[Info-vax] ana/disk/repair

[VAXman- at SendSpamHere.ORG]

In article <slrnkeuesn.12m.brown at libra.gmcl.internal>, Rob Brown <brown at libra.gmcl.internal> writes:
>On 2012-12-26, Phillip Helbig---undress to reply
>  <helbig at astro.multiCLOTHESvax.de> wrote:
>>
>> I gather most of my problems are relatively harmless:
>>
>> %ANALDISK-W-DELHEADER
>> %ANALDISK-W-ALLOCCLR, %ANALDISK-W-BADDIRENT
>> -ANALDISK-I-BAD_DIRFIDSEQ, %ANALDISK-W-BAD_NAMEORDER
>> %ANALDISK-W-LOSTHEADER
>> %ANALDISK-W-FREESPADRIFT
>
>Most of your questions about these are answered by the appropriate
>HELP/MESSAGE command.  I did not find ANALDISK-I-BAD_DIRFIDSEQ on VMS
>7.1 or 8.3.  In most of the other cases, it boiled down to "delete the
>bad directory file and then run ANALYZE/DISK_STRUCTURE/REPAIR to
>recover the lost files".
>
>They seem like good things to do.  Who wants corrupt directory files?

I somehow missed the initial posting.  There've been issues with directory
corruption of late.  I know three separate VMS customers that I have worked
with that have experienced severe directory corruption.  I have to explore
the patch database to see if there's anything yet to address it.  I do know
that there was an F11X patch to increase logging information to assist HP 
with diagnosing its cause but I've not heard if it's been diagnosed and or
repaired.  

%ANALDISK-W-BAD_NAMEORDER is one of the directory corruptions that has been
seen.  This error indicates that the entries in the directory file are out
of alphabetical sequence.  This has been exaccerbated with/in directories
of vary large size.  If you have/has such directories, try to thin them out
to more reasonable size.  

%ANALDISK-W-BADDIRENT is an INVALID entry in a directory.  Generally, there
is no valid FID stored in the directory.  It could be because the file was
deleted but considering your other errors, I'd delve a little deeper before
dismissing it.

%ANALDISK-I-BAD_DIRFIDSEQ is telling that there is a FID<SEQ> number in a
directory that does not correlate with the FID<SEQ> found in the header.
Not the end of the world, mind you, but, again, considering these other
analyze errors, there's something not quite copacetic there.

I'd say you have some real problems with directory corruption on that disk
you are analyzing.  Welcome to the club; however, it's not a club you want
to be a member of! ;)

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

Well I speak to machines with the voice of humanity.