Invalid reference to object

While running dsrepairs to troubleshoot a different problem (see the
thread
on -672 in iManager), I have noticed a large amount of "Invalid
reference to
object" error messages.

They seem to be mostly ZFD workstation objects, but are also server's
sometimes.  After running dsrepair multiple times, until no errors,
the
problem returns either on the same or new objects.

Here are some copies of the error(s):


ERROR: Invalid reference to object, ID: 0000896F, value is purged
Attribute 3F828177, Reference
Object ID: 00008904, DN: OU=LCSC.O=REMOTE.T=POST


ERROR: Invalid reference to object, ID: 00008816, value is purged
Attribute 3F8D9160, Reference
Object ID: 0000AE8F, DN:
CN=HAGIESI-00:04:23:25:A1:82.OU=Workstations.OU=NBTX.O=REMOTE.T=POST





0
Keith
10/17/2003 7:21:56 PM
novell.edirectory.netware 7858 articles. 0 followers. Follow

5 Replies
508 Views

Similar Articles

[PageSpeed] 53

Invalid references mean that there's an external reference that got
purged
that didn't finish getting cleaned up - generally not something to
worry
about, but I think in your case they might be a symptom of a larger
issue.
 It would definitely be worthwhile to run a full health check on the
 environment.

im

-- 
 Jim Henderson, Novell Support Connection Volunteer SysOp
 http://support.novell.com/forums
 (Sorry, support is not provided via e-mail) 
 Homepage at http://hendersj.dyndns.org (URL has changed!)




0
Jim
10/18/2003 2:11:52 AM
Keith L. Schneider,

In addition to what my collegue Jim says. You could try running a
local
repair that closes the database on all servers at the same time.
Usually invalid references aren't a problem, but when the keep on
coming back or there are a whole bunch of them, it could be an
indication of another problem.
I was out on a server the other day that came back with 123776 invalid

refs errors. I ended up recreating the replica on that server.


-- 
//Niclas Ekstedt
_______________________________
Niclas Ekstedt, CNE, NSC Sysop
Iftech Network AB
_______________________________



0
Niclas
10/18/2003 8:00:18 AM
Performed the dsrepair on all locals, keeping the databases closed on
each
server.  Cleared the errors for a while, but they came back.

I'm going to do some checking in the ZFD forums, as I think the users
might
not have all the rights needed for updating workstation information.
(TID10052011, TID10026210 , etc.)

My biggest concerns based on the dsrepairs were that a pseduo server
object
was found, although all servers are accounted for and no crashed
servers or
anything else has happened recently.  Also, I was seeing a ambiguous
containment flag error that I couldn't repair which seems to be a
situation
that should only apply to a mixed tree and mine is all 8.6.2 sp4.

Here's a snip of the log:

/***************************************************************************

*/
NetWare 6.00 Directory Services Repair 10250.40, DS 10350.23
Log file for server ".BKUP.bkup.Coleman" in tree "POST"
Repairing Local Database
Start:  Sunday, October 26, 2003   6:12:35 pm Local Time

** All disk amounts are approximations **
->Current available disk space: 2639 MB
->DSRepair may need to use: 0 MB
->Disk space remaining after operation: 2639 MB

Physical Check
Repair Trees - Scan Values
Repair Trees - Sorting Values
Repair Trees - Scan Entries
Repair Trees - Sorting Entries
Repair Trees - Check Values
Repair Trees - Check Entries
Total Objects in Database:       6587
Total Objects in Schema  :       1718
Total External References:        259
Total Objects in Replicas:       4450
Rebuilding operational schema
Schema Check

Schema Class Definition, Object ID: 000006D2, RDN: Partition
ERROR: 'flags' field was changed from 00200013 to 00200033
Ambiguous containment flag
Field 'flags' changed from 00200033 to 00200013


Repairing objects in a replica
Start:  Sunday, October 26, 2003   6:13:01 pm Local Time
[1 of 9] Subordinate: OU=Exec.O=Coleman.T=POST

Total Objects = 116, UNKNOWN class objects = 0, Total Values = 8808
[2 of 9] Read Write : OU=SPO.O=REMOTE.T=POST

Total Objects = 194, UNKNOWN class objects = 0, Total Values = 10886
[3 of 9] Master     : O=REMOTE.T=POST

Total Objects = 1, UNKNOWN class objects = 0, Total Values = 103
[4 of 9] Read Write : T=POST
Total Objects = 205, UNKNOWN class objects = 0, Total Values = 7740
[5 of 9] Read Write : OU=LCSC.O=REMOTE.T=POST

Total Objects = 195, UNKNOWN class objects = 0, Total Values = 12251
[6 of 9] Read Write : OU=NBTX.O=REMOTE.T=POST

Total Objects = 357, UNKNOWN class objects = 0, Total Values = 24584
[7 of 9] Read Write : OU=MAIZE.O=REMOTE.T=POST

Total Objects = 249, UNKNOWN class objects = 0, Total Values = 19179
[8 of 9] Read Write : OU=BVAR.O=REMOTE.T=POST

Total Objects = 162, UNKNOWN class objects = 0, Total Values = 12444
[9 of 9] Read Write : O=Coleman.T=POST

Total Objects = 3086, UNKNOWN class objects = 0, Total Values = 225089


Checking local references
Start:  Sunday, October 26, 2003   6:13:58 pm Local Time
Repairing single object:

Object ID: 0000805D, [Pseudo Server]

Total Objects = 1, UNKNOWN class objects = 0, Total Values = 44
Finish:  Sunday, October 26, 2003   6:14:02 pm Local Time
Total repair time: 0:01:27
Checking mail directories
Checking stream syntax files
Repair process completed, total errors found = 1
"Niclas Ekstedt" <niclas.ekstedt@nospam.iftech.se> wrote in message
news:mq6kb.4815$S13.1604@prv-forum3.provo.novell.com...
> Keith L. Schneider,
>
> In addition to what my collegue Jim says. You could try running a
local
> repair that closes the database on all servers at the same time.
> Usually invalid references aren't a problem, but when the keep on
> coming back or there are a whole bunch of them, it could be an
> indication of another problem.
> I was out on a server the other day that came back with 123776
invalid
> refs errors. I ended up recreating the replica on that server.
>
>
> -- 
> //Niclas Ekstedt
> _______________________________
> Niclas Ekstedt, CNE, NSC Sysop
> Iftech Network AB
> _______________________________





0
Keith
10/27/2003 5:27:31 PM
Keith,

If you find out anything about this, please let me know.  I am running
eDirectory 8.7.0.3, and have been having this problem for a while.  I
too locked all databases and did the repair with the same failed
success.  However, my reference problems are not limited to ZEN
objects only.  I can be sympathetic to your problem, but
unfortunately, I can not offer a solution since even Novell has not
yet figured out my problem.   I have sent them numerous copies of my
DIBs and DSRepair Logs.

Have Fun,

Bob



0
Bob
10/27/2003 7:52:05 PM
Bob Stumpp,

> If you find out anything about this, please let me know.  I am
> running eDirectory 8.7.0.3, and have been having this problem for a
> while.  I too locked all databases and did the repair with the same
> failed success.  However, my reference problems are not limited to
> ZEN objects only.  I can be sympathetic to your problem, but
> unfortunately, I can not offer a solution since even Novell has not
> yet figured out my problem.   I have sent them numerous copies of my

> DIBs and DSRepair Logs.

When they come back with anything, please let us know.
I wonder if upgrading to eDir 8.7.1 and then run a repair on all
servers might fix this.


-- 
//Niclas Ekstedt
_______________________________
Niclas Ekstedt, CNE, NSC Sysop
Iftech Network AB
_______________________________



0
Niclas
10/28/2003 2:45:56 PM
Reply: