Discussions specific to the VMware vSphere hypervisor
Post Reply
antran
Influencer
Posts: 22
Liked: never
Joined: Jun 05, 2018 2:36 am
Full Name: Tran Thanh Linh An
Contact:

Mix Deduplication Storage DataDomain with non-Deduplication Storage for Scale-out Repository

Post by antran » Nov 19, 2018 4:02 am

Now we have a scale-out repository with one DataDomain6300 30TB usable capacity
We want to expand it more 300TB usable capacity
Please susggest us if we can buy a non-deduplication storage like NAS, HPE MSA, Server Storage to mix with the current DataDomain to expand the scale-out capacity?
Or is it recommend us to buy ith the DataDomain expansion shelf to expand the scale-out capacity?

Thanks

falkob
Service Provider
Posts: 66
Liked: 20 times
Joined: Sep 28, 2017 7:47 am
Full Name: Falko Banaszak
Contact:

Re: Mix Deduplication Storage DataDomain with non-Deduplication Storage for Scale-out Repository

Post by falkob » Nov 19, 2018 8:57 am

Hi,
i did not found any Knowledge Base Article or "Prerequisite" or "Requirement" where Veeam says, you have to add "extents" to your SOBR from the same Storage/Vendor/Device.
Instead Veeam says, you can add extents from "NAS", "DAS" or "DEDUPE" devices, due to the functionality of the extents in the SOBR.
If you mix up extents from different storage systems, make sure to edit and adjust "Backup placement settings" as you need it, because the performance of a Data Domain system versus a e.g a NAS system will differ.

Maybe also have a look here: https://www.veeam.com/scale-out-backup- ... ry_wpp.pdf

Best Regards

Falko
VCP6.5-DCV, VCP6-DCV, VMCE, VMCA, VMware vExpert
https://www.virtualhome.blog

Andreas Neufert
Veeam Software
Posts: 3823
Liked: 687 times
Joined: May 04, 2011 8:36 am
Full Name: Andreas Neufert
Location: Germany
Contact:

Re: Mix Deduplication Storage DataDomain with non-Deduplication Storage for Scale-out Repository

Post by Andreas Neufert » Nov 19, 2018 9:26 am

In theory you can do this and it is supported.

Specifically if you use the setting to store the full backups on DD while storing the incrementals on the new standard storage. SOBR was specifially build for it.

If you want to use SOBR in a way that you mix storage types for the same workload, you need to live with potential performance differences and it will be hard to dedup if something do not work as expected.

As well keep in mind that for new backups (or new backup chains) the SOBR extend with the most free space is used. So it could happen that new backup files always go to the dedup device as it reports more free space than your newly added storage. If you want to use it in parallel for same workload profiles, I would use the standard storage as Repository for the more critical servers (performance) and use the DD for the other VMs. This need manual Job definition.

Post Reply

Who is online

Users browsing this forum: Baidu [Spider] and 7 guests