-618 Sync error NDS replication

[artesianwater]

We Have Three (3) Netware 5.1 servers. One of which holds the master and the other two which have read/write replicas. All of a sudden we receive -618 sync errors on the two servers holding the read/write replicas. I have run dsrepair -a and unattended full repair and it reports "Invalid Id'd - Removed from list" and it won't correct the database. I even ran "Repair local database" without success. I have found out that the three errors refer to Bindery ojects that are "Inconsistent". I tried to delete them using nwadmn, ConsoleOne, and Dsbrowse but keep getting -601 erros, and the objects won't delete. Does anyone have any ideas??? I am at the end of my rope on this one..Thanks, Ted

 

[Lou0686]

Hi,

In most cases this is an indication that the object has lost it's attributes in one or more of the replicas.

What do you see in NWadmin? Sometimes a question mark (?) and at other times it seems ok? When this occurs, you are attaching to different replicas at different times.

What you will need to do is determine which replica has the good object. There a few ways to test this.

1. Go to 2 of the servers and disable login. Then Launch NWadmin, the object looks good, that servers replica is ok. Do the same until you have lauched NWadmin attaching to all servers.

or you could

2. Do a DSbrowse on all three servers and browse to the object in question. The server with the correct object will have all it's attributes and show up properly, the server(s) that has the corrupted object will not show up properly.

Once you determine which server has the correct object, load dsrepair from the server and perform the steps in Novell TID 10067667.

Hope this helps

Lou

 

[artesianwater]

Thanks for your information.The ultimate resolution was to have Novell support dial in and "kill" three (3) old & damaged bindery objects. We couldn't delete them in NDS or using dsbrowse. They used dsdump, a hex editor. We then removed the R/W replicas using dsrepair -xk2 xk3 and added them back in. All is now ok. Thanks, Ted