Azure Managed Instance Backups - Could Backup Be Completely Lost?
-
Some of our more regulated customers (banks, gambling sites and suchlike) like to know that we have a robust and tested recovery routine.
We use Azure Managed Instance (MI) and the backup/restore process for 'normal' requirements is quite simple.
But this had me thinking - is it possible for Azure to be in such a mess that the backups aren't obtainable? I would have thought that if that was the case, we (and a large chunk of the UK) would have bigger problems.
The only restore that I have used thus far, when setting up a new MI, has allowed me to point to the MI that has the backups taken from it, and restore from those (via the Azure Portal). But what if that MI was no longer available - are the backups still available elsewhere, or by some other means?
For example, if UK South disappears, can I restore from the DR MI site on UK West?
I'm struggling to find decent documentation from Microsoft, for the more 'exotic' cases of Azure failures.
Thanks for any pointer, to satisfy my curiosity (and the documentation that I might have to produce for this).
Steve.
-
First up a disclaimer: At this time, I and my employer do not use MI, so anything I say is based on reading about MI
My understanding is that the backups taken of an MI database are stored in your blob storage that the MI has access to. So if that blob storage is only locally redundant, then if UK South were just erased from the surface of the UK by a druidic ritual performed at Stonehenge, then no, you would not be able to restore anywhere else.
If, however, you either went with geo-replicated for the blob storage OR your blob storage was in another Azure datacenter (ex UK West) then in our angry druids situation above, your backups would be safe and restorable.
Again, note, I cannot say I am correct, only that based on my readings, this is how it would work.
-
Exactly what Jasona mentioned.
It all depends how you have set the backup preferences, by default it should be Geo replicated to the associated region
However if someone has gone and changed this to local or zone redundancy then you'll want to change it back to Geo for the event in that a region is completely destroyed.
However there is a slight delay in files syncing from the primary to secondary region, we see around a 1 hour delay, so ensure that your RPO is at least that if not longer in the event of a complete region disaster.
-
Check your SQLMI instance / Compute + storage / Backup section
Johan
Learn to play, play to learn !Dont drive faster than your guardian angel can fly ...
but keeping both feet on the ground wont get you anywhere :w00t:- How to post Performance Problems
- How to post data/code to get the best help[/url]- How to prevent a sore throat after hours of presenting ppt
press F1 for solution, press shift+F1 for urgent solution 😀
Need a bit of Powershell? How about this
Who am I ? Sometimes this is me but most of the time this is me
Viewing 5 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply