question

salves avatar image
0 Votes"
salves asked salves edited

SQL AlwaysOn fails when I restart domain controller (quorum)

Hi friends,

i have a sql cluster with alwayson and today i noticed a scenario.

When I restarted the domain / dns server that is my quorum when trying to authenticate to the databases I had difficulty and generated several failures in the cluster of quorum unavailable.

So I asked myself regarding the importance of the availability of my domain controller server to access the databases.

However I have another domain controller server (secondary) which is also DNS.

Can anyone tell me how long it takes servers that authenticate to the domain/dns server need to direct connections to the secondary server?

Can this be adjusted to be faster?

Thanks.

sql-server-general
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

1 Answer

Cathyji-msft avatar image
0 Votes"
Cathyji-msft answered salves edited

Hi @salves,

Quorum failure will cause all clustered services, SQL Server instances, and Always On availability groups, in the WSFC cluster to be set offline.

To bring the WSFC cluster back online, you must correct the root cause of the quorum failure under the existing ,configuration recover the affected databases as needed, and you may want to reconfigure the remaining nodes in the WSFC cluster to reflect the surviving cluster topology.

You can use the forced quorum procedure on a WSFC cluster node to override the safety controls that took the cluster offline. This effectively tells the cluster to suspend the quorum voting checks, and lets you bring the WSFC cluster resources and SQL Server back online on any of the nodes in the cluster.

Refer to this MS document WSFC Disaster Recovery through the Forced Quorum Procedure to get more information.


If the answer is helpful, please click "Accept Answer" and kindly upvote it. If you have extra questions about this answer, please click "Comment".



· 3
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

Hi @Cathyji-msft ,

I understand this concept, but it made me very worried because I have a high availability critical cluster and I'm dependent on a quorum share? This is very bad.

What I noticed is that my databases were not unavailable. I had a problem accessing one of the bank servers while the other was still available.

Generated alarm in cluster stating that quorum was unavailable only.

I believed that my problem was related to windows authentication and not quorum availability.

Do I need to restart the windows server to install windows updates, ie will my databases be unavailable whenever I do this?

Is there any option to get around this scenario with prevention?

Thanks.

0 Votes 0 ·

This really has nothing to do with SQL Server.

It sounds like you are only have 2 votes. You need at least 3 votes.

I suggest you read this:
https://docs.microsoft.com/en-us/azure-stack/hci/concepts/quorum

0 Votes 0 ·

Hi @Cathyji-msft @TomPhillips-1744 ,

but i got 4 votes.

I have 3 votes nodes and one more quorum nodes, that's why I noticed that my bases are still offline.

The only warning was that the quorum was offline.

My question was the time that the active directory server was restarting and I couldn't authenticate to the databases of one of the nodes.

I thought it was authentication time until he tried to connect to the secondary active directory server, but it didn't happen quickly.

When the active directory pirmario server came back I was able to authenticate myself immediately.

Or he was connected to the secondary server and when the primary server returned while the OS did not have windows resources available, he was stuck waiting.

Thanks.

0 Votes 0 ·