-
- Novice
- Posts: 3
- Liked: never
- Joined: Oct 26, 2022 9:01 pm
- Contact:
How much data is needed to restore from Cloud Object Storage?
I am currently setting up a SOBR config, with capacity tier being offloaded to Wasabi. What I'm trying to figure out is how much data will need to be pulled down from the cloud in order to perform restores. Say, for example, I want to restore a single file that currently resides out in cloud storage. Will Veeam need to download the full VBK and VBI file-set, in order to perform that restore? Or, does it only need to pull down a few blocks? Is there any documentation that spells out how this works?
I'm trying to get this sorted so that I can determine the best way to structure my jobs. It wouldn't make sense for me to have multiple VMs in a job, if they entire job-set needs to be pulled down to restore a few files. But, I also don't want to lose Veeam's de-duplication abilities, but setting up many smaller jobs. Depending upon how cloud restores work, it may make sense to use the "per-machine backup files" option.
To be fair, I do have a support ticket open about this, but they're giving me some info that seems to conflict with other things I've read. They're saying that using the "per-machine backup files" doesn't affect de-duplication ability across a backup job, and that when restoring from the cloud, all backup files need to be copied locally before you can restore. So, I'm just looking for confirmation one way or the other.
I'm trying to get this sorted so that I can determine the best way to structure my jobs. It wouldn't make sense for me to have multiple VMs in a job, if they entire job-set needs to be pulled down to restore a few files. But, I also don't want to lose Veeam's de-duplication abilities, but setting up many smaller jobs. Depending upon how cloud restores work, it may make sense to use the "per-machine backup files" option.
To be fair, I do have a support ticket open about this, but they're giving me some info that seems to conflict with other things I've read. They're saying that using the "per-machine backup files" doesn't affect de-duplication ability across a backup job, and that when restoring from the cloud, all backup files need to be copied locally before you can restore. So, I'm just looking for confirmation one way or the other.
-
- Veeam Software
- Posts: 2590
- Liked: 606 times
- Joined: Jun 28, 2016 12:12 pm
- Contact:
Re: How much data is needed to restore from Cloud Object Storage?
Hi SirVsALot,
1. "Say, for example, I want to restore a single file that currently resides out in cloud storage. Will Veeam need to download the full VBK and VBI file-set, in order to perform that restore? Or, does it only need to pull down a few blocks? Is there any documentation that spells out how this works?"
Veeam will only retrieve the data blocks necessary (+ a little management traffic naturally). The Structure of Capacity Tier (and Object Storage repositories in general) means that Veeam is able to retrieve data at the block level resulting in "only necessary blocks", so no need to repatriate the entire backup to perform restores of data from within the GuestOS.
Can you share the case number where you asked this question? I'd like to review what was told, as I suspect there may just be a bit of semantic confusion happening. The latter statement does not sound correct as there's no need to download the entire job for a single file or VM restore, so likely there is just some confusion happening.
1. "Say, for example, I want to restore a single file that currently resides out in cloud storage. Will Veeam need to download the full VBK and VBI file-set, in order to perform that restore? Or, does it only need to pull down a few blocks? Is there any documentation that spells out how this works?"
Veeam will only retrieve the data blocks necessary (+ a little management traffic naturally). The Structure of Capacity Tier (and Object Storage repositories in general) means that Veeam is able to retrieve data at the block level resulting in "only necessary blocks", so no need to repatriate the entire backup to perform restores of data from within the GuestOS.
Can you share the case number where you asked this question? I'd like to review what was told, as I suspect there may just be a bit of semantic confusion happening. The latter statement does not sound correct as there's no need to download the entire job for a single file or VM restore, so likely there is just some confusion happening.
David Domask | Product Management: Principal Analyst
-
- Novice
- Posts: 3
- Liked: never
- Joined: Oct 26, 2022 9:01 pm
- Contact:
Re: How much data is needed to restore from Cloud Object Storage?
A review of the case would be great! I feel like I'm being fed a lot of incorrect and/or contradictory information, but maybe I'm just phrasing my questions wrong. Case# 07387682.
This is what they had stated:
This is what they had stated:
Veeam Backup & Replication requires the necessary backup files (VBK for full backups and VIB for incremental backups) to be available locally before starting the restoration process. This means that for any restoration, whether it's a single file or an entire VM, the relevant VBK and VIB files must be downloaded from cloud storage.
-
- Veeam Software
- Posts: 2590
- Liked: 606 times
- Joined: Jun 28, 2016 12:12 pm
- Contact:
Re: How much data is needed to restore from Cloud Object Storage?
Thanks for sharing the case SirVsALot, indeed, I do not believe that statement (in this context) is incorrect. I will be forwarding the case to Support Management for review, perhaps there is just a misunderstanding from the engineer who made this statement or some bad info, but it's best to get corrected.
David Domask | Product Management: Principal Analyst
-
- Novice
- Posts: 3
- Liked: never
- Joined: Oct 26, 2022 9:01 pm
- Contact:
Re: How much data is needed to restore from Cloud Object Storage?
In case anyone comes across this post in the future, here was support's response once they had a bit more clarity around the subject:
Considering initial scenario, the restore procedure might vary slightly if we work with an Object Repository or a Scale Out Backup Repository. In the second option we will also have a local Repo as part of the setup; in the 1st one, we will be working directly with the bucket. In both scenarios, Veeam will download needed Backup Files and the size will vary depending on the Restore Point we select. The procedure is exactly the same as if we were working with local files, we just need to add the download time into consideration. For SOBR, we have documentation that addresses the shared blocks, but it is only applicable for SOBR when we also have data locally: https://helpcenter.veeam.com/docs/backu ... ml?ver=120
Who is online
Users browsing this forum: No registered users and 10 guests