HP Lefthand and VMware Fault Tolerance - production site failure

[atmorell]

Hello,

We are currently evaluating HP Lefthand with VMware Fault Tolerance for building a disaster proof setup.

I have worked with HP Lefthand before and really like the redundancy when choosing a Multi Site SAN layout with two sites. So far I have not used VMware FT for production.

What happens when the production site goes offline?

Will the VM's failover to the secondary site without any downtime or will they need a reboot?

The datastores should be protected by the Lefthand network RAID, and the VMs should failover to the secondary site with VMware FT.

However when I read page 13 of the document below it looks like blade 3 had to be rebooted when there was a complete site failure.

http://www.google.com/url?sa=t&source=web&cd=1&ved=0CBIQFjAA&url=http%3A%2F%2Fh20195.www2.hp.com%2Fv2%2FGetPDF.aspx%2F4AA0-4385ENW.pdf&rct=j&q=4aa0-4385enw&ei=Zl6gTOXVJYGtOJ7U6a0M&usg=AFQjCNFOHiNrmPPn00VStGLOybxm1McTNQ&sig2=U1uAvby4jDa-7pCF1lNoWw&cad=rja

"The third test consisted of powering off both the entire blade enclosure and the storage node on Site A. With the Primary VM running on Blade 1 and the Secondary VM running on Blade 3, HP abruptly powered off the entire Site A to simulate a total loss. Since the test configuration was running AD and DNS services on one of the blades in Site A, HP was creating a total site failure including the storage nodes.

The result when the entire site went down was that the primary VM running on Blade 1 was no longer available. The Secondary VM running on Blade 3 now became the Primary VM, and a new Secondary VM was spawned on Blade 4 on Site B. There was a delay on the VM as it needed to restart on Blade 3. Once the VM started on Blade 3, the VM kept running the processes.
"

Why did blade 3 have to be restarted? So was VMware FT not working?

Best regards.
Asbjørn Morell

[Michael Leu]

I have no clue about this stuff, you could watch http://www.youtube.com/watch?v=6n3VpfesM8A maybe it explains something.

[Marcel-p4500]

Question: what is/was your "host failure cluster tolerates" setting?