Veeam B&R 9.5 U4 - Azure Blob Tiering

[simonh-infopot]

Can 9.5u4 control the Blob Tiering (Hot/Cold/Archive) inside Azure Blobs?

it would be awesome if the Capacity Tier Offload process could upload to a "Cold" blob, and then set it to "Archive".

Thanks.

[Mike Resseler]

Hi Simonh-infopot

First: Welcome to the forums

At this point in time it is not possible. But this is certainly something that we think about for the future. I assume you do this for the cost but can you live with the slow restore possibilities? When moving to an archive tier, it can take up to 1 hours just to rehydrate those files before you can do an actual restore.

[simonh-infopot]

Hi Mike,

Thanks for letting me know.

Yes, I can live with the slow restore. It would be analogous to "get the tapes from offsite and start restoring archived data".

[wa15]

Just to add, this would be EXTREMELY useful for us as well. When talking about 7+ years of retention, several hours or even a half day of waiting for files to rehydrate wouldn't be an issue and is indeed analogous to "getting the tapes".

Side note, since Veeam doesn't support this, is there a way of doing this migration of cold to archive tier at the blob level? If so, will such a move "mess up" Veeam?

[Mike Resseler]

Hi Simonh-infopot and wa15,

In our next update archive tier will not be supported. At the moment the data is moved to archive tier we loose access to it. So as a "workaround" to migrate it to archive tier from the console or through PowerShell would break the functionality. As said, we are certainly thinking on how to approach this best for the future.

[wa15]

This is really disappointing. Veeam releases a new cloud "archive tier" with U4, but the back-end cloud storage type that is actually designed for archiving (e.g. Blob archive tier) is not supported? This defeats the purpose of having an archive tier in various cases. Even when looking at use cases for various blob tiers https://docs.microsoft.com/en-us/azure/ ... ccess-tier, MS notes that "archival datasets" are best fitted for use with the blob archive tier. The "Cool" access tier is best for short-term and infrequently used data.

There is a difference of several thousands of dollars per month for us between the archive tier & cold tier.

Does Cloud Connect support archive tier Blob storage? Wondering if we can just use that instead.

[Gostev]

Veeam releases a new cloud "archive tier" with U4, but the back-end cloud storage type that is actually designed for archiving (e.g. Blob archive tier) is not supported?

I am sorry you are disappointed, but please note that we didn't release "archive tier" as you say, but rather "capacity tier". This functionality seamlessly extends scale-out backup repository into object storage, with backup files remaining instantly accessible on-prem regardless of where their content is actually located. So if you know how blob Azure blob archive tier works, you will quickly realize it simply is not fit for the job.

If we are to support archive tier at some point in the future, then it will be quite different integration - something close to true archival, and you would only use it for data you should "never" need to read back - THIS is when you will see us talking about releasing "archive tier"

On the other hand, the definition of cool tier that you shared above actually hits nail in the head with its reference to "infrequently used data", making it the perfect target for our current integration: that is, for hosting older backup files which are still under the retention policy, just less likely to be the source of restores due to being outside of your typical operational restores window. But nevertheless, they are still required to be immediately accessible and restorable, which is why previously customers kept them on disk-based storage in the first place.

Having said that, I suppose you can use blob archive tier with Veeam via VTL (this was always possible and does not require Update 4).

Cloud Connect is just a multi-tenancy and secure transport framework on top of our product, so it is working with the same repositories.

[jb81]

Hi, in the "what's new" of 9.5u4 there is:

*****************
Space efficiency: In addition to source-side compression, backup files are offloaded to object storage in the forever incremental manner, thus preventing duplication between multiple full backups, which effectively provides source-side deduplication.
*****************

Let's say I configure an azure blob that has a lifecycle management policy (hot tier first, then move to cool after 30days, then archive after 90days), what does happen to blocks moved to the object storage? I mean, after 90 days they are moved to archive, so not accessible anymore until someone rehydratates them back from azure management portal....

[veremin]

Azure Archive Blob Tier is not supported in this release. Thanks!

[jb81]

the same question for aws glacier

[Gostev]

Same answer here as well: it's simply not compatible with the Capacity Tier functionality due to huge data retrieval delay.

[evgenyrudinsky]

Since there are few tiers in Azure Storage (GPv2 / blob) and there is lifecycle management for blobs, can you guys confirm / disagree that the following might work, - https://azure.microsoft.com/en-us/blog/ ... available/

Since there is daysAfterModificationGreaterThan attribute that can be used for tier migration. What are your thoughts on this? Is it go or no go solution?

Thanks.

[Gostev]

No, this is not supported. No other processes should manage data residing in Veeam containers.