Can't delete NDS users

[rsebasti]

I am having problems deleting objects (users and nal apps) on NDS - NWAdmin keeps telling me Directory Services still hasn't finished moving it or one of it's subordinate objects.

Any ideas... ========================================
I learnt the most
when I stopped talking...

 

[TheLad]

Is your NDS properly synchronised? Check in NDS Manager. -----------------------------------------------------
"It's true, its damn true!"
-----------------------------------------------------

 

[salserocito]

Run DSREPAIR!!! That is, after verifying that you have successful synchronization!

Try doing a "Full Unattended Repair" of Directory Services. Keep running it until you see no more errors!

Good Luck!

 

[rsebasti]

Been doing that - I have this server that's having the high ncp packet problem. And it's the same one having the errors in NDS manager.

So you recon kep running dsrepair - when I did full unattended to it - I got 355 errors (last time). Will try again. ========================================
I learnt the most
when I stopped talking...

 

[salserocito]

Continue running it until you have NO MORE ERRORS!

Good Luck!

 

[rsebasti]

Thanx... would this be a reason for high NCP traffic between servers? ========================================
I learnt the most
when I stopped talking...

 

[rsebasti]

The errors are still on 355 no matter what I do.
I did an external reference check and found errors as well.
But the server timesync's correctly.

Using NDS manager I found yellow stops signs when I check synchronization on all partitions.

Heeelp!! ========================================
I learnt the most
when I stopped talking...

 

[TheLad]

Do you run NDS for NT? -----------------------------------------------------
"It's true, its damn true!"
-----------------------------------------------------

 

[rsebasti]

No NDS for NT.

Here's what I have found so far - checking NDS Manager there is one partition having problems in synchronization. It's master replica is held in the main office. The read/write replicas for it are held on another server in the main office and the target problem server.

Using Partition continuity check the problem jumps between the replica on the problem server and the other server in the main office (not the replica on the main server) producing 625 errors.

Using external reference checks in DSrepair for the problem server and the main server - errors are produced. They seem to be backlink obituaries. Which is apparently NDS object moves or renaming not synced/replicated between replicas.

Is there a way to purge these issues and force sync of the replica? ========================================
I learnt the most
when I stopped talking...

 

[TheLad]

See if the following TID helps:

http://support.novell.com/cgi-bin/search/searchtid.cgi?/10056268.htm

Otherwise, check out this one:

http://support.novell.com/cgi-bin/search/searchtid.cgi?/10013126.htm

-----------------------------------------------------
"It's true, its damn true!"
-----------------------------------------------------

 

[rsebasti]

Well after a close look it seems that the branch office server holding a replica of the main office partition has a slow link.

Is there a way to remove this partition cleanly (which is obviously trying to replicate through the link) without making sure that synchronization is updated (which can't) anyway.

========================================
I learnt the most
when I stopped talking...

 

[salserocito]

What size is the Circuit? I'm surprised that DS is having problems replicating considering that it should be happening over TCPIP.

Some of our many remote sites are linked to Corporate via 64K Frame-Relay circuits with 32K CIR. SInce we have upgraded to NetWare v5.1 from v4.11, we have seen a definite increase in performance now that TCPIP is the primary protocol.

Make sure that you have SLP configured on ALL of your NetWare v5.x Servers. You can verify its functionality by typing... DISPLAY SLP SERVICES. This will let you know what resources are available over TCPIP.

Good Luck!

 

[rsebasti]

Did so found SLP running.

I still need a way to remove a problem replica off a server.

========================================
I learnt the most
when I stopped talking...

 

[salserocito]

You have to run DSREPAIR and go to ADVANCED OPTIONS MENU / REPLICA AND PARTITION OPERATIONS. Here you can remove the Replica from the Replica rings.

Good Luck!

 

[rsebasti]

Won't this need a special DSREPAIR switch?

And after removing it what do you run to make sure it worked fine? Unattended repair? ========================================
I learnt the most
when I stopped talking...

 

[salserocito]

Try using DSREPAIR -xk2

As for testing, once its gone, its gone. If you are receiving "Unable to communicate with Server XXXX" messages on the Servers that were in its replica ring, then you haven't removed everything.

Remember, not only do you have to remove the replica and partition info from the Server in question, but every other Server that was in its replica ring!

Good Luck!