VCSA: High Availability

[Dryv]

Hi Community

Does anyone have any experience of how VCSA copes with a hard shutdown for e.g. lets say VCSA is hosted on a host in a cluster that abruptly fails and vSphere HA kicks in to restart it on other host in the cluster. Is it likely to recover consistently?

Is there anything I can do to ensure the VCSA doesn't get corrupted in such scenarios outside of having a clone taken regularly in case I need to revert to it?

Can VCSA be deployed in a manner similar to application clustering?

Thanks

Dryv

[rcporto]

Since this is a hard shutdown, HA cannot guarantee the consistency... BUT, depending of the size of your vCSA and version of vSphere you're running, you can use the Fault Tolerance (FT) to protect the vCSA virtual machine.

[Dryv]

Hi Richardson

I'm running 5.5... which I think is limited to 1 vcpu.

Is it a good idea to run this appliance with 1 vcpu? Environment is no bigger than 30 hosts and 350 VMs... I can't seem to find a sizing guide and have just deployed it out of the box.

I would hope vmware have tested this extensively as no matter how well an environment is built host failures do happen and alot of shops use HA (rightly or wrongly ) to try and recover from such crashes.

[rcporto · Accepted solution]

Since you're at version 5.5, the recommendation is upgrade to version 6 to take full advantage of the new FT that includes supports to until 4 vCPU.

[Dryv]

Thanks Richardson, this might just give me the ammunition I've been needing to upgrade to ESXi 6 and vCenter 6