vSphere 5.1 HA / FT technical difference

[fasttrack]

Hello guys,

I am new in 5.1 HA/FT and I am trying to understand more about this technologies.

Neverthless I have surfed the internet and VmWare site for more details, I still need more information and confused.

1)

Tecnically speaking (not logical: HA: rapid recovery, FT: continuous availability), what this the difference between HA and FT?

I have understood that for both the solution all VM in a host are replicated in a second host.

HA: if VM or host fail, the VM is restarted in the other host  (it means that duing operation the <slave> VM is off?)

FT: here there is something more, communication between same VM hosted in different hosted. So,

a) it means that in this case all VM are up for all Hosts?

b) When failure occur: the VM (secondary) in the other  host is activate: So it means that teh secondary VM is always online?

2)

What's more, when looking some pictures that describe how they work, I do not understand why they are different.

3)

In case of using SQL Server 2008, I think that is possible to adopt vSphere HA/FT as clustering method (therefore using VM) instead of using machine level failover clustering. Is it true?

4)

Do you have any picture that represent the HA/FT architecture with some applications and SQL Server.

Many thanks for your support.

[sparrowangelste]

FT is more ideal for a solution that doesnt inherently have a clustering feature, so SQL woulsnt be something normally used with FT.

HA: high availibilty gives you high uptime, allowing for the monitoring of the host, vm, or application (if vendor supported) and allowing for the HA agent to restart a vm if there is a failure.

FT create a sync copy of a vm on 2 different hosts, allowing for max uptime, but there are limitations such as 1 cpu only, and limited ram in coparison with a non FT vm.

in real world applications I rarely hear of anyone using it.

[ramkrishna1]

Hi

Welcome to the communities.

HA in single line I will say consist of a set of loosely coupled servers which have fail over capabilities on which services will not remain 100 % available it may be 99%.

And FT is two separate machines that are mirrored.  In the event that the main system has a hardware failure, the secondary system takes over and there is zero downtime.

"a journey of a thousand miles starts  with a single step."

[fasttrack]

Many thanks for you reply:

So to sumup:

1) In case of SQL Server High Availability: using VSpere 5.1 HA/FT is not the right solutions? It means to use the Cluster feature provided by Microsoft ?

2) When you write.....<in real world applications I rarely hear of anyone using it> do you mean for both the technologies: HA and FT?

3) You say that FT do not support more than  1 vCPU. But if I am not wrong, I have read that 5.1 support more than 1 vCPU. Correct?

4) .... Loosely coupled servers...... Please could you detailed?

Many thanks

[sparrowangelste · Accepted solution]

fasttrack wrote:

Many thanks for you reply:

So to sumup:

1) In case of SQL Server High Availability: using VSpere 5.1 HA/FT is not the right solutions? It means to use the Cluster feature provided by Microsoft ?

use the MS cluster

2) When you write.....<in real world applications I rarely hear of anyone using it> do you mean for both the technologies: HA and FT?

everyone uses HA

no one uses FT (almost)

3) You say that FT do not support more than  1 vCPU. But if I am not wrong, I have read that 5.1 support more than 1 vCPU. Correct?

http://www.vmware.com/pdf/vsphere5/r51/vsphere-51-configuration-maximums.pdf

still 1 cpu per FT vm

4) .... Loosely coupled servers...... Please could you detailed?

Many thanks