how to unloack dev

[SCC_2]

I have a new scsi hd and install all files from backup. And after I boot it up, it looks o.k. But when I try to edit a files. The system said :
"error open dub0:[data.file]add_per.jou, as output
-rms-e-wlk, device currently write lock.

when I type sh dev/full dub0:
last line show
subject to mount verification, allocation inhibited because of error on bitmap,
write-through cashing enable.

how can I reset the drive/or fix in-order I can write to it.
Thanks !
SCC

[Uwe Zessin · Accepted solution]

The system has found out that the data in the storage allocation bitmap is invalid, so it cannot reliably find out which disk blocks are unused.

Try an "$ ANALYZE /DISK_STRUCTURE /REPAIR DUB0:".

This should reconstruct the bitmap from the allocation data in the file headers. Good luck!

[Robert Gezelter]

SCC,

I suspect that you are having problems with that drive. In any event, I would not just unlock it and use the drive.

A thorough ANALYZE/DISK/REPAIR and definitely some ANALYZE/DISK/READ is in order, at the least.

Have you tried dismounting and remounting the drive?

- Bob Gezelter, http://www.rlgsc.com

[SCC_2]

Hello,

I try the ana/disk_structure/repair and there a lot of errors show up on screen.
after that I can edit the file now.

Before I perform this, my system job quene manager is not running while boot up and it still can't.

I reboot the system again, and do another ana/dis_structure/repair all error message is gone, except 2 messages
1). %verify-w-futurevdat file(2,2,1) [000000]bitmap.sys;1
revision date is in the future.

and %verify-e-sorterror, error sorting multiply allocated block list.

I install this HD to another Vax system. I can mount or dismout it. But I have to use lable as sysdisk.

Thanks !
SCC

[Uwe Zessin]

The queue manager usually gives an error message when it cannot start - I would also check recent versions of SYS$MANAGER:OPERATOR.LOG for errors.

If nothing helps, you need to create a new queue file.

1) This error message usually comes up when then time on the system was incorrectly set.

> %verify-e-sorterror, error sorting multiply allocated block list.

There should be one or two more lines below this message which give a more precise hint. If this is your boot disk and you are running a very old version of VAX/VMS (it's something like V5.5, right?), then the problem is that the ANALYZE utility tries to create some temporary files on the sytem disk. Unfortunately, the system disk is write-locked during an ANALYZE /DISK_STRUCTURE /REPAIR.

If I recall correctly, the workaround is to mount another disk and redirect the logical name SYS$SCRATCH to a directory on this disk. More recent versions can do an in-memory sort, but obviously that won't help you...

A privately mounted disk can have the same volume label as the system disk (was that your question?).

[Robert Gezelter]

SCC,

I would be very VERY cautious about using this disk as a system disk. Unrecoverable errors on system disks can lead to serious consequences.

- Bob Gezelter, http://www.rlgsc.com

[SCC_2]

Hi Robert,
Yes I am 100% agree with you. That is my 2nd. step I am going to install another disk here next week and init this disk and start from scratch and see what happening.
Thanks !
SCC

[Antoniov.]

I think your HD is not compatible with your hardware so it goes continously in mount verification.
You can do the follow dangerour operations:
$ ANALYZE/DISK/REPAIR
$ SET VOL DUB0 /NOMOUNT
Setting /NOMOUNT disable message, not problem. However, when a disk is not compatible with hardware, continuous mount verification messare are more dangerous than disablig check.
Don't forget use another hd next week!

Antonio Vigliotti

[Uwe Zessin]

Have you really tested this, Antonio?

It is my understanding that:
$ SET VOLUMNE dev: /NOMOUNT_VERIFICATION

entirely disables the mount verification. It does not simply discard the messages - the whole volume is invalidated when the disk goes offline.

[Uwe Zessin]

I forgot:

from SCC's message:
> last line show
> subject to mount verification,

That only means that mount verification is enabled on the volumne, not that it is undergoing one at the moment. OpenVMS' ability to do mount verifications is a good thing!!

[Antoniov.]

Uwe,
yes I tested it.
Some year ago I tryed a non dec hd on old system. Vms goes continually in mount verification error. I disabled it and I could work; as I posted this feature is very dangeorus, in my case sometime I lost data. However mey be a temporary solution to extract data from hd and restore to new device.

Antonio Vigliotti

[Antoniov.]

Uwe,
as you posted SET VOL /NOMOUNT disable mount verification.
However my previous post is valid.

Antonio Vigliotti

[SCC_2]

I am going to test another Hd to install all my data back again and see I have the same porblem. So far this HD is still working O.k
except the que manager can't run, the rest of the program perform o.k. But I have to
do a ana/disk_structure/repair everytime time if I shut the system down. If not I can't edit any programs.
SCC

[Uwe Zessin]

Can you connect the disk drive to another system and run the ANALYZE there? It is possible that the disk drive has a permanent error.

To diagnose the queue manager problem:
- do a $ REPLY/ENABLE
-- if this is not the operator console or
-- it is disabled

- do a $ START /QUEUE /MANAGER
-- you should get an error message

if nothing helps, I am afraid you have to reinitialize the queue database:
$ START /QUEUE /MANAGER /NEW_VERSION

hopefully you have documentation and/or DCL procedures that assist you in setting up all forms and queues again.