I would like to list the best practices, i came across Oracle articles and VMWare white papers.
- All ESX server host hardware, in particular the CPU, must be compatible
- All virtual switches must be configured the same way for all participating ESX server hosts
- Use a separate non-routable subnet for all Vmotion traffic or dedicated NICs for service console and Vmotion
- Run a private Gigabit Ethernet migration network between all Vmotion enabled managed hosts
- Incase a LUN has to be shared between multiple VMs, set “DisallowSnapshotLUN” value to 0 in virtual center
- There needs to be free memory greater than the VM NVRAM file size
- Make sure that Vmotion has unique IP and there aren’t duplicate Vmotion IPs
- Make sure that Vmotion vmkernel portgroup is on its own vswitch and has a unique VLAN ID if VLANs are being used
- VMware Network Documents also state that Vmotion and iSCSI vmkernel interfaces need to be on isolated networks vmkernel PGs in separate vswitches, and on separate networks or VLANs
- Make sure that Vmotion is NOT enabled on any other portgroups other than vmkernel interface intended for Vmotion
- Check time settings, enable NTP for the ESX/ESXi Servers in the clusters
- VMware suggests where possible to not route Vmotion traffic to limit the number of hops that Vmotion has to take, and only one vmkernel interface is permitted per vswitch
- Every hop that Vmotion traverses adds to Vmotion latency. This is not always possible but it is a VMware suggestion
- Follow the troubleshooting guidelines from ref. notes below:
- Troubleshooting Vmotion Failures and Vmotion Configuration Best Practices (Doc ID 1518833.1)