Discussions specific to the VMware vSphere hypervisor
CaptainFred
Enthusiast
Posts: 88
Liked: 2 times
Joined: Jul 31, 2013 12:05 pm
Full Name: Si
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by CaptainFred » Oct 14, 2013 3:35 pm

Ok makes sense, thanks.

Would it have to do "digest recalculation (metadata rebuild)" for each VM or at the start of the whole replication job?

By retention policy, do you mean the "restore points to keep" in the replication job settings? (under the "replica name suffix" setting). We have set it to 9, planning on running replication 3 times a day hopefully, so I guess it might get quite large!

CaptainFred
Enthusiast
Posts: 88
Liked: 2 times
Joined: Jul 31, 2013 12:05 pm
Full Name: Si
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by CaptainFred » Oct 14, 2013 3:38 pm

Btw, do you have a place/best way to report minor issues, eg spelling mistakes? Under edit replication job, then "Virtual Machines" on the right hand side, the text at the top has the word "exlude" in it.

Vitaliy S.
Product Manager
Posts: 22697
Liked: 1498 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by Vitaliy S. » Oct 14, 2013 3:42 pm

1. If you lose the entire repository with metadata, it will be rebuild for each VM at the start of each VM task session.
2. It shouldn't be that large. VM metadata will be removed from the repository together with the restore point that falls out of the configured retention policy.

Vitaliy S.
Product Manager
Posts: 22697
Liked: 1498 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by Vitaliy S. » Oct 14, 2013 3:42 pm

CaptainFred wrote:Btw, do you have a place/best way to report minor issues, eg spelling mistakes? Under edit replication job, then "Virtual Machines" on the right hand side, the text at the top has the word "exlude" in it.
You can do that via forums or through our support team. Thanks for the heads up!

luckyinfil
Enthusiast
Posts: 91
Liked: 10 times
Joined: Aug 30, 2013 8:25 pm
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by luckyinfil » Oct 16, 2013 9:29 pm

I've been following this thread and I'm in a similar boat. I do have a few questions tho:

1) Assuming I am only backing up to local repository right now and I want to add in remote replication and backing up the current backups to a remote repository. Am I correct to assume that I will only need a replication job and a backup copy job (assuming I have the remote vSphere environment and the remote backup repositories)?

2) In the scenario above, is the WAN accelerator only for the backup copy jobs or is there a way to use them for replication as well? Also, for the "Repository for replica metadata" setting within the Replica Job, what does this mean? Is it supposed to be closer to the source or target? Also what is the point of the replica restore points if you are keeping backup restore points already?

3) Where does GFS fit into all this. Assuming my backup (both local and remote) repositories use windows deduplication, the GFS scheme sounds like a bad idea as it is constantly creating new VBK files.

Vitaliy S.
Product Manager
Posts: 22697
Liked: 1498 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Backup and Replication to DR Site Design Proposal

Post by Vitaliy S. » Oct 17, 2013 7:05 am

Hello John,

1. Correct.

2. Currently WAN accelerators are available for backup copy jobs only, but we are planning to expand this functionality to replication jobs as well in the next releases. Right now, make sure you have two proxy servers to optimize VM data traffic across the sites.

Replica metadata have to be as close to source site as possible, this will allow you to have better performance during digest calculation operation.

The point of having multiple replica restore points is pretty simple, see this topic for further reading.

3. GFS can only be used by backup copy jobs. You should go with this configuration if you need to store backups for long term (for example, for 1 year). As to deduplication and re-creating VBK files, then backup copy jobs are forward incremental, they only modify VBK files to create new restore point (full backup) according to your GFS settings. You will still benefit from storing your backup copy files on the dedupe volume.

Hope this helps!

Post Reply

Who is online

Users browsing this forum: Bing [Bot] and 46 guests