BA356-TruCluster

[banini]

hi all,
My plate-form is composed by: two ES45 and two shared disk BA356 under Tru64 5.1B and trucluster
When I have to replace the failed disk of my shared disks, the local HP partner recommends to switch-off the both nodes. it's the high availability environment and I think there is another solution to avoid this downtime.
Any help well be more than welcome, thank you.

[Vladimir Fabecic]

Hello Banini
BA356 is just a disk shelf. Where is it connected? Little bit more information would be helpfull.
By the way, this is a HOT SWAP disk shelf and that means that disks can be removed and added online. But how are disks in that shelf configured?

[banini]

Hi Vladimir,
thks for your reply, my shared box is connected directly without any Raid controller, like HSZXX. I have two shared box mirrored by LSM.

[Vladimir Fabecic · Accepted solution]

In that case everything depends on LSM configuration. If LSM is configured clever enough, there is no reason to shutdown both servers because of changing failed disk.
Ask them to give you some arguments why cluster has to shutdowned when changing failed disk in HOST SWAP disk shelf.