Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Bad Last Block error 103

1 view
Skip to first unread message

Andrew Caplinger

unread,
May 28, 2003, 6:02:09 PM5/28/03
to
I have a client who is having a couple newly recurring issues. The most
pressing one is a BAD LAST BLOCK error in one segment of the dms. When
this happens, every document within a certain date range ending in 4 or 9
or whatever goes away. So far, we've been able to recover these files
(DMDD0x01.DB) from backup and get them reinserted into the system, but we
have no confidence that it won't happen again.

We've implemented the TURBODIS.NLM to disable TurboFAT, but we haven't yet
turned QuickFinder indexing off of Continuous. Is this the only and last
fix? What else can we do to prevent this from happening in the future?

Thanks in Advance!

-Andrew Caplinger

Dave Parkes

unread,
Jun 15, 2003, 9:45:04 AM6/15/03
to
Ensure all clients are running client/server, and I'd set the indexer to
at least one hour. Never really trusted continuous indexing

Cheers Dave

--

Dave Parkes [NSCS]}}
Occasionally resident at http://support-forums.novell.com/

Andrew Caplinger

unread,
Jun 23, 2003, 4:05:48 PM6/23/03
to
Dave Parkes <Da...@Norgren.co.uk> wrote in
news:MPG.1956763...@support-forums.novell.com:

> Ensure all clients are running client/server, and I'd set the indexer
> to at least one hour. Never really trusted continuous indexing
>
> Cheers Dave
>

Thanks for the note, Dave. That was what I had determined from many
sources as well. We spent an evening looking at this and found that two
users were doing Direct Access. We removed all user file rights to the
post office and ensured they were working correctly in client/server
mode. We also changed the indexing to twice a day.

I still am getting a bunch of Error 50's and Error 52's. We purchased
and used your GWCheckL utility to backdate the orphans, which got rid of
a bunch of the 52's (but not all of them), but I still have a number of
Error 50's (46 of them, last I checked.)

I'm trying to get the checks to run clean so we can move forward with an
upgrade to 6.5. At this point I am anticipating having to manually fix
the system by either deleting the references that point to missing
blobs, or manually importing Something into their place, just to make
the data consistent, even if the document doesn't match the profile.
Can you recommend a good article, book or engineer available for hire
who can help me with this? Novell won't until we're neck-deep into an
upgrade to 6.5.

-Andrew

Dave Parkes

unread,
Jun 24, 2003, 1:22:04 PM6/24/03
to

Care to post a bit of the log file, what sort of error 50s are you
getting ?. Favourite is when the library storage area is straight off
the root of a volume, bad things happen, 8209s I think it was, can't
remember off hand. But htat is fairly easy to move down a level on the
same volume.

The 52s will get down to a stable level, but never really go away, they
are no big deal really provided they hover around the same figure. I get
about 200 of those a night here when the structure checks run.

Danita/Greg have got a new DMS book out from www.caledonia.net. Not that
I've seen a copy, they both know I get lost when there are no pictures
<g>

Andrew Caplinger

unread,
Jun 24, 2003, 5:10:37 PM6/24/03
to
I don't have a log handy. I'll be going onsite tomorrow morning
(Wednesday 6/25) and I'll post as soon as I get one. I kinda figured
that about the error 52's. I'm pretty sure the document storage location
is one subdir off the root, but the path is pretty short.
Server/WP:\DOC\ I think.

I suggested to my client that they purchase the upgrade guide, which I'm
hoping will help us get up to speed quickly, and we finagled opening an
incident with Novell on the premise that we are trying to get cleaned up
and ready for an upgrade to 6.5.

Thanks.

-Andrew Caplinger

Dave Parkes <Da...@Norgren.co.uk> wrote in

news:MPG.1962991...@support-forums.novell.com:

Andrew Caplinger

unread,
Jun 25, 2003, 10:56:39 AM6/25/03
to
Ok, onsite now. Here is a sample of the Struct log we've got so far...

- Checking directory FD58
- Blob is a word list file
No further info available for this Blob
Error 50- Orphaned Blob file: 3EE8B1B4.000
Problem 39- Unknown file 3EE8B1B4.000- 144 bytes, 06/12/03
17:00
NOTE- the timestamp on this file is recent, and may reflect a
temporary mismatch between the file system and the databases.
- File is too recent- will not be deleted
- Checking directory FD59
- Blob is a word list file
No further info available for this Blob
Error 50- Orphaned Blob file: 3D4105BD.000
- Blob is a word list file
No further info available for this Blob

and here is the summary:

***** POST OFFICE "****po" STATISTICS *****
*********************************************************************
Uncorrectable conditions encountered:
CODE DESCRIPTION COUNT
---- -------------------------------------------------- -----
50 Orphaned Blob files deleted........................ 58
52 Blob files with missing trailers deleted........... 3
Correctable conditions encountered:
CODE DESCRIPTION COUNT
---- -------------------------------------------------- -----
39 Unrecognized or invalid files in mail directories.. 41
*********************************************************************

We are getting ready to apply SP5 today. Any thoughts would be greatly
appreciated.

-Andrew

Andrew Caplinger <acap...@no-spam-hotmail.com> wrote in
news:Xns93A4A5025C60Ba...@192.233.80.227:

Dave Parkes

unread,
Jun 27, 2003, 9:44:25 AM6/27/03
to

If they are all just word list files, don't worry about them. Now is
that library area /docs/library/fd0 or just /docs/fd0 ?. First is good,
second isn't
0 new messages