Is vSphere Data Protection (VDP) Replication Overtime Available?

[jrussellaz]

I realize its possible to allow overtime for a VDP Backup Task: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=210966...

Is the same possible for a replication task? My initial seed replication across a slow link is failing after 24 hours.

[RyanJMN]

Replication functionality is not in MCCLI so I don't believe there is a way like there is with a client backup.  You should be seeing a partial replication which means its saving the deduplicated data.  When it starts over it doesn't have to copy the backup data a 2nd time.  For any backup data that is already at the replication target it only has to copy meta data which references the backup data.  You should see the total amount of data replicated increasing.  If you aren't then you don't have enough bandwidth.

Possible solution is to manually copy the VMs being backed up to the replication target via USB or NAS device and back them up once using replication target VDP.  This will place the majority of the backup data at the replication target making the seeding process a lot faster.  Most of the data will deduplicate when its replicated.

[jrussellaz]

@RyanJMN,

Thanks for your reply. I'd read in other places that replication should pick up where it left off. Thanks for confirming. Regarding this statement:

You should see the total amount of data replicated increasing.

Would I monitor this by watching the *0-1008-Replicate-avtar.log log files? If this is correct then I believe the replication is making progress. Here are the first entries from two logs for the same job:

2015-12-06 09:06:36 avtar Info : Status 2015-12-06 09:06:36, 441.0 GB (140.7 MB, 0.03% new) 284MB   3% CPU  XXXXXXX#10::/VMFiles/2/

2015-12-10 09:07:06 avtar Info : Status 2015-12-10 09:07:06, 570.6 GB (114.7 MB, 0.02% new) 280MB   3% CPU  XXXXXXX#10::/VMFiles/2/

I think this means on 12/6 it started at 441.0 GB and on 12/10 it started at 570.6 GB. If this is all correct, how can I tell how much data is remaining to be copied?

Thanks in advance for your help.

[RyanJMN · Accepted solution]

I recommend using mccli instead of looking directly at the log files.  You see the same info but is easier and quicker to do.  The commands I would run is to run is mccli activity show and mccli activity get-log --id=.  VDP doesn't track remaining data.  Oftentimes its impossible because you can't tell compression rates or total data being backed up until the job completes.

If your replicating everything from one VDP appliance to another you can look at the total disk being used be each to get an approximate amount of post-compression data is left to be replicated.