Discussions related to using object storage as a backup target.
Post Reply
scott.hale
Influencer
Posts: 11
Liked: 1 time
Joined: Jan 19, 2011 12:04 pm
Contact:

What type of job to select for SOBR to Cloud object storage ?

Post by scott.hale »

Hi, I am looking for feedback on the optimal type of backup job to use to minimize/optimize bandwidth consumption when moving data to cloud object storage given this general scenario:

Veeam V.21.1.X Enterprise.
1Gbps Internet connection.
7Tb VM with about 10Gb daily change (Mainly unstructured data, i.e.: PDF's, XLSX and DOCX files).
Retention-recovery requirements: Any day within 60 days | Month-end full for 12 months | Year-end full for 7 years.
SOBR with onsite performance tier.
SOBR with B2 cloud object storage capacity tier.
Eventually SOBR with B2 archive tier.

Note: I will send in a copy of the full VM as a seed to B2.

Would appreciate any feedback. Many thanks !
david.domask
Veeam Software
Posts: 2629
Liked: 611 times
Joined: Jun 28, 2016 12:12 pm
Contact:

Re: What type of job to select for SOBR to Cloud object storage ?

Post by david.domask » 1 person likes this post

Hi scott.hale,

The job type itself doesn't really matter -- with Scale-out Backup Repostiory ( SOBR ), you'll be either moving or copying data to Capacity Tier, so the original job type doesn't particularly matter, it will use the same mechanism.

You also can do Direct to Object Storage by just adding an Object Storage repository and use the jobs listed on that User Guide page for directly writing to B2.

Basically, I don't think really anything to worry about here, Veeam already takes steps to ensure that the necessary data for complete backups is transferred.

As for Seeding, last I remembered, Backblaze's Fireball program won't work as with Object Storage the structure of the repository/backups is a bit different. That is, the SOBR offload/Direct to Object Storage isn't just copying the VBK/VIB files to S3, the data is broken up into chunks as shown in the User Guide page on the Object Storage Structure. So just copying a VBK file will not work. Once the initial offload is complete however, then the incremental changes should be pretty fine.

Similarly, consider Unstructured Data Backup from the server -- you can keep a local copy and have Secondary and/or Archive copies that can have further customized file version retention. This may be a consideration if the main concern is protecting the file server and getting the data to your S3 provider in a timely manner.
David Domask | Product Management: Principal Analyst
scott.hale
Influencer
Posts: 11
Liked: 1 time
Joined: Jan 19, 2011 12:04 pm
Contact:

Re: What type of job to select for SOBR to Cloud object storage ?

Post by scott.hale » 1 person likes this post

David, many thanks for your reply, and particularly the insight into seeding object storage and the shortcoming with B2 ! Cheers, Scott.
Post Reply

Who is online

Users browsing this forum: No registered users and 16 guests