This post is more than 5 years old
2 Posts
0
5343
Data Domain Collection Replication Configuration
Hi guys,
My environment consists of one Backup Exec server and two Data Domain Appliances (one at Primary replicating to the one at Secondary).
I am trying to configure Collection Replication but it seems collection replication has to disable the file system at source. Is this mandatory every time the system is replicating?
If so, is there a way to avoid this? Because the backup jobs at Primary will stop writing to the source appliance every time the system is replicating? Is there any workaround to this problem?
Regards,
Ian.
umichklewis
1.2K Posts
0
December 4th, 2015 17:00
No - there's no way to avoid suspending the source filesystem to start a collection replication. However, once the collection replication is in progress, there's no further interruption of service to the source filesystem. So, as your new backups land on the source filesystem, it is replicated to the destination Data Domain without any interruption to the source filesystem.
There are other replication types available. Why not use mtree or directory replication instead?
Let us know if that helps!
Karl
jbrooksuk
208 Posts
1
December 9th, 2015 07:00
Just to double check - because I'm not sure what you need to achieve.
You are aware that a collection replication DD target needs to be entirely empty and with a brand new filesystem that was never enabled (this is to ensure it is 100% empty - with no 'other' data stored).
It will then be entirely and only used as a DR target system for the source system, so that in a DR you can stand it up - like it is the source DD.
It cannot accept any other traffic - no backups, no Mtree replication etc... It is solely used as a DR destination for your source DD until such time that you need it.
As stated, up front you only need to disable the source FS once when you start the collection replication - when fully in sync - you are DR ready.
regards, Jonathan
ianonyango
2 Posts
0
December 16th, 2015 01:00
Thanks, I decided to go the mtree replication route instead.