1705 again

[Chris Taylor_6]

Quick question, I reciently replaced a drive in a TL891 which is attached to my CL380. I worked flawlessly before, now I'm getting a 1705 locked scsi bus error. The error happens like this: I start with both nodes off. I spin up the attached storage in the cluster. I turn on the TL891. I have one node in maintenance mode and off. I start the other node (which is not in maintenance mode). When Windows (2003) comes up, my cluster service fails and the drives are not accessible. When I reboot the node I get a 1705 locked scsi bus error after post. The error refers to slot 4 port 1 which is the slot my 64bit differential card is located and the port the library is connected to. The IDs are setup correctly. Previous to the drive replacement this worked fine.

I'm guessing that the drive is terminated when it shouldn't be. It's a DLT7000 and based on the QRG it should be the right replacement drive. I checked quantums website and it shows that the drive is indeed terminated by the number on the second pins from the right.

My question is, should this drive be terminated? I'm thinking that the termination would be handled by the connectors within the library and at the cards. I would test my theory but I broke off the drive loading lever before I put it back in (doh!). I'll have a new lever Tuesday. I'm just looking for confirmation on the termination issue in advance...

Thanks!
Chris

[Chris Taylor_6]

Well, termination or no I get the same error. I'm thinking now that the problem is Windows Server 2003 which has it's own driver and doesn't appear to support clustering.

[Chris Taylor_6]

The problem is indeed Windows Server 2003. In order for this to work with the cluster connectivity kit the library and drive have to be disabled in one server.