commit transaction

[chulett]

I have also seen APT_RDBMS_COMMIT_ROWS environment variable in the below link.

http://etl-tools.info/infosphere-datast ... iables.htm

I don't see this variable in my project. !!! ???

Add it.

[chulett]

I have tried to learn about ODBC stage and its option (like Insert Array Size, Fetch Array Size, Row Commit Interval etc.,). But i don't see this stage in any manual. If anybody knows where i can found more info on this, please let me know.

Meanwhile can you explain what is the difference between 'Insert Array Size' and 'Row Commit Interval' ?

Check the Connectivity Guide for ODBC pdf.

Short version: Array Size is the number of requests sent to (or retrieved from) the database across the network at once. Commit Interval is exactly what it sounds like, at what row count (how often) to issue a commit.

[karthi_gana]

I have also seen APT_RDBMS_COMMIT_ROWS environment variable in the below link.

http://etl-tools.info/infosphere-datast ... iables.htm

I don't see this variable in my project. !!! ???

Add it.

ok

i thought this will be available by default.

[karthi_gana]

add it

Assume i have added it by using administrator tool.

Assume i have also added that variable in my job. But my doubt is how it will get affect while DB insert operation. will adding the variable is enough for commit ? i am little bit confused.

There is a logic in Insert Array Size and Row Commit property. But adding the variable will take care the committ part ????

[karthi_gana]

any inputs pls?

[Mike]

You can find the description for APT_RDBMS_COMMIT_ROWS in the Parallel Job Advanced Developer's Guide.

Other than that, my advice is "Prototype it". We can learn more by doing than by simply reading (IMHO).

Mike