cancel
Showing results for 
Search instead for 
Did you mean: 

Creating new EV9 partition on 2008 R2 cluster

PhilTr
Level 3

We've implemented EV 9.0.1 on a 2008 R2 two-node cluster and have reached the point where we need to open a new partition. I've presented a disk and added it to the Resource Group, but when it comes to the file share (which always used to be automatic in non-clustered mode) I'm not sure what EV does automatically and what I need to do manually.

Will the New Partition Wizard create a clustered file share on the new disk by itself?

Or do I need to create the share myself in Failover Cluster Manager before or after creating the partition? If I need to do it manually, what process should I follow?

Very grateful for any advice.

1 ACCEPTED SOLUTION

Accepted Solutions

JesusWept3
Level 6
Partner Accredited Certified

When it comes to partition and storage, EV doesn't really have any code differences that i'm aware of when it comes to creating partitions, as long as it can write to the drive, create the directory and then the hidden admin share, it will work just fine.

Whether its iSCSI, SAN, NAS, NetApp, doesn't matter, but i think as well EV could also be configured in such a way where as the Partition directory could be written to a local only drive, so when it fails over the other server will fail out because it can't read that partition. which is bad :)

As far as i'm aware, when the storage service starts if the Hidden share does not exist it will just create one at that time, so you could create the partition as you normally would, pointing it at a shared resource and then when it fails over and the storage service comes up as long as the paths are all the same, it will create the hidden share and you will be fine

https://www.linkedin.com/in/alex-allen-turl-07370146

View solution in original post

4 REPLIES 4

JesusWept3
Level 6
Partner Accredited Certified

When it comes to partition and storage, EV doesn't really have any code differences that i'm aware of when it comes to creating partitions, as long as it can write to the drive, create the directory and then the hidden admin share, it will work just fine.

Whether its iSCSI, SAN, NAS, NetApp, doesn't matter, but i think as well EV could also be configured in such a way where as the Partition directory could be written to a local only drive, so when it fails over the other server will fail out because it can't read that partition. which is bad :)

As far as i'm aware, when the storage service starts if the Hidden share does not exist it will just create one at that time, so you could create the partition as you normally would, pointing it at a shared resource and then when it fails over and the storage service comes up as long as the paths are all the same, it will create the hidden share and you will be fine

https://www.linkedin.com/in/alex-allen-turl-07370146

PhilTr
Level 3

Thank you, that's reassuring :)

The thing that confused me is that when the Convert to Cluster wizard ran, it caused a clustered FileServer resource to be created for each of the existing disks in the EV resource group. I can also see the hidden share names it's created (like EVPartition01cbe0ca8bd81180$) in the Shared Folders section of Failover Cluster Management.

I guess all I want to ensure is that I create my new partition in the same way the Convert to Cluster wizard created the others.

JesusWept3
Level 6
Partner Accredited Certified

yeah the hidden shares are created for every partition in EV8+ (in order for the OSIS to work correctly SIS across partitions and vault stores), i think the cluster wizard just creates the drives so that at that point in time it knows everything is in order

https://www.linkedin.com/in/alex-allen-turl-07370146

PhilTr
Level 3

Just in case there's anyone else who has a similar question in future, I'm pleased to report that I created the new partition last weekend just using the New Partition Wizard and exactly as JesusWept2 said, by pointing it at a clustered disk the clustered share and fileserver resource were created correctly and worked on failover to the other node without any intervention from me.

I'm now running EV from the other node and all is good.