Replication. Bypass the 2TB vmfs partition limit.

[mtcitalia]

Hi to all,

I need to know which are the best practices for configure a replication job when the total size of the VMs are bigger than the 2TB limit of vmfs file system.

For the backup job I found a work around by made a striped raid in to a VM that have a Veeam backup soft. installed, using two virtual disks that are located in two different vmfs datastores.

But I didn't follow the same way with the replication job, because the location required must be an vmfs file system.

Do you have any suggestions for me or the only way that I have, is to make a multiple replication jobs?

Thank You.

P.S.
Sorry for my poor English.

Bye

Stefano

[Vitaliy S.]

Hello Stefano,

Could you please describe your setup further? The point is that replication job does not form/create a single storage file (VBK) for the VMs included to the job as opposed to the backup job. This means that your source VMs and replicas will be equally the same.

Thanks.

[mtcitalia]

Hello Vitaly,
thank you for your answer. I want to describe you how is my infrastructure:

In the vCenter I have three esxi host. Two are in a cluster and they contain all the VMs machine. All the VMs are stored in three datastore that are located in to a SAN. The size of the SAN is about 4 TB. This is the production environment.
The third esxi host is placed in an other room and is where is located the VM that have installed the Veeam Backup soft. The datastore dedicated for the replication is inside the host itself and the size is about 4TB. This datastore is divided in two vmfs partitions (for the 2TB vmfs limit).
So my target is that if I have a down in main infrastructure ( where is the SAN), I'll can power on the VMs from a replication files in the other datastore, inside the third esxi host. So I need to create a replication job that copies the VMs from the SAN, but the total size of the all VMs is more big than 2TB. So, I want to know if I must split the VMs in two replication jobs; one for one vmfs partition and another job for the second vmfs partition, or there is another solution. Two or more replication jobs involve that I should be careful about the dimension of the replication files and that the replication jobs don't overlap each other.

I hope that I explain more clear which is my problem.

Thank You again

Bye

Stefano.

[Vitaliy S.]

Stefano, thanks for clarifications, now I got it. In order to exceed the LUN maximum size, you have to use extents or create two separate replication jobs.

But please be aware that it is not considered best practice to use extents for your VMFS partitions. Because if you lose a LUN from your extent, then you'll lose the entire VMFS partition hosting your VMs. And If you have two separate LUNs, then you will only lose VMs on that particular failed LUN.