Cluster partitioned 2 master - back scenario

[19AjaS75]

Hi

I have one question with one scenario:

- situation one cluster was partitioned all 2 partition have master ( in second partition election appear ) in this scenario we have 2 master - that work without problem

Question is what happened when second partition back to main partition ( network, storage connection back ) then we have one cluster without partition with 2 masters and slaves, what HA solve this situation with 2 masters ??

I search documentation blogs but no info what happens next with this 2 masters ?? this is conflict situation new election process executed, which master will still be master ??

[a_p_]

Once the partition is corrected any of the masters will take over.

You are right. It's not easy to find in the documentation. Had to look for this in the "Clustering technical deepdive" book by depping‌ and frankdenneman‌

André

[depping]

and that can now be downloaded / read for free here: http://ha.yellow-bricks.com/

[19AjaS75]

Hi

I read precisely this document from Duncan blog about HA but this is one of best document but inside only one scenario is similar "Data Center Partition"

Duncan write only:

"...If the connection between sites is restored, a classic “VM split-brain scenario” will exist..."

but no info what happen when connection will by restored and 2 master appear in cluster

[a_p_]

It's in chapter 5. Fundamental Concepts | vSphere 6.x HA Deepdive

~snip~

If a cluster is partitioned in multiple segments, each partition will elect its own master, meaning that if you have 4 partitions your cluster will have 4 masters. When the network partition is corrected, any of the four masters will take over the role and be responsible for the cluster again ...

~snip~

André

[19AjaS75]

Hi

But what does it mean this is not clear answer for my I try this scenario in my lab.

[a_p_]

I can't tell you what exact happens under the hood, but I assume some kind of election takes place so that only one of the masters remains master, and the other one(s) become slave(s).

André

[depping]

What is means is that after the partition is cleared and a single cluster is formed again you will only have 1 master left, which typically is the one with the highest identifier. This is based on the same mechanism as the normal election, as stated in my book:

> "The host that is participating in the election with the greatest number of connected datastores will be elected master. If two or more hosts have the same number of datastores connected, the one with the highest Managed Object Id will be chosen. This however is done lexically; meaning that 99 beats 100 as 9 is larger than 1. For each host, the HA State of the host will be shown on the Summary tab. This includes the role as depicted in screenshot below where the host is a master host."