Hyper-V Failover Cluster: CSV with Compression on LUNs from SAN

Lanky Doodle 226 Reputation points
2023-03-15T00:04:33.66+00:00

Hi,

I have a SAN that provides compression/dedupe natively. For example I have a 4TB LUN that has about 3.7TB of data in it, but according to the SAN it's only about 20% full because of compression/dedupe. Here it shows in good health.

However the CSV in Failover Cluster Manager is reporting less than 10% available (so is bad health red). This is not correct but I cannot find a way of getting the CSV to honour the free space amount according to the SAN.

I'm now really interested in how Failover Clustering is going to handle this, as when a CSV gets too low is goes offline to protect itself and the data. Is this process based on what Windows thinks is free (in this case incorrect) or based on what the SAN is reporting?

How do I get Windows and a SAN providing compression to work together on this? Right now, SAN-based compression seems totally unfit for purpose (and dangerous) when used with Windows!

This is rather critical!

Thanks

Hyper-V
Hyper-V
A Windows technology providing a hypervisor-based virtualization solution enabling customers to consolidate workloads onto a single server.
2,536 questions
0 comments

2 answers

Sort by: Most helpful
Most helpful Newest Oldest
  1. Deleted

    This answer has been deleted due to a violation of our Code of Conduct. The answer was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.

    Comments have been turned off. Learn more

  2. Eric Siron 1,251 Reputation points MVP
    2023-03-17T17:19:17.0633333+00:00

    I doubt that you can solve this in the way that you hope. A LUN's space is always logical. For instance, the backing store for a LUN could go the other way by mirroring itself across SAN nodes, but it will still report the logically available space. So, you have 4TB of logical space with 3.7TB used. The fact that it's compressed and deduped doesn't change logical usage.

    What most people do is change the logical size of the LUN. You're monitoring actual use on the back-end, so you can grow its reported space to anything that you feel comfortable with. If you want to minimize the splash damage of overprovisioning that overflows physical storage, you can use multiple LUNs that present more total logical space than physical space.

    As to your question regarding full LUNs, Hyper-V will pause all VMs on a full LUN. They are not saved or turned off, but all processing and I/O stops. You can resume them after you free up some space.

    0 comments