Hi!Originally Posted by jens
I 've done a cluster test setup (Win 2008 R2 Datacenter RC) today and I had the same "error", but it's no error it's "only" a warning. I 've seen no malefunction so long ...
But it would be great to fix this also ...
Hi!Originally Posted by jens
I 've done a cluster test setup (Win 2008 R2 Datacenter RC) today and I had the same "error", but it's no error it's "only" a warning. I 've seen no malefunction so long ...
But it would be great to fix this also ...
regards,
Lukas
descience.NET
Dr. Lukas Pfeiffer
A-1140 Wien
Austria
www.dotnethost.at
DSS v6 b4550 iSCSI autofailover with Windows 2008 R2 failover cluster (having still some issues with autofailover).
2 DSS: 3HE Supermicro X7DBE BIOS 2.1a, Areca ARC-1261ML FW 1.48, 8x WD RE3 1TB, 1x Intel PRO1000MT Dualport, 1x Intel PRO1000PT Dualport.
2 Windows Nodes: Intel SR1500 + Intel SR1560, Dual XEON E54xx, 32 GB RAM, 6 NICs. Windows Server 2008 R2.
BTW, I am on the SCST mailing list (SCST is what open-e uses for iscsi and fibre channel targets), and it looks like Vlad, the main SCST developer, has done a short review of the persistent reservation code code that open-e's engineers submitted. Most of his comments are regarding coding style improvements (ie not using // comments), but he had some architectural improvements' too. This makes SCST the only linux iscsi target that works with Win 2008 clustering, unless LIO has made considerable progress. This means that it might draw more developers into developing SCST. So, there are improvements in the works!
Our customer was getting first :
"Failed to Register PR key for cluster disk"
so we upgraded to DSS V6 and now he's getting:
"Storage does not support SCSI page 83h VPD descriptors for cluster disk "
I've advised to remove bondig on eth ports and see how it goes, but I can't sey that customer is happy about this.
Martin
I don't know about anyone else, but this is a huge dealbreaker for me. My plan is to cluster my Hyper-V Servers and I can't do that until this is resolved. Unless this is resolved I'm going to be forced to purchase a Dell or Equallogic SAN for twice the money, which is something I DON'T want to do!
I am using the DSS v6 trial, will be purchasing the upgrade shortly. Windows Server 2008 (& R2) clustering works fine, the only 'warning' now being presented is about the SCSI device Vital Product Data (VPD), I assume this functionality is being worked on currently? Could we have an update from the team pertaining to how this is going please?
Ignore previous message, if you are trying to do a Hyper-V based cluster with Cluster Shared Volumes, it does not appear to work, a guess time-line is now needed please open-e...
Problem solved, I apologise for being too hasty...Originally Posted by CuppaT
Right, this is how it seems to working, or not!?
My earlier problem appeared to be an issue within the cluster, not realted to the open-e software at all. That is now resolved.
The problem I am having now, is as follows;
Three node Windows 2008 R2 Cluster, Cluster Shared Volumes enabled. Add a 1TB 'disk' to the cluster storage and then add that disk to the cluster shared volumes 'pool'. The Hyper-V machine is then configured to be highly available, success so far. Attempt to migrate the Hyper-V machine from server A to B fails. The CSV is supposed to accessible to all nodes within the cluster at the same time. This is not the case and I believe this is due to the current lack of 'SCSI device Vital Product Data (VPD )' support on the Open-E DSS V6 software.
Can someone from Open-E support either confirm my belief, or tell me that DSS V6 does support Windows Server 2008 R2 Clustering completely, including Cluster Shared Volumes?
Thank you
Thanks for putting all your work into CuppaT, you are doing exactly what I am planning to do! Your results are exactly what I was afraid of.
No worries, cphastings, I am now running a full HA Hyper-V cluster, but (and it is a big but) I have not been able to go down the preferred route of CSV's.
I hope it won't be long before someone from open-e updates this thread with some info on the last remaining issue to be ironed out...