I just took a "mock exam" off of the VMware website and found the answer to the following question to be rather surprising.
The questions is:
Which of the following technologies would provide the best chance of uninterrupted service in the event of an ESX Server hardware failure?
The answers are:
VMware HA Clusters
VCB
DRS
Microsoft Cluster Services
According to VMware's website, the correct answer is:
Microsoft Cluster Services could provide truly uninterrupted service in the event of hardware failure.
I find this surprising because I would never have guessed a vendor would ask a question where the correct answer is a competitors product.
Jason
null
I suppose they're attempting to ascertain whether you know that HA will mean that the VM's are restarted. so failover will not be seamless
It's not a great question though, because the MS Cluster will also have to failover which would interrupt service albeit for less time than a VM restart.
What they are trying to explain is that when you use HA, the server has to be restarted in minutes rather than seconds.
With MSCS only the services have to be failed over meaning less interruption of service.
The caveat here is you can utilise both technologies using vmware either inside a box, across boxes or physical & virtual nodes.
Expense is the big factor here. An example in the real world could be physical active server nodes for SQL, Exchange, File, Print etc. with a single ESX server sitting with all the passive virtual nodes waiting in case one fails.
I suppose they're attempting to ascertain whether you
know that HA will mean that the VM's are restarted.
so failover will not be seamless
It's not a great question though, because the MS
Cluster will also have to failover which would
interrupt service albeit for less time than a VM
restart.
This is exactly what they are testing, if you know the difference between HA and MSCS. They should probably change the question to which would provide the least interruption in service.