Skip to main content

There are 2 nutanix cluster in different regions. I'm curious about a replication issue. I currently have 19.77 TiB free (physical) 50.7 TiB and 9.74TiB free (logical) 25.35 TiB disk space on first nutanix cluster site . there is enough free space in the second zone. I saw it through Nutanix. now if I replication 2 servers ( both 500gb ) in other region. in the configuration, local snapshot and remote snapshot 1 selected. How much to lose from Nutanix storage. For example, in the nutanix cluster there is 9.7 tb of free space, if I replicate 10 servers . 500 gb each.  how much space will i lose on nutanix storage. Does this have an account? 

I want to learn this. I have 9.7 tb of free space, but I want to replicate 10-15 servers. Is there a problem in the storage area? How to calculate?

Hi @Dotex  In prism there is filled called Reclaimable Space.

 

Reclaimable Space. Displays the total space that can be recovered after the snapshot is deleted.

 

The reclaimable space is the amount of space reclaimed after each snapshot which mostly dependents on the amount of data deleted inside the VM and the dependent snapshot chain. (Considering no features like Compression, dedup, EC and snapshot clones)

 

I request you to follow the Data Protection Guide

https://portal.nutanix.com/page/documents/details?targetId=Prism-Element-Data-Protection-Guide-v5_17%3APrism-Element-Data-Protection-Guide-v5_17

 


Hi Dotex,

 

As far as I understand your question you are looking at understanding space requirements for recovery site? Is that correct?

Both data protection aka backups and replication are based on snapshots in a Nutanix environment. I have explained in the past that it is difficult to predict the size of the snapshot as it solely depends on the amount of change on the VM’s disks. Some applications and operations experience more change than the other.

With data protection aka backups the amount of storage required is generally higher than with the DR replication. This is because the oldest recovery point with backups is expected to be much further in the past. Hence more data is stored at any given point in time.

With data replication, however, the purpose is to recover data as close as possible to the point of impact. The guideline would be to have sufficient storage space required to accommodate all recovered VMs and to keep the snapshots and to have 15-20% free space.

For example, if all your protected VMs are 10x500GB - 5TB + snapshots (size of snapshots is a unique value to the environment and the protected VMs set + 15-20% free space. If to continue using the same example with a 9.7TB container at a DR site we have:

5TB of VMs
1.94TB of free space (9.7TB x 20%)
gives us 2.8TB of space for snapshots or 280GB per VM roughly.

Is that enough?
There is no straight answer, unfortunately. How high is the rate of change on the VMs, how many snapshots do you intend to keep at the recovery site? But also what is the disaster scenario that you are protecting your environment from?

For data replication general recommendation would be to have storage sufficient to recover from the disaster scenario. If you intend to recover only a few VMs at a time and have a few days worth of recovery points with snapshot one a day then there is no point of keeping equal size storage at a recovery site.

If the scenario is the simultaneous recovery of a number of VMs (essential services for example or even whole site failover) then the calculations above are applicable.

I understand that you are looking at a hard set answer but I’m afraid there isn’t one.

I would suggest starting with crucial VMs first. Let them remain replicated for a full cycle (i.e. let snapshots age) and protect more of second-grade critical VMs if you still have space.

Alternatively, reach out to one of Nutanix Software Engineers who would be more than willing to help you scope your DR or data protection site.

 

Let me know if there are still any questions.


Reply