Start a Conversation

Unsolved

This post is more than 5 years old

525

April 4th, 2007 01:00

Device shared between two clustered dedicated storage node

Hi,

I have two dedicated storage nose and they are clustered, i want to do backup to disk so i created a disk and i define it on the cluster admin as physical disk shared and i tested it for failover.
Then, on Legato, on devices, i created 2 new (name: rd=server1:\\.\J as adv-file and rd=server2:\\.\J also as adv-file), now i could mount only one of them.

Question: What should i do in order to get the drive J: mounted on both node??

22 Posts

April 4th, 2007 01:00

I did well the configuration from client side.

I am trying to backup to disk. Can the disk (device) be shared? If yes how can i do that?

Thank you

14.3K Posts

April 4th, 2007 01:00

Shared disk should be backed up via "shared" clientname. In such setups DSN is not supported, but rather SN (officially).

14.3K Posts

April 4th, 2007 02:00

Oh wait, you want to share disk device (eg. ATFD).... yes, that is also possible with recent versions - check documentation for steps.

22 Posts

April 4th, 2007 02:00

Thats what i did, but when i create new devices, what name should i give?

Do i have to create 2 devices, one for node1(rd=node1:\\.\J: and second for node 2 (rd=node2:\\.\J:), or i have to create only one and give it the name of the virtual node (rd=virtualname:\\.\J:)?

Thanks

14.3K Posts

April 4th, 2007 02:00

You can share the disk via cluster and cluster will manage it.

14.3K Posts

April 4th, 2007 05:00

You create two devices.

22 Posts

April 4th, 2007 07:00

Ok, i created 2 drive, but the proble that i am facing it the following:

one device is mounted at a time and if node 1 goes down, the drive will be failover from cluster admin, but i have to manually mount the second device.

Is there is any solution for that issue?

Thanks

14.3K Posts

April 4th, 2007 07:00

On Solaris, we did test this without any cluster software - you simply have one at the time mounted and it works that way... as soon as I one is in use another one gets "in use" status. I think what you are trying here with cluster controlled software is not how it is designed to work. Nevertheless, if you problem is only mounting (which I doubt) simply add script to mount it as action during failover.... AFAIR, DDS during failover is not really something NW handles very nice and I'm not sure if this has been improved in recent versions.
No Events found!

Top