Issue Description

The CPU and memory status of virtual machines on the master controller node is normal but displayed as 0% on all other nodes.

Error/Warning Information

Handling Process

1. Check that the network from the main control to other hosts is normal.
   • Check via vtpclustat, ping and chping.

Root Cause

Modifying the management port is prone to delay, causing the host_proxy process from the host to not automatically restart.

Solution

1. Restart the report service and redis service
   vs_cluster_cmd.sh e "/sf/etc/init.d/vtp-datareport-server restart"
   vs_cluster_cmd.sh e "/sf/etc/init.d/cacher restart"

2. Restart the cluster service (business will be interrupted)
   /sf/etc/init.d/pmxcfs restart && /sf/etc/init.d/corosync restart

3. Restart the host_proxy service of the host and return to normal
   vs_cluster_cmd.sh e /sf/etc/init.d/host_proxy restart

Restart the host_proxy service of the host and return to normal