Backup, Copy, RMS Errors (OpenVMS 7.3-2)

[Cass Witkowski]

Hein,

Definitely do not want to muddy any waters.

After re reading this post, my case doesn't apply. I have to confirm, it but I think my issue was that the container file was not set /nocache.

As to the problem in this post...

From what I know OpenVMS tends to be rather robust as far as error detection goes.

It it was a hardware error in the disk subsystem, you think that the copy command would exhibit errors as well. The disk subsystem does not know whether backup or copy command was doing the IO.

I'm assuming that the firmware on the MSA1000 is at the latest version. Also that all the connections profiles on the MSA1000 are set to OpenVMS. We had issues on system after we replaced a Fibre Channel HBA on a node but forgot to update the connection profile to OpenVMS.

If the MSA1000 is not active/active, then
it could have been an issue with the MSA1000 controller and when the system rebooted it forced a switch of the FC path over to the other MSA1000 controller.

If the MSA1000 is active/active then we have to look at the system. It is not clear that this test was done from both systems in the cluster. If both systems show the same issue then it is not hardware related to any one system.

If you still had the BACKUP-FILE it would be interesting to dump the file around the errors to see what is written compared to what is supposed to be there. Is the data random? Is it zero? Is it similar to other data in the file? I would look if the errors occur at where the file is fragmented.

Do the EA40s have Elsa Gloria graphics cards? If so see. http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=109447627+1285819675885+28353475&threadId=1251038