Azure-docs: When is failover going GA?

[GFoley83]

This feature has been in preview for nearly a year now. There must surely be an ETA for when it's going GA by now?

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: d8314bbb-3e1b-7a68-dcd1-f7dfe34221a2
• Version Independent ID: 5c26cb3f-ec5c-b028-f456-53998a67749e
• Content: Disaster recovery and storage account failover (preview) - Azure Storage
• Content Source: articles/storage/common/storage-disaster-recovery-guidance.md
• Service: storage
• Sub-service: common
• GitHub Login: @tamram
• Microsoft Alias: tamram

[artemuwka]

Hi, Gavin! Thanks for checking in! The account failover capability is planned to go into GA by end of CY19. With the holidays coming up I don't have a firm date just yet and will know more in the next ~2 weeks. Would you mind sharing your scenario(s)/requirement(s)? An important call out is the current version of failover was designed for scenarios where a disaster is in progress or is coming and therefore it’s disruptive – upon failover completion the account becomes LRS. If you performed the failover as part of the disaster recovery drill you’ll need to convert it back to GRS/RA-GRS to be able to failover again (failback) to get to the state you were in before with respect to primary location. You’ll also need to convert the account to GRS/RA-GRS again upon failback completion to get to the state you were in before with respect to redundancy configuration. Converting from LRS to GRS/RA-GRS incurs cost. Finally, the state of the data can vary based on whether you lost any during the original failover and/or the subsequent failback. Potential data loss can occur depending on the usage pattern, e.g. you wrote blob A at 10.00 am which by 10.05 got replicated to the secondary location, at 10.10 you executed a failover – in this scenario there is no data loss associated with the failover however, if you executed a failover at 10.03 blob A didn’t get to secondary yet so it will be lost upon failover completion.

[GFoley83]

Hi Art. Thanks for the quick reply!
The scenario is nothing more complicated that having the ability for a customer-initiated storage account failover, so that "writes" can happen ASAP, without needing to update a connection string in the application/Key Vault/Azure App Configuration. Or in other words, the ability to initiate a storage account failover and none of my applications are even aware that it happened (with the exception of some slight downtime or data loss).
Effectively a similar flavor to what you get from setting up an Azure SQL failover, in that you have a connection string that doesn't need to be updated when failover is initiated.

[SumanthMarigowda-MSFT]

@artemuwka Requesting you to please add your comments here

[artemuwka]

Just got back from a short vacation - appreciate your patience, Gavin (@GFoley83 - tagging you here for visibility)! And thanks for the context, this is helpful. Sounds like in your scenario disaster/outage would be one of the main reasons to execute a failover which is exactly what this flavor was designed for. One other consideration to throw into the blend of implications (on top of slight downtime and potential data loss) is that once failover is complete if all other components of the app are still residing in the original primary region there will be cross-region reads/writes happening with larger latency, and you'll be paying for egress on that storage account too. On the other hand, if the entire application gets failed over to the secondary region then these additional items aren't applicable.

I've also had a chance to sync with the dev team on the GA timeline and it's safer to assume that failover GA would happen in January 2020.

If you have any other questions/comments/concerns - please let me know!

[SumanthMarigowda-MSFT]

@GFoley83 We will now proceed to close this thread. If there are further questions regarding this matter, please tag me in your reply. We will gladly continue the discussion and we will reopen the issue.

[GFoley83]

Hey @SumanthMarigowda-MSFT any update on when this will go GA?