REPLICATION PROBLEM- data coruption... HELP!

[cheezebeetle]

Ok here is the problem... thanks for reading, and thanks for any help... it is
desperately needed.

We are doing a replication from a server in Maryland to one in New Jersey. The
one in Maryland (Server A) is a SQL 2000 server. The subsriber in NJ (Server B)
is version 7. To upgrade to 2000 would be a real pain, so seeing if it is
definately the problem before doing so.

Ok, during the replication using an ODBC connection, the data in one field is
getting corrupted. The table is made of several fields, one of which is a ntext
datatype. This is where the problem happens... just this field.

The snapshot is taken once a day at 2 AM, then the replication takes place. It
does so fine until it gets to a record 524. When it gets to that record on
Server B, it cuts off the data at exactly 1692 characters and then repeats only
that cell with the same cutoff point over and over again throughout the rest of
the field and remaining records.



So after analyzing it, I removed the data from Server A and tried again. It now
worked and continued past 524 until it gets to record 1154, where it repeats
exactly the same problem. Cutting of that data once again at 1692 characters.

Now both tables are the same design, and in fact, if I do a backup from 2000 and
restore it on 7, it is fine. It just gets corrupted during the replication. I
am not sure if I missed something during the set-up process of replication, or if
it is simply a bug with replication from 2000 to 7.

Can anyone help!!?? Solving this is imperative to our daily operations. If you
need any additional info, please do not hesitate to contact me.

Thanks,

Brian Hensley
bhensley@h2xmultimedia.com
410-752-5070

 

[vanoordt]

Could it be that an ODBC interface setting truncates data at 1692 characters and cause the problem?
Check it out.
Good luck!