Hello Kevin,

We have many customer replicating data to relational databases (Oracle,
mySQL, SQL Server and DB2).

Depending on the skill sets the customer wants to utilize will determine
the product that they will use. Most use our FusionWare Direct mv2SQL as
they did not want to maintain an application to move the data.

Of course if .Net skills are available they use our FusionWare Direct
ADO.Net Provider.

We have some users that wanted to use Java/XML. 

Most of the users have the data timed and dated and run a phantom
process to look for changed data this data is then moved to the required
relational database. Some of the users had to put a process in place
which the other emails have covered.

As Tony mentioned it would be a pleasure to work with you to achieve
your project goals.


Best Regards,

Janet Bond
FusionWare Corporation
1.866.266.2326 x159 [ad]
 

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
[EMAIL PROTECTED]
Sent: Sunday, July 30, 2006 4:56 PM
To: [email protected]
Subject: RE: [U2] Replication Between UniData and MySQL [ad]

Steve O'Neal wrote:
> Proposing three alternatives:
> 1) UniData Triggers that call a UniData EDA (Extended Data
> Architecture) interface to write to MySQL.  The only thing
that has
> to be developed is the MySQL interface for EDA.  The
burden here is
> adding triggers for the UDT database. Hummm....

Am I right in thinking EDA is OFS++?  I don't think Kevin
will have access to that in UniData 6.0, and I can't see him
wanting to write and link in an OFS driver for MySQL.  What
about IBM Lab Services - would they write an OFS driver when
the technology is *very* deprecated?

> 2) UDT Triggers that feed IBM's DataStage to perform the
ETL
> (Extraction of the data, Transformation of the data (ABC
on the UDT
> system = 123 on MySQL), and Load). The burden here is
adding triggers
> for the UDT database. Hummm....

Sledgehammer, nut? ...

> 3) using the "WRITE Subroutine" methodology that updates
both
> the UniData
> DB and the MySQL via EDA.  The burden here is changing all
of
> the writes to call subroutines. Hummm....

Again, I'm having trouble with the juxtaposition of the
components EDA and UniData 6.0 in this solution.

> A lot of work no matter what the alternative is.
> 
> Rephrasing Tony...
> U2 Lab Services would be honored to work with you to
provide
> a complete solution.

Adding triggers for the appropriate files shouldn't be a big
problem once you've worked out what needs to go across. 
Normalising and massaging that data on it's way across could
be fun though.

I still think the option of pushing the data into MySQL
through ODBC and BCI (not UCI which is the term I mentioned
the other day), or pulling it out from the windows end via
an ODBC schema on your interrim tables make fairly simple,
understandable and testable solution options.

I also suspect IBM Lab Services would help you with these
too, if you wanted them to! ;^)

Cheers,

Ken
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to