Redundancy testing

[tractng]

Hello,

vCenter/ESXi = 7u3

We have setup an environment with 5 hosts connecting to a SAN (NFS & iSCSI datastores).  DRS is turned on (partially automated).  We have disconnected one of the NICs (etherchannel) during the testing and everything seems to be working.

What else should I be testing?  Would forcefully powering (cold boot) one of the hosts down be another test?

Thanks,

TT

[scott28tt]

I used to use the iLO/DRAC and power the server off.

[IRIX201110141]

Part of my deployments are always a vSphere HA test of a failing host.

• Disable DRS
• Move all VMs away and leave one test VM behind which you can ping
• If its FT setup place a primary VM on and ping it
• Configure mail notification for all related vSphere HA alarms
• Use ILO/iDRAC to shutdown the Host or pull the cable

Expected behavior

• Ha should restart the VM within 1min and ping should return
• FT should protect the VM and there was no interrupt within the ping
• You will get some Mails

Because of we are not using HA for VM

• A long time ago there was a registry key to force a BSOD based an a key command

They are more tests

• vMotion
• svMotion
• Pulling networking cabled

Regards,
Joerg