backup command issue

[Kumar.R]

Hello Bence,

Welcome to the OpenVMS discussion forums.

The first two errors doesn't look relavant to the the last two ones.It looks like dka0: is the system disk and went to mount verification for a while.Or you getting the dka0 errors still?

Comng to the last two errors, it looks like your tape drive could not read from the tape on MKA400.

* This needs a cleaning of the tape drive. with a cleaning disk.

* The tape drive is bad.Needs a replacement
of the tape drive.

Hope this solve your problems.

Thanks and regards,
Kumar

[Bence Richter]

Thanks Kumar.

Actually you are right. It is a fact that I can take backups after the general cleaning of the box + re-plugging the SCSI cables, but DKA0 (CD-ROM) is mounting for too long sometimes. So probably that's another issue.
Maybe my CD-ROM is dying... That's the only thing I can imagine, as I re-checked the cables.

Cheers,
Bence

[Hoff]

I've not seen a dirty tape drive or dirty media or unclean tape drive heads cause disk mount verification errors. The degree of disruption here tends to involve a fairly substantial disruption to what appears to be the system disk, and typically beyond filthy tape heads.

A bad disk or bad tape drive or a spotty SCSI connection or a bad cable or bad termination or such, yes. That can cause bus disruptions.

DKA0: is a fairly unusual unit setting for a SCSI CD-ROM and a fairly common setting for a disk drive. Ensure you don't have duplicate units on the bus, whether with one of the SCSI devices present or with the host SCSI controller (which is usually at unit 6 or 7) in this configuration.

It's easily feasible that the CD-ROM drive or the CD media is tossing errors, and (if that's the device at DKA0:) then that could explain most or all of what you're seeing here.

And gear of this vintage (old SCSI CD-ROM widgets, old SCSI disks, old Mylex DAC 960 controllers, old SCSI cables, etc) can and do fail.

[Jon Pinkley]

Bence,

You may find that your backups and restores are faster if you use a larger blocking factor on your backups. By default backup uses 8192 byte blocks for tape drives, and that is tiny.

See help bakcup/block for more info.

I would suggest replacing

5. $$$ backup/image/record/verify dkb0: mka500:sybck6

with

5. $$$ backup/image/record/verify dkb0: mka500:sybck6/block=65024 !(or at least 32256)

The %SYSTEM-I-MOUNTVER, DKA0: messages may just be a result of the SCSI buss being really busy with the tape backup. Do they occur when backup isn't active? I.e. if you do

$$$ analyze/disk/read DKA0:

do you get MOUNTVER errors? If not, then it is probably just backup hogging the adapter with a deep queue of pending operations.

If you have the 8.3 distribution CD, that has an updated version of BACKUP that may give you better results. I am assuming you are using the 8.2 distribution.

Jon