CSV will not attach – 2‑node Hyper‑V cluster with StarWind VSAN (iSCSI)

[EdwardW]

Hello,
I’m running a 2‑node Hyper‑V cluster with StarWind VSAN (iSCSI, MPIO) using the compute/storage separated design.
Cluster builds successfully, StarWind devices show Healthy, iSCSI is connected, and SCSI‑3 PR tests pass — but CSV will not attach.
Symptoms:

Cluster Shared Volumes = 0
Available Storage stays Offline
Storage validation fails Validate Disk Failover

Validation report shows:

Only one StarWind disk visible to both nodes
Other StarWind disks visible to only one node, skipped during validation
Result: disk failover test fails

On both Hyper‑V nodes I see frequent:

Event ID 129 (iScsiPrt) referencing \Device\RaidPorts during disk arbitration
MPIO DSM appears Degraded

Questions:

For CSV use, must every StarWind device be presented identically to both cluster nodes?
Will CSV fail if even one StarWind LUN is visible to only a single node?
Any StarWind‑specific or MPIO tuning recommendations to avoid iSCSI Event ID 129 during cluster disk failover?

Thanks for any guidance.
— Ed

[yaroslav (staff)]

Welcome to StarWind Forum.
Normally, validation does not Fail, it produces the waning.
Connect all targets, including the cluster witness locally and to the partner

For CSV use, must every StarWind device be presented identically to both cluster nodes?

Yes. It must be replicated, ideally, and connected to all cluster nodes. Please refer to the guide (e.g., https://www.starwindsoftware.com/resour ... %20targets).

Will CSV fail if even one StarWind LUN is visible to only a single node?

Yes.

Any StarWind‑specific or MPIO tuning recommendations to avoid iSCSI Event ID 129 during cluster disk failover?

Connect the iSCSI targets.

Good luck with your project