Azure managed disk LRS redundancy

Question

According to the documentation, LRS disks are replicated 3 times within a single data center (details).

At the 18/1 there was a single zone power event with this sentence explaining what went wrong (taken from Azure status history site):

While all server racks have redundant dual feeds, one feed was powered down for maintenance, and a failure in the redundant feed caused a complete shutdown of the affected racks.

The sentence above describes replication process of 2 times (dual feeds)

Can you explain the difference between the documentation and the incident?

Answer 1

@Assaf L Adding more information to the above-

These are referring to two different things. #1, LRS disks have data replicated 3 times (across 3 different racks). #2, Racks have dual power supply. If both the feeds supplying to a rack go down, the rack goes out of power. In the 1/18 incident, multiple (more than 2) racks lost power due to both power supplies down.

Does that answer your question? Please let me know. Thanks!

Remember:

Please accept an answer if correct. Original posters help the community find answers faster by identifying the correct answer. Here is how.

Want a reminder to come back and check responses? Here is how to subscribe to a notification.

Answer 2

Hi, Assaf

This doesn't indicate storage replication, this refers to electrical components required for running the gear:

The electrical feeds are dual, but during maintenance, when it was running on 1 out of the 2 - the 2nd one failed.

This unexpected event was caused by a failure in the electrical systems feeding the affected racks. While all server racks have redundant dual feeds, one feed was powered down for maintenance, and a failure in the redundant feed caused a complete shutdown of the affected racks. We continue to investigate the nature of this redundant feed failure to prevent incident reoccurrence.