[critical] Windows Server 2022 cluster can't update nodeweight during Manual failover without quorum witness if it only have even nodes.

Anonymous
2023-11-02T23:22:25+00:00

Windows Server 2022 cluster can't update nodeweight if it configured manual failover without witness.

Please note : This is Manual failover only. This specific case don't care Quorum since 2 multisubnet. Only one site is preferred.

However, failover not possible unless Force Quorum. Failover itself work. However, when the secondary site down, entire cluster down since the nodeweight not updatable and current vote doesn't move.

Same approch is working on Windows Cluster 2019 and lower version.

Produce version : Windows Server 2022 Standard

Version : 21H2

OS Build : 20348.1970

<Reproduce issue>

2 Node - multisubnet (1 node per subnet)

Node1 - Primary

Node2 - Secondary

No witness(Node Majority)

DynamicQuorum = 1

  1. Node1 set as nodeweight =1, Current vote =1
  2. Node2 set as nodeweight =0, Current vote =0

Start Failover>

Service moved to Node2 as Primary.

After that, update nodeweight as below.

Node2 set as nodeweight =1

Node1 set as nodeweight =0 -> this is failing with following message "because changing the node weight would violate the minimum cluster quorum requirements"

It works in lowerversion of Windows cluster. Current Vote is updating to corresponding node when AssignedVote(nodeweight) is updated.

Because of this failure in Windows Server 2022, it can't failover other node and maintenance isn't possible since Node2 can't goes down since it occupied current vote as 1. it's not updatable.

Please check that ASAP.

Windows for business | Windows Server | Storage high availability | Virtualization and Hyper-V

Locked Question. This question was migrated from the Microsoft Support Community. You can vote on whether it's helpful, but you can't add comments or replies or follow the question. To protect privacy, user profiles for migrated questions are anonymized.

0 comments No comments
{count} votes

4 answers

Sort by: Most helpful
  1. Anonymous
    2023-11-06T09:17:56+00:00

    Hi,

    According to the error message, it seems the node1 is the only node assigned to vote. Have you checked the node weight of Node2?

    (Get-ClusterNode Node2).NodeWeight

    Also check the current cluster function level.

    (Get-Cluster).ClusterFunctionLevel

    Best Regards,

    Ian Xue

    0 comments No comments
  2. Anonymous
    2023-11-06T18:22:25+00:00

    The failure is happening when I update nodeweight.

    Note : I want to manual failover only since this is site to site failover. That's reason why we update NodeWeight to one site only and set as disabled other site.

    (Get-ClusterNode Node2).NodeWeight=1 <- this action works.

    (Get-ClusterNode Node1).NodeWeight=0 <- this is failing with error messages. "because changing the node weight would violate the minimum cluster quorum requirements"

    (Get-Cluster).ClusterFunctionLevel is 11

    0 comments No comments
  3. Anonymous
    2023-11-07T06:06:03+00:00

    Don't assign a value to the NodeWeight attribute yet. Check the current value of the NodeWeight property by running

    (Get-ClusterNode Node1).NodeWeight

    (Get-ClusterNode Node2).NodeWeight

    0 comments No comments
  4. Anonymous
    2023-12-28T18:21:07+00:00

    We need to update it due to preferred site only.

    Again, the objective is maintain Cluster role in prefered node(preferred site) only.

    The issue is happening while the maintnenace running, we can't change it.

    Please note that same behaviour work in lower windows cluster.

    This is only issue with Windows 2022 cluster.

    0 comments No comments