Queue Manager problem

[Art Wiens]

VAX 4100, VMS v6.2.

I was reviewing my nightly disk to disk backup log for this system and saw this:

%BACKUP-E-OPENIN, error opening $40$DIA5:[]SYS$QUEUE_MANAGER.QMAN$JOURNAL_OLD;1 as input
-SYSTEM-W-NOSUCHFILE, no such file

I checked for a current file, and found that it was created yesterday morning. I went through the OPERATOR.LOG file and found no mention of this occuring, and no phone calls came yesterday with any kind of queue issue - this system does alot of batch processing and printing. The QUEUE_MANAGER process's uptime matches system uptime (~60 days) ie. the process didn't crash/restart.

Why would it do this? Where else can I look to see this event?

Cheers,
Art

[Ken Robinson]

The clue here is the lack of a directory between the brackets "[]".

This usually indicates a left over file marked for deletetion or other such inaccessible file.

I would check if the disk needs rebuilding
$ sho dev d/reb

and then I would do a
$ ana/disk/norepai

$ ana/disk/repa ! if needed

Ken

[Volker Halle · Accepted solution]

Art,

QUEUE_MANAGER will temporarily rename the .QMAN$JOURNAL file to .QMAN$JOURNAL_OLD while checkpointing, i.e. while writing the in-memory data to a new .QMAN$JOURNAL file.

Once the checkpointing operation is complete, the .QMAN$JOURNAL_OLD file is automatically deleted.

This is expected behaviour. Your BACKUP seems to just have hit that directory, while checkpointing was active.

Volker.

[Art Wiens]

I understand that, but my question is more, why did this happen? What would cause it to seemingly "remove" the previous journal file and create a new one? Where can I see it having done this?

Art

[Art Wiens]

Thanks Volker, your reply slipped in before my reply to Ken.

Ok, I'll see what the log says tomorrow.

Cheers,
Art