Cluster Quorum / File Share - needs to be clustered?

Question

Hi everyone,

we want to deploy a 2-node Hyper-V hyperconverged cluster.

We come from a big installation with Bladecenter, 3PAR, SAN etc but most of our workloads are in the cloud now.

In our old installation we had the cluster quorum on the 3PAR. Now we want to use a file share on a bare metal server outside of the cluster.

BUT - i cannot really find suggestions how to "secure" the quorum. How long can it be unavailable (due to windows patching or switch patching) until the cluster go nuts? Also with a cloud witness - how long can the internet connection be down?

Do i need another "cluster" to secure the quorum?

BR

Stephan

Answer 1

Hello,

When deploying a Hyper-V hyperconverged cluster, the cluster quorum configuration is a crucial aspect to ensure high availability and proper failover behavior. The quorum is used to maintain cluster consistency and determine the cluster's health and status.

In a 2-node Hyper-V cluster, the quorum configuration is particularly important because a simple majority of voting nodes is required for the cluster to function. If one node goes offline, there needs to be a way to ensure that the remaining node can maintain the quorum and continue operating.

Here are some considerations and best practices for securing the quorum in a 2-node Hyper-V hyperconverged cluster:

File Share Witness:

Using a file share witness on a separate bare metal server outside the cluster is a common approach for 2-node clusters. It provides a third voting member, ensuring that there is always a majority (2 out of 3 votes) even if one node goes down.

Securing the File Share:

The file share that acts as the cluster witness should be highly available and secured. It should be hosted on reliable hardware with proper backups to minimize the risk of downtime due to hardware failures. It should also have sufficient access controls to prevent unauthorized access.

Network Redundancy:

Ensure that the network connecting the cluster nodes and the file share witness is highly redundant. Use redundant network adapters, switches, and network paths to minimize the risk of connectivity issues.

Internet Connection for Cloud Witness:

If you choose to use a cloud witness (e.g., Azure Blob Storage) as a quorum witness, the internet connection needs to be reliable. The cloud witness does not require a constant internet connection once the cluster is established, but the cluster nodes need to communicate with the cloud witness during cluster operations and changes (e.g., failover, adding/removing nodes).

Connection Timeout and Tolerance:

The exact timeout and tolerance for the cluster to go offline due to quorum loss can vary based on your specific deployment and business requirements. By default, Windows Server Failover Clustering (WSFC) allows a 60-second timeout for the cluster to lose quorum before initiating a failover. However, you can adjust this value based on your needs.

Monitor Quorum Status:

Implement monitoring for the quorum status, cluster nodes, and the file share witness. Regularly check the health of the cluster and address any issues promptly.

Testing and Maintenance:

Perform regular testing and maintenance to validate the quorum configuration and ensure that all components are functioning as expected.

While a separate "quorum cluster" is technically possible, it adds complexity and additional points of failure. In most cases, using a file share witness or a cloud witness should suffice for a 2-node Hyper-V hyperconverged cluster. With proper planning and redundancy, you can achieve a highly available and reliable setup.

I used AI provided by ChatGPT to formulate part of this response. I have verified that the information is accurate before sharing it with you.

Hope this resolves your Query !!

--If the reply is helpful, please Upvote and Accept it as an answer–

Answer 2

Hey,

As mentioned, quorum is very important in 2 node configuration. Your cluster will still run, if witness fails, because it will have quorum. However, in case one of the nodes fails during quorum outage, your cluster will fail. Check the following article for more information: https://techcommunity.microsoft.com/t5/failover-clustering/understanding-quorum-in-a-failover-cluster/ba-p/371678

You can deploy Disk Witness using StarWind VSAN. StarWind VSAN will create replicated LUN between the nodes. In this case, you will still have Witness available, even if one of the nodes fail. Check the following guide for configuration process: https://www.starwindsoftware.com/resource-library/starwind-virtual-san-for-hyper-v-2-node-hyperconverged-scenario-with-windows-server-2016/

StarWind VSAN can also be used as a shared storage on your 2 node cluster.
https://www.starwindsoftware.com/starwind-virtual-san

Cheers,    
Alex Bykovskyi    
StarWind Software    
Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

Answer 3

Thanks for your answers. We went with a cloud witness (as recommendation from MS) - we tested it yesterday and stopped the internet connection on the cluster servers. It had no big outcome. We could even move VMs between the hosts.