.
  1. curator_cli get_master_Location


#4 - Log in to CVM that is the curator master and manually trigger a curator scan.
  1. SSH to the CVM that is running the curator master
  2. Log in with the appropriate credentials (defaults are User name: nutanix Password: nutanix/4u)
  3. allssh curl http://localhost:2010/master/api/client/StartCuratorTask?task_type=2
  4. Wait about 10 minutes, then check again if there is any under replicated data (curator_cli get_under_replication_info summary=true )


If all is well, you can then try the upgrade again.

No comments:

Post a Comment