Basic Veeam replication questions

[omg]

Hi All,

These are some basic replication questions from a newbie. Please answer them
quickly if you can rather than giving a 200 pages web link

1) When we setup a replication of a VMWare virtual machine on a different target host than the primary site. When something
goes wrong on the primary and we do a failover to the replica VM. Does the replica site VM will have data as on what time
of the primary site VM ?

2) When the replica site is up and end users are using the replica site VM and making changes to the data. And let says
replica site is used for 1 day. In the meantime we are fixing the primary site VM to see why it went down. Let say we now
fixed the primary site. When we failover back to production site, will the production site automatically gets all the data
changes which have been done by end users on the replica site automatically ?

3) Does Veeam store all the 1 day changes to the replica site in some kind of repository, if yes where ?


Thanks in advance

[tdewin]

1) When we setup a replication of a VMWare virtual machine on a different target host than the primary site. When something
goes wrong on the primary and we do a failover to the replica VM. Does the replica site VM will have data as on what time
of the primary site VM ?

RPO depends on how often the scheduler runs. If you configure the replication on a daily basis, you might loose up till 24h of data. If you schedule to replicate every 1 hour, and the bandwidth can handle the changes, you will only loose 1 hour of data maximum. This is a maximum because imagine the replication just completed five minutes before failure, in this case you would only loose 5 minutes of data.

2) When the replica site is up and end users are using the replica site VM and making changes to the data. And let says
replica site is used for 1 day. In the meantime we are fixing the primary site VM to see why it went down. Let say we now
fixed the primary site. When we failover back to production site, will the production site automatically gets all the data
changes which have been done by end users on the replica site automatically ?

3) Does Veeam store all the 1 day changes to the replica site in some kind of repository, if yes where ?

As long as you don't do a permanent failover, all change will be kept in a snapshot, allowing you to failback if you used the Veeam GUI to do the failover. If you commited a permanent failover or manually started the VMs and deleted the snapshots at the target side, you can still manual simulate a failback by creating a replica job in the other direction. It is still possible to do an incremental sync by using the replica mapping function

[Shestakov]

Hi,
My short answers with some concise additional reading

1) When the synchronization was done last time, according to the job schedule. You can fail over to the latest state of a replica or to any of its good known restore points.

2) Yes, but not automatically, you need to commit failback.
When you commit failback, you confirm that you want to get back to the original VM. Veeam Backup & Replication gets back to the normal operation mode and resumes replication activities for the original VM to which you failed back.

3)It is kept in a protective snapshot.
Here is described how to commit failback in details.

Note that critical changes made on a production site are to be replicated to the VM_replica on the DR site, so it`s recommended to make backups periodically for the best data protection.

[omg]

As long as you don't do a permanent failover, all change will be kept in a snapshot,

(1) Which snapshot is it ? VMWare VM snaphosts (or) Veeam snapshots ?

[Shestakov]

Veeam B&R initiates creation of VMware snapshot.

[omg]

This might be a dumb question:

We have ESXi host with 10 VMs.
If we do NOT use replication strategy at all and only depend on regular backups.

In case of a disaster if it takes us around 3 hours to full restore all the VMs using Veem B&R to another ESXi host.
We are OK with 3 hours being down during restoration. We are not so business critical that we cannot afford 3 hours
downtime during disaster.

In the above scenario are we OK to just live with regular backup strategy instead of doing all the replication setup ?

Is the replication strategy only allows us to save the 3 hours of restoration time during disaster ? or is it more than that.

Any inputs are appreciated.

[Shestakov]

In the above scenario are we OK to just live with regular backup strategy instead of doing all the replication setup ?

Absolutely.