July 3, 2020 at 11:39 pm
Hi,
I am using AlwaysOn Availability Groups in SQL Server 2016 with a primary and secondary replica. I have a Quorum file share witness on a separate file server. This file server needs to be rebooted, so I'm wondering what I need to consider before rebooting this box...
Anyone have experience with this?
Also, I have some flexibility here in terms of down time and making adjustments to the primary and/or secondary. I'm really looking for the best way to take down this server and get the AG to be back and synchronized as easily as possible--even if it means an outage or temporarily removing a database from an AG and adding it back in later, etc.
Any guidance is appreciated!
Thanks,
Mike
EDIT: I have since found this thread, which essentially answers my question.
But would someone mind confirming that my assumption is right in that I wouldn't need to do anything special whatsoever to reboot the file share witness? Simply reboot it, let the AG continue to operate as normal, and then when the file share witness comes back online, carry on?
Thanks,
Mike
Mike Scalise, PMP
https://www.michaelscalise.com
July 4, 2020 at 1:34 pm
As long as your cluster maintains quorum - the AG will continue to function and you won't have any issues, that is why you have the file share witness.
For a cluster to be healthy - you need to have enough quorum votes to count as more than 50% of the total quorum votes. In a 2 node cluster - you have to configure the quorum to be at least 3 quorum votes. If you only had 2 quorum votes - one for each node in the cluster then the count would be 1 out of 2 when one of the nodes was down - and 1 of 2 votes is not greater than 50% of total votes.
So - you add the file share witness to get a total of 3 quorum votes and now you can maintain quorum as long as you have 2 of the expected 3 quorum votes up and available. One node and the file share witness = 2 out of 3 votes...2 nodes and no file share witness = 2 out of 3 votes. 1 node and no file share witness = 1 out of 3 votes and cluster is not healthy and will be shut down.
This is why we can restart any node in a cluster without taking the cluster down - as long as you maintain quorum the cluster is healthy and services will be up and available.
Jeffrey Williams
“We are all faced with a series of great opportunities brilliantly disguised as impossible situations.”
― Charles R. Swindoll
How to post questions to get better answers faster
Managing Transaction Logs
July 5, 2020 at 11:34 pm
As long as your cluster maintains quorum - the AG will continue to function and you won't have any issues, that is why you have the file share witness.
For a cluster to be healthy - you need to have enough quorum votes to count as more than 50% of the total quorum votes. In a 2 node cluster - you have to configure the quorum to be at least 3 quorum votes. If you only had 2 quorum votes - one for each node in the cluster then the count would be 1 out of 2 when one of the nodes was down - and 1 of 2 votes is not greater than 50% of total votes.
So - you add the file share witness to get a total of 3 quorum votes and now you can maintain quorum as long as you have 2 of the expected 3 quorum votes up and available. One node and the file share witness = 2 out of 3 votes...2 nodes and no file share witness = 2 out of 3 votes. 1 node and no file share witness = 1 out of 3 votes and cluster is not healthy and will be shut down.
This is why we can restart any node in a cluster without taking the cluster down - as long as you maintain quorum the cluster is healthy and services will be up and available.
Jeffrey,
That's what I thought. Thank you for that information!
Mike
Mike Scalise, PMP
https://www.michaelscalise.com
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply