Unsolved
This post is more than 5 years old
54 Posts
0
22541
January 26th, 2011 15:00
MD3000 keeps changing from Optimal to Non-optimal
I have a MD3000 connected to two Dell R710 servers. Each server has two HBAs and the SAS cabling is setup according to the 2 Node Cluster with Redundancy diagram.
The MD3000 has 15 disks. 7 disks in a RAID5 on RAID controller 0, 7 disks in a RAID 5 on controller 1. The remaining disk is a hot spare. I have 4 disk groups (Databases, Logs, Backups, Quorum) configured on the array. Both servers are in a host group which have access to all the disk groups on the array. The cluster appears
to be working correctly and failing over properly. However, in the Dell Storage Manager, the status keeps changing from Optimal to Non-Optimal every few minutes.
If I change the virtual disk ownership path to correct the suggested path for two disk groups (it is the same groups that keep losing their path) the array goes back to Optimal, but soon thereafter it starts switching back and forth from Optimal to Non-optimal.
I have updated the firmware, reseated the cables. It will do this with one server powered on or both.
Any ideas?
The MD3000 has 15 disks. 7 disks in a RAID5 on RAID controller 0, 7 disks in a RAID 5 on controller 1. The remaining disk is a hot spare. I have 4 disk groups (Databases, Logs, Backups, Quorum) configured on the array. Both servers are in a host group which have access to all the disk groups on the array. The cluster appears
to be working correctly and failing over properly. However, in the Dell Storage Manager, the status keeps changing from Optimal to Non-Optimal every few minutes.
If I change the virtual disk ownership path to correct the suggested path for two disk groups (it is the same groups that keep losing their path) the array goes back to Optimal, but soon thereafter it starts switching back and forth from Optimal to Non-optimal.
I have updated the firmware, reseated the cables. It will do this with one server powered on or both.
Any ideas?
No Events found!
JOHNADCO
2 Intern
•
847 Posts
0
January 26th, 2011 15:00
Usually means your paths are not all there.... Check under iSCSI adapater on the configuration tab. Listed under Storage adapters.
kmn96
54 Posts
0
January 26th, 2011 15:00
JOHNADCO
2 Intern
•
847 Posts
0
January 27th, 2011 07:00
So you shutdown one of the cluster members and it stops doing it?
kmn96
54 Posts
0
January 27th, 2011 07:00
JOHNADCO
2 Intern
•
847 Posts
0
January 27th, 2011 09:00
On your MD3000 there is a switch on the front Unified / Split.... That switch being wrong could cause this.
kmn96
54 Posts
0
January 27th, 2011 13:00
JOHNADCO
2 Intern
•
847 Posts
0
January 28th, 2011 13:00