DSXchange

Nope. At least I really don't believe so.

Does your performance hit seem to be related to hashed files reading and/or writing?

Also known as Runtime Column Propagation which is (as noted) what you need to fulfill this requirement. It's all documented plus has been discussed here ad nauseam so there's plenty of information on it to go around.

ray.wurlod wrote:CONVERT is not available as an SQL function, but you can include it in an EVAL expression.

Interesting. Thanks for the clarification.

That makes more sense to me. I didn't think "binary" could support the implied decimal or scale but had already been schooled once so didn't want to get whacked again.

Why would you think CONVERT would work here? You're mixing... what's it called... UniVerse RetrieVe / SQL syntax with a DataStage function. Sorry, not sure what the answer is off the top of my head but that's not it. I think Rocket Software has all of the old UniVerse manuals available for download,...

Well then... never mind.

Trying to recall - COMP and COMP-1... Sam Ting?

I wasn't confused as to what "CDC" stands for and am well aware of the change code of which you speak. I just don't see how you can be using it and allegedly "re-load by only inserting". I'm guessing you meant that you switched from "insert else update" to separate acti...

Thanks for the update. Certainly, I have not yet clear the relationship between the two environment variables $APT and Connector property "number of row per commit". Finally, I do not understand why this feature is not clearly documented in the IBM manuals or on Redbook. Questions for your...

Thanks for the update. Certainly, I have not yet clear the relationship between the two environment variables $APT and Connector property "number of row per commit". Finally, I do not understand why this feature is not clearly documented in the IBM manuals or on Redbook. Questions for your...

ray.wurlod wrote:suggested @DAY usage

That makes more sense as the stage (for whatever reason) doesn't support that option.

Just a quick fyi - doing a "select *" in a tool that is metadata driven is generally considered a Bad Idea.

Me, I'd still write a routine. [shrug] So - what's your WORKAROUND? In my mind there's nothing broken here that needs to be "worked around" so it's either resolved or it isn't. The fact that there are multiple resolutions and you may not have implemented the one you wanted doesn't make it ...

satheesh_color wrote:re-load by only inserting the records using cdc code.

Can you explain exactly what this means, please. What does 'using the cdc code' have to do with doing pure inserts?

Sounds like a plan! So... resolved?