topic Re: %MOUNT-F-DRVERR, fatal drive error in Operating System - OpenVMS

%MOUNT-F-DRVERR, fatal drive error

Adrian Gehrig — Fri, 17 Jun 2005 03:24:23 GMT

Once again we are having trouble with our fibre attached tape drives...

Environment: 2 NI-clustered AlphaServer ES40 running OpenVMS V7.3-1, 2 Network Storage Router M2402 for each MSL5026S2

Problem: When mounting a tape on HOST2 $2$MGA0 the following error shows up
%MOUNT-F-DRVERR, fatal drive error
The same physical device works fine on HOST1...
Is there a procedure to bring the tape device back working again. MCR SYSMAN IO AUTOCONFIGURE, REBUILD and so on didn't fix the problem.
It's hard to find a time slot to reboot the node, since there are simulations running which take multiple weeks time...

Thanks for any help in advance

Re: %MOUNT-F-DRVERR, fatal drive error

Uwe Zessin — Fri, 17 Jun 2005 03:51:31 GMT

Adrian,
the systems have two paths to the tape drives. Last time I checked, tape multipath was only supported on the old MDR SCSI/FC converter, but it might have changed.

Assuming that the servers are cabled identically (both PGA0 goes to the same fabric), if you inspect your paths you will see that each server uses a different path.

You could try to move HOST2's current path to that of HOST1 with the 'SET DEVICE/SWITCH' command.

Re: %MOUNT-F-DRVERR, fatal drive error

Vladimir Fabecic — Fri, 17 Jun 2005 04:07:20 GMT

Did you try MCR SYSMAN IO REPLACE_WWID?
Do both machines see same WWID?

Re: %MOUNT-F-DRVERR, fatal drive error

Adrian Gehrig — Thu, 23 Jun 2005 03:46:23 GMT

Thanks to Uwe! The tape drive is accessible again.

Answers to Vladimir's questions:
Both machines see the same devices / WWIDs in the SAN. The SYS$DEVICES.DAT also contains the same records. I tried the SYSMAN IO REPLACE_WWID command also. But it didn't help...