Backup Copy with StoreOnce NAS Replication

[Edage]

Hi I have some question about Backup Copy. Is it work with StoreOnce Replication?

My customer want to use StoreOnce Replication to send secondary backup to another StoreOnce at another site. I can create Backup copy to NAS and configure NAS Replication with out problem but what concern me is when Backup Copy merge file Full Backup after it reach "Restore point to keep" from what I know if I configure it to be 7 day when it come to day 8 Full from Day 1 and Incremental from Day 2 will merge to be another Full Backup as Forever Incremental method.

At StoreOnce Local site it will never be problem but for StoreOnce at another site I don't sure that they will merge it too or not? Could you tell me what will happen?

[JasonL]

Hi Edage,

Generally speaking, when copying backups between StoreOnce's it is recommended to have Catalyst as the protocol being used to transfer data in conjunction with Veeam Backup Copy, meaning that when Veeam does a transfer of data it can leverage source-side deduplication. This also allows for processes such as the backup file merge to be virtualized by StoreOnce. Is your customer using Catalyst or is everything NAS shares only?

NAS shares do have the ability to replicate themselves and I have spoken to a couple of customers that have had success with this, but it's important to keep in mind that you don't benefit from any source side dedupe and it is basically a flat file copy that runs separate of Veeam Backup Copy. This could be challenging for a couple of reasons - it can take a long time depending on change rates, and Veeam has no knowledge of backup files being transferred unless you schedule manual rescans (or scripted) of the target StoreOnce repository.

[Andreas Neufert]

If you do it that way, please use a GFS chain and enable read the entire restore point from source as described here https://helpcenter.veeam.com/docs/backu ... tml?ver=95

That way you don‘t rehydrate deduped data when writing into store once.
As well you can spread the weekly GFS point creation from different jobs to different weekdays to reduce load.

[Edage]

- Dear JasonL

My customer using Catalyst to contain first backup from Backup job then using Backup copy to create secondary backup to NAS on same StoreOnce and using NAS replication to replicate it to target StoreOnce.

I think about using Backup Copy to send Backup file from Catalyst to target Catalyst but from link Bandwidth (around 10 - 20 mbps), Backup file size (File server Full Backup 700 - 800 GB and Incremental 200 GB after dedupe) and time for replication (around 14 - 15 hours) I think Backup Copy will not success, so I have to go with my customer suggest.

I also talk to my customer that he will need to rescans target StoreOnce Repository every time he want to restore but he said he ok with that.

But for my concern I don't sure that secondary Backup on target StoreOnce Repository can merge Full Backup (Restore Point 7 Days) on their own because I think it will merge only when Backup Copy job active and after it success create new Incremental. I think when it reach day 8 Sourece StoreOnce NAS will send New Backup set (1 Full + 6 Increment after merge) to target StoreOnce NAS with NAS Replication if it happen this way I think it will made some problem later.


- Dear Andreas

For what you recommend select "Read the entire restore point from source backup instead of synthesizing it from increments" it mean that Full backup (Monthly or Weekly) that Backup Copy create will read from Source backup file (From Backup job) right? I will consider to use it thank you.

[JasonL]

Edage,

Understood on why NAS replication is being used. I have definitely heard of some mixed results going this route, but if it's working for you customer, that's great. Keep in mind that if you do decide to try backup copy over Catalyst, you'll want to have the Veeam gateway server for StoreOnce on the source side, not at the same location as the target. This way Catalyst is being used to transfer data.

As for data being copied on the same StoreOnce at the local site, your concern is valid. Anytime you are using NAS shares in conjunction with backup copy or even regular backups with synthetic fulls, merge operations are a concern. What Andreas suggested would likely be useful here so it removes the merge operation when you get to day 8.

[salyerma]

I was wondering if there was anyone out there with any experience with this situation. We are using Veeam B&R 9.5 currently with DataDomains at two locations. Right now, we are letting the DataDomains replicate to each other over a dedicated 100mb connection. We are not using a BCJ to do the replication, we are using the built in DataDomain replication.

We are wanting to replace our DataDomains with new StoreOnce devices. My worry is that since Veeam does not support Catalyst Copy (yet), Will a Backup Copy Job in Veeam be as efficient as the datadomain replication and allow us to sync those.

I do not have the WAN acceleration piece, but will Veeam do Dedupe and compression over the Wan link? My hp rep and engineer are somewhat clueless about Veeam so they cannot help me figure out if replication will be fine.

Has someone on here went through this and has a general idea on how a BCJ will compare to built in DataDomain replication?

Thanks!

[foggy]

Hi Mark, please review the thread above for answers. Basically, backup copy job will perform data rehydration, but with a gateway server located on source and 'Gateway server and StoreOnce are connected over WAN' check box enabled in repository settings, Catalyst source-side deduplication will be used.

Here's another useful thread for review.

[salyerma]

I had read this thread before, but I was wondering if anyone had real world experience comparing Storage Replication vs BCJ setup with a gateway. Not hard numbers but is it slower (a lot?).