iSCSI Storage and Quorum for Hyper-V Cluster (Windows Server 2016)

[chrisk2305]

Hi Guys,
I am kind of new to FreeNAS and have a couple of questions:

The FreeNAS Server uses 2x2TB SSDs (mirrored through FreeNAS) as iSCSI storage for two Windows Server 2016 hosts. Is it possible to use the volume I created (the 2x2TB Mirror) as storage and as Quorum disk? Would it make sense to e.g. create a Volume with 1.8TB for storage and 1GB for the quorum? (Is that even possible to create two volumes on the same disks?) Sorry if these questions are obvious to answer. Thanks a lot in advance!

Chris

Edit: I guess I just need to create another target/extent for quorum?

 

[Ericloewe]

You want two zvols on one pool and then share each of them via a separate iSCSI connection? Sure, why not, as long as performance suffices for both and you keep enough free space on the pool.

 

[chrisk2305]

yeah, exactly. But I guess I did something wrong, as I made a second target on the same volume and then the windows failover cluster failed at the storage part. So creating 2 volumes on the same raid1 should work I guess? Is Quorum using that much IO? Its only 1GB though

 

[chrisk2305]

can you tell me what the right order is to create 2 zvols on one volume?

 

[Ericloewe]

There is no right order. But your failover cluster comment has me somewhat confused. What are you trying to do, exactly?

 

[chrisk2305]

so basically I need two seperate iscsi targets for a windows server 2016 hyper-v failover cluster (2x Win 2016 as hosts and freenas as shared storage via iSCSI). All I want to do is to create two iscsi targets on 1 mirror (volume/dataset?) consisting of 2x 2tb ssds. these should represent "Quorum" and " Storage" with 1gb and 1.8tb respectivley.

 

[Ericloewe]

How are you going to failover from a 1GB disk to a 2TB disk or vice-versa?

 

[chrisk2305]

1GB is only the witness disk. 2TB is the shared storage for the two hosts.

 

[bigphil]

I assume you're using node and disk majority quorum, which should work just fine in a two node cluster. What are the error details when validating the cluster?