2 Servers with Failover Cluster and one File Share Witness down time.

Anish Anand 0 Reputation points
2026-08-14T17:43:26.4566667+00:00

Dear Team,

I have two Windows Server 2016 servers configured as Domain Controllers. Both servers are also configured as nodes in a Windows Failover Cluster, and the cluster is currently operating without any issues.

A Windows 10 Pro workstation is configured as the File Share Witness (FSW) for the two-node cluster. The workstation is joined to the same domain but is located on a different subnet. This configuration has been working successfully since implementation.

I am planning to upgrade the Windows 10 Pro workstation to Windows 11 Pro. Before proceeding with the OS upgrade, I also need to upgrade the workstation BIOS and install the necessary drivers. The complete activity is expected to take approximately 3–4 hours, during which the workstation will be offline and may require multiple restarts.

Could you please advise on the recommended approach for performing this upgrade while minimizing any impact on the existing two-node failover cluster?

In particular, I would like to confirm:

Can the File Share Witness remain offline for 3–4 hours without causing a split-brain condition or loss of cluster availability?

Are there any specific precautions or steps we should take before taking the FSW offline?

Is there any recommended procedure to place the cluster in a safe state during the FSW maintenance?

Will the two-node cluster continue to operate normally while the FSW is temporarily unavailable?

Is there any risk of quorum loss or cluster shutdown during this maintenance window?

I would appreciate your recommendations for the safest approach to complete the BIOS, driver, and Windows 11 upgrades with minimal downtime and without affecting the production cluster.

Many thanks,

Anish.

Windows for business | Windows Server | Devices and deployment | Install Windows updates, features, or roles
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Marcin Policht 104.3K Reputation points MVP Volunteer Moderator
    2026-08-14T18:58:15.1133333+00:00

    Yep - the two-node cluster can continue operating while the File Share Witness is offline for 3 - 4 hours, provided the two cluster nodes remain healthy and can communicate with each other. The FSW is not required for normal cluster operation when both nodes are online. Its purpose is to provide the additional quorum vote needed to avoid a split-brain situation if communication between the two nodes is lost.

    Before taking the FSW offline, verify that both cluster nodes are online, communicating normally, and that the cluster currently has quorum. Also verify that the FSW is currently functioning and that there are no existing cluster warnings or quorum issues. I would not deliberately place the cluster into a special "maintenance" or paused state simply because the FSW is being upgraded.

    During the maintenance, the cluster will effectively operate with the two node votes while the FSW vote is unavailable. With both nodes online, the two-node cluster retains quorum and should continue operating normally. The important restriction is that you should not reboot, shut down, or otherwise make either cluster node unavailable during the FSW maintenance window. If one node fails while the FSW is offline, the remaining node would no longer have sufficient votes to maintain quorum, and the cluster could stop its clustered roles.

    So confirm healthy quorum and node-to-node communication, take the FSW workstation offline, perform the BIOS, driver, and Windows 11 upgrades, and bring the workstation back online afterward. Once it is back online and connected to the domain/network, verify that the File Share Witness is again online and that the cluster reports normal quorum. There is no inherent timeout after which an offline FSW causes the cluster to shut down; the critical issue is losing another cluster vote while the witness is unavailable.

    As an aside, you should not cluster AD domain controllers - two DCs provide sufficient high availability without Failover Clustering

    If the above response helps answer your question, remember to "Accept Answer" so that others in the community facing similar issues can easily find the solution. Your contribution is highly appreciated.

    hth

    Marcin

    Was this answer helpful?

    0 comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.