August 18, 2005 at 1:10 am
Hi,
I'm new to sqlserver but i've to get skilled very soon as my compagny is going to use MS sql server as core rdbms. At this time we're using Oracle rdb and Oracle rdbms as core rdbms systems.
My question is:
Does someone have experience in replicating rdb to sql server and visa versa? We have to do this for a limited time before our final migration.
next question:
Does anyone have experience or tips for me how we migrate Oracle rdb to MS SqlServer?
i hope someone can help.
In advance thanks.
Michel
August 18, 2005 at 4:54 am
Have a look at the sample chap in my sig. It explains how to replicate to Oracle. To replicate to SQL from Oracle I believe you need the Oracle transparent gateway.
You can use DTS to migrate simple Oracle databases to SQL.
--
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
August 19, 2005 at 9:13 am
If you are migrating just data, then set up a linked server and use scheduled jobs to migrate the data with DTS, as Hilary indicated.
I had to do homegrown replication with flat files as an intermediate because I was not allowed to install Oracle drivers on our production SQL Server. It wasn't too bad once I developed the pattern and it has worked out well. It also gives us a good audit step.
If you also have to migrate procedures and functions, then that may take some time, though there are some commercial products/services that promise to help. I haven't used any of those.
(I used to use Rdb, but it was before it supported SQL. I forget what the language was called. I hope you're using SQL at this point.)
August 19, 2005 at 9:26 am
For what it's worth, it is called RDO (Relational Database Operator) and it still functions with the current releases of RdB.
Being a co-worker of the original poster, I'll have to say that there will be probably no migration. We will (just) have to synch a sqlserver enviroment, and, in time, do it the other way around until we can depart the RdB world.
August 19, 2005 at 9:36 am
If the data is complex (lots of tables and columns) I would recommend a ETL tool like Pervasive's DataIntegrator (used to be DataJunction) or similar. These have good mapping tools and the data movements can be scheduled.
You could start by setting up Oracle to SQL Server movements and then reverse it when you cut over to SQL as the production data store.
If you have few (<25) tables and simple data content, then homegrown extract, transform, load scripts could be cost effective. They were for my situation.
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply