The limitations of cluster sets at large scale deployments .

Question

The limitations of cluster sets at large scale deployments .

Patel Krishna 40

We’ve been testing cluster sets to manage multiple Hyper‑V clusters as one unified fabric, but we’re running into some concerns at scale: with over 3000 endpoints, we noticed that metadata synchronization across clusters sometimes lags, which makes management tasks feel inconsistent.
Another issue is that when we try to move workloads between clusters, the orchestration layer doesn’t always provide clear visibility into resource usage, so troubleshooting becomes harder. We’re also wondering if there are known gaps in monitoring or alerting when you stretch cluster sets across 10 clusters. The documentation highlights the benefits, but at this size we’re worried about hidden limitations that only show up in very large deployments.

I really like to hear from other engineers or any members who have deployed cluster sets at similar scale what challenges how did you work around them ?

0 comments

Answer accepted by question author

0 additional answers

Your answer

Answer 1

Hi Patel Krishna,

What you’re seeing is consistent with some of the known challenges when stretching cluster sets across multiple large clusters. Metadata synchronization can lag when you have thousands of endpoints, especially if the control plane is under heavy load. This doesn’t usually break functionality, but it can make management tasks feel inconsistent until replication catches up.

For workload mobility, visibility into resource usage is indeed limited in the current orchestration layer. Many administrators work around this by integrating System Center or Azure Arc monitoring to provide clearer insights into CPU, memory, and storage utilization across clusters. Without that, troubleshooting can feel opaque.

Monitoring and alerting across 10+ clusters is another area where gaps exist. Native tools provide basic health checks, but at scale you’ll want to layer in centralized monitoring solutions that can correlate events across all clusters. This helps avoid blind spots when metadata sync delays occur.

I hope the response provided some helpful insight. If you find this answer useful, please hit “accept answer” so I know it addressed your concern.

Jason.