Hello all,

I am looking for pointers regarding using ODM (on Openshift) to serve:

• not only: online workloads (incoming requests via REST API calls),
• but also: batches of requests (thousands)  that are part of a larger SSIS-managed batch processing workflow.

The Requests would be "passed" to ODM via Relational database table(s)

The Outcome from ODM would be expected to also be persisted in Relational database table(s)

Questions:

1. Are there any suggested approaches or best practices how to (not) go about doing that?

2. Has something similar been done done in any customers in real-life projects and worked well?

3. Are there ways to isolate in the ODM setup the serving of the online and the batch workloads?

Would very much appreciate any feedback on the above.

Our context is a real-life core banking transformation project.

P.S. This is one of the very few items of published content I was able to find on ODM batch processing in general.

(It looks interesting but it's too condense for a team without deep ODM skills like ours)

file:///C:/Users/071950726/Downloads/C1000-010-demo-1.pdf (question number 3)

John Spanoudakis

jsanoudakis@gr.ibm.com