Implementing cloud combined storage tiers

Hey @dshep,

Beyond the docs, I’d also highly recommend taking a look at the cloud architecture guide for Azure - it explains a lot of concepts and considerations: https://documentation.commvault.com/commvault/v11_sp20/others/pdf/public-cloud-architecture-guide-for-microsoft-azure11-19.pdf

Using combined tiers is straightforward - when you create the cloud library you can choose the combined tier option - nothing more to configure beyond that. You don't need to set the storage class in azure, as Commvault can do that for you. Commvault will keep metadata on cool storage and move the actual data to the archive tier automatically. This brings several benefits when you need to recover from the archive copy since there is no extra delay waiting to move the metadata before we can read it to calculate what data actually needs to be brought back to facilitate the restore.

You can also use cloud power management to automatically turn on/off the cloud media agent as required. This can save a bundle of $$$ and also helps with security.

@Damian Andre I don’t think I’m following.   Do you mean when the cloud library is created it creates the storage account(s) and assigns the storage class(es)?

Hi @dshep 

Let me know if this helps clarify or if I’m not fully understanding your intentions:

To create a cloud library:​

From command center > storage > cloud 

Then follow the prompts for configuration of the cloud library

From the commcell console:

• When you right click on library > add > cloud storage library
• From here you will see the option to select the type of cloud storage, choose the primary media agent to use and enter your credentials for the cloud account.

• 

Here is some documentation for Combined Tier storage:

Planning Combined / Archive Storage Tiers
https://documentation.commvault.com/commvault/v11/article?p=124490.htm

How is Data Stored and Managed in the Various Microsoft Azure Storage Tiershttps://documentation.commvault.com/commvault/v11/article?p=95147.htm

Microsoft Azure Storage
https://documentation.commvault.com/commvault/v11/article?p=91299.htm

https://documentation.commvault.com/commvault/v11/article?p=87477.htm

Thanks @Sean Crifasi .   Here’s where I’m getting hung up.  When creating a cloud library/mount it requires authentication.  Normally I get the authentication key from a pre-created storage account.  If I pre-create the storage account I need to select a storage type (standard/premium) and the tier (hot/cold/archive).  I don’t see any way to create a storage account without selecting those, or to create the library without pre-creating the storage, and I’m not seeing any documentation that goes into step by step detail of creating a cloud library/mount that covers this specific information.  Is the commserve able to change the tier of the storage account when adding it to a library as a mount?

Hi @dshep,

Could you perhaps share me a screenshot of what you’re (for requiring the authentication) seeing just to make sure I fully understand we’re looking at the same thing?

Do you mean the request for the azure credentials? or do you mean commvault gives you a popup saying you need authorization for this action?

Regards,

Sean

@Sean Crifasi I mean the request for azure credentials in the cloud library configuration.

OK.  Now I get it.  Thanks!

@Damian Andre - I still have one more question on the storage account creation.

Its straight forward to create a storage account in Azure , which storage class to be defined during the storage account creation?

Or it does not matter which ever storage class the account is created in Azure “either hot or cool”. As we will be choosing the option “combined cool\archive tier” in commvault so it will manage and distribute the data as per the retention set.

I believe its the latter. When we write the objects we set the storage class on them - unlike, say AWS where typically the class is defined on the entire bucket. So whatever class you pick when you create the library is the one that will actually be used.

@Damian Andre , “So whatever class you pick when you create the library is the one that will actually be used.” - Based on this the regardless of the storage account being ‘hot or cold’ , as it will be refined as a combined tier , commvault will manage the data movement between the cold and archive tier.

Let me know if this is the right understanding.

Hello CV community!

The combined tearing is going to work also in existing Azure Cloud Disk Library?

Currently we have lot of TB is warm blob (o365 and VM backups).
Is this possible just to edit the current container and change the Storage Class from Hot → Hot/Archive?

Thank you in advance,
Nikos
 

@Nikos.Kyrm 

The existing data is not retroactively changed. Changing the storage tier would only effect new data.
https://documentation.commvault.com/11.24/expert/129205_migrating_data_between_storage_classes_in_cloud_storage_library.html

Please also ensure that if using combined storage tier your associated ddb’s are configured to automatically seal periodically

https://documentation.commvault.com/11.24/expert/139246_combined_storage_tier.html

https://documentation.commvault.com/11.24/expert/147071_migrating_cloud_storage_data.html

@Sean Crifasi  Thanks a lot for your quick reply!

From Commvault Public Cloud Architecture guide for Microsoft Azure (Feature release 11.25), Im reading that “Combined storage tiers are not recommended for primary backup copies”.

So, I guess that best practice is to create a new Azure Blob Cloud Library with combined Hot → Archive and send there (with Aux copies) monthly backups lets say.

To keep in Hot the last 15 days inc and all the synth fulls in combined Hot → Archive.

What’s your feedback about that?

Best regards,
Nikos