Abend whilst running purge.

Stuart Robinson

unread,

Dec 8, 2000, 4:48:00 PM12/8/00

to

I was running purge on two different PCs (different directories) and the
server abended.

It ran a SYS vrepair on startup, 4 errors but I dont know what they are. I
am gong to run a more thorough vrepair later, but any suggestions as to
what the problem might have been, invalid deleted file sounds like disk
corruption.

*********************************************************

Server FS1 halted Thursday, December 7, 2000 1:45:10 pm
Abend 1: Server-4.11a: ScanLimboFilesInDirectory found an invalid deleted
file.

Registers:
CS = 0008 DS = 0010 ES = 0010 FS = 0010 GS = 0010 SS = 0010
EAX = 0FB48DA0 EBX = 00000000 ECX = 0FB48060 EDX = 00000D00
ESI = 00000000 EDI = 00000000 EBP = 0E47CC8C ESP = 0E47CC70
EIP = 00000000 FLAGS = 00003246

Running process: Server 18 Process
Created by: SERVER.NLM
Stack pointer: E47D004
Stack limit: E47A010
Scheduling priority: 0
Wait state: 00
Stack: F8044BF5 ?
--D7F06A27 ?
--0FB8FB60 ?
--0FB48DA0 ?
--000D121A ?
--00000000 ?
--00000000 ?
--0E47CCD4 ?
--0000001E ?
--00000000 ?
--00000000 ?
F805C62C ?
--00000000 ?
--000308AC (SERVER.NLM|ddh_keyStateFlags+C9EC)
--00000546 ?
--000D1213 ?
--00000004 ?
--0E47CD08 ?
--0E47CDDC ?
--0E47CCC4 ?
--00000000 ?
--80000000 ?
--063C47A0 ?
0001789E (LOADER.EXE|(Code Start)+808E)
--0000952A (DS.NLM|DSF9078260+61F7)
--0E47CD18 ?
--00000000 ?
--00000004 ?
--0E47CDDC ?
F8061F33 ?
--0000001E ?
--00000000 ?

Additional Information:
The NetWare OS detected a problem with the system while executing
a process owned by SERVER.NLM. It may be the source of the problem or
there may have been a memory corruption.

Stuart.

Stuart Robinson

unread,

Dec 8, 2000, 6:46:00 PM12/8/00

to

Further info.

I ran a purge on the volume from a Win 98 client. Twice the server abended
with the same message, but each time it had purged more files.

I have isolated the problem to a particular user directory of FAXNOW!.
This users directory has circa 1.6Gb of purgeable files (because every
other directory has now been purged).

When the purge gets to this directory it slows down considerably, couple
of files a minute maybe.

I dis-mounted the volume and ran a vrepair, no errors reported.

Most odd.

Stuart.

Barry Schnur

unread,

Dec 8, 2000, 11:26:39 PM12/8/00

to

>I dis-mounted the volume and ran a vrepair, no errors reported.
>

How about running vrepair with the purge option....

Barry Schnur
Novell Support Connection Volunteer Sysop
http://support.novell.com/forums/
Please post replies ONLY via the Newsgroup

Stuart Robinson

unread,

Dec 9, 2000, 3:53:00 AM12/9/00

to

With over 1.6Gb of files still left as purgeable, it would be massive
error log time, trees in a forest etc.

If there are no files left to purge, I presume there is no benefit to
running vrepair with purge on.

But since I seem to be getting abends running Dos/Windows purge, would the
vrepair with purge do better at cleaning them out ?

Stuart.

Barry Schnur

unread,

Dec 9, 2000, 1:11:33 PM12/9/00

to

>With over 1.6Gb of files still left as purgeable, it would be massive
>error log time, trees in a forest etc.

Understood.

>
>If there are no files left to purge, I presume there is no benefit to
>running vrepair with purge on.
>

Correct

>But since I seem to be getting abends running Dos/Windows purge, would the
>vrepair with purge do better at cleaning them out ?
>
>

Yes -- (at least this has worked for me when I was getting abends running purge from the
DOS prompt).

Stuart Robinson

unread,

Dec 9, 2000, 4:13:00 PM12/9/00

to

> Yes -- (at least this has worked for me when I was getting abends
> running purge from the
> DOS prompt).

Ok, ta.

This server is at sp6a, do you know if the bug that causes the server
abend been fixed in later service packs ?

Stuart.

Barry Schnur

unread,

Dec 9, 2000, 7:05:45 PM12/9/00

to

>
>This server is at sp6a, do you know if the bug that causes the server
>abend been fixed in later service packs ?
>

I don't believe it has.

Stuart Robinson

unread,

Dec 10, 2000, 4:22:00 AM12/10/00

to

Its a dilemma then, I need to run purge now and again in order to free
disk space or the server crashes, but if I run purge it could crash
anyway. Both could cause damage to the volume.

Is the bug to do with purge itself, or only when running large purges ?

It occurs to me that I may need to spend some time setting dirs to purge
immediate, but I dont want to do that if the bug could cause an Abend
whilst people are using it.

Stuart.

Barry Schnur

unread,

Dec 10, 2000, 1:20:38 PM12/10/00

to

Hmm, what type of drive/controller here?

Stuart Robinson

unread,

Dec 10, 2000, 4:32:00 PM12/10/00

to

Adaptec Array 1000 RAID.

Stuart.

Barry Schnur

unread,

Dec 10, 2000, 6:40:38 PM12/10/00

to

Ah -- how many drives in this array -- what make/model of drives....

Stuart Robinson

unread,

Dec 11, 2000, 4:29:00 AM12/11/00

to

No idea of the drive makes, and you have to down the server to find out.

I know there are two arrays, SYS is on one RAID 5 array of 3x4G and the
data volume is a single 8G SCSI disk.

Stuart.

Barry Schnur

unread,

Dec 11, 2000, 1:04:38 PM12/11/00

to

>I know there are two arrays, SYS is on one RAID 5 array of 3x4G and the
>data volume is a single 8G SCSI disk.
>

OK -- these are older drives then (4G).

I'm really not sure but suspect the problem might be at the drive level as the only time
I've encountered this sort of problem, either the drives were bad, or they needed a BIOS
update (IBM DNES drives).

Stuart Robinson

unread,

Dec 11, 2000, 4:10:00 PM12/11/00

to

> I'm really not sure but suspect the problem might be at the drive level

Not sure I see why it should be drive problem, as far as the drive is
concerned all thats happening is a particular area of the disk (FAT) is
being written/read.

I have decided to leave the drive as is at the moment, I can tolerate the
space used by purgeable files and the customer is likely to upgrade to 5.1
in the next few weeks.

Stuart.

Barry Schnur

unread,

Dec 11, 2000, 10:46:47 PM12/11/00

to

Well the thing is, the ONLY time I've seen something like this abend is in the situations
I described.

Not sure why you are the special one here (and I know this is NOT ;,0AA"&t of 'special'
you want to be).

Stuart Robinson

unread,

Dec 12, 2000, 7:00:00 AM12/12/00

to

Ok, perhaps worth checking if the disk driver has been upgraded.

Stuart.

Shaun Pond

unread,

Dec 12, 2000, 8:54:02 AM12/12/00

to

Barry,
there's also the purge that comes with toolbox...

Shaun Pond
Novell Support Connection SysOp

Barry Schnur

unread,

Dec 12, 2000, 11:39:20 AM12/12/00

to

>there's also the purge that comes with toolbox...
>

Good point -- that one runs from the console prompt without using Vrepair, right?

Stuart Robinson

unread,

Dec 12, 2000, 12:36:00 PM12/12/00

to

> Barry,
> there's also the purge that comes with toolbox...

Does that one also crash the server ?

Stuart.

Shaun Pond

unread,

Dec 12, 2000, 1:30:30 PM12/12/00

to

Barry,
yep

Shaun Pond

unread,

Dec 12, 2000, 1:30:33 PM12/12/00

to

Stuart,
not in my experience, but then I've not had purge from any other source
Abend the server :-( It has the advantage that it's quicker than doing
it from a workstation, and you can choose how old deleted files should
be before they're purged, plus you don't have to dismount the volume as
you would with a VREPAIR.

Stuart Robinson

unread,

Dec 12, 2000, 6:19:00 PM12/12/00

to

ta

Stuart.

Barry Schnur

unread,

Dec 13, 2000, 12:10:37 AM12/13/00

to

Not that I know of -- but again, your situation appears quite unusual...

Stuart Robinson

unread,

Dec 13, 2000, 4:33:00 AM12/13/00

to

ta

Stuart.

Tony Locke

unread,

Dec 14, 2000, 3:53:09 AM12/14/00

to

Does this card have more than one SCSI channel?
I had a similar problem if I ran the chain of channel C instead of channel
A.
Also I use to have crashes with Vrepair when I had the volume corrupted by a
macintosh desktop virus.
I had to restore from an uninfected tape in the end to get around it.

Make sure your Adapt
"Stuart Robinson" <stu...@no.spam.please> wrote in message
news:memo.20001210...@srnet.compulink.co.uk...

Stuart Robinson

unread,

Dec 14, 2000, 7:52:00 AM12/14/00

to

Yes, channel A & B.

3 disks in a RAID 5 array on Channel A, and a single disk on channel b.

Its the Channel A disks (SYS volume) that give the problem.

Stuart.

Tony Locke

unread,

Dec 15, 2000, 4:19:21 AM12/15/00

to

OK then it's not the problem I had.
I always turn on immediate purge on volumes on my systems as it saves having
to runa purge.
Are U using .HAM or .DSK drivers (sorry this may have already been covered
in the thread)

"Stuart Robinson" <stu...@no.spam.please> wrote in message

news:memo.20001214...@srnet.compulink.co.uk...

Stuart Robinson

unread,

Dec 15, 2000, 8:29:00 AM12/15/00

to

.dsk

Stuart.

Tony Locke

unread,

Dec 15, 2000, 9:30:12 PM12/15/00

to

I've had good results with the .ham drivers...

from the modules command on my 4.11 sp8a, hotfix3

CDA1000H.HAM
Adaptec CDA1000 RAID Host Adapter (Build 30)
Version 4.20 April 29, 1999

"Stuart Robinson" <stu...@no.spam.please> wrote in message

news:memo.20001215...@srnet.compulink.co.uk...
> .dsk
>
> Stuart.

Stuart Robinson

unread,

Dec 16, 2000, 4:29:00 PM12/16/00

to

Certainly worth considering, the customer is likley upgrading to 5.1 very
soon, and its possible the older driver has some 'issues' working with 5.1

Stuart.