Need information on Data Extract approach

[nvuradi]

Hi All,



We need to extract the data from OLTP system which is on Mainframe DB2 host

We have two approaches to extract the data from DB2 tables:

1. Using DB2 API Stage which directly fetches the data from DB2 tables.
2. Using JCL scripts to fetch the data from Mainframe and ftp the files to UNIX box. and process those files using a sequential stage.

can anyone suggest which approach is the best one taking into consideration the performance aspect. our jobs will be run daily insert/updating around million records.


Any help will be highly appreciated.

Thank you
Narendar

[keshav0307]

what is your target, where do you want to load the extracted data

[nvuradi]

what is your target, where do you want to load the extracted data

data will be loaded to DB2UDB database.


Thanks,
Narendar

[ArndW]

How many Mb are you looking at? If you use the API then the DataStage server controls when and how to do the job run. If you need to extract & ftp the data then another group is involved and controls if things will run. This would be push instead of pull technology.

[Mallayagari]

IF you have enough time to pull the records from mainframe using API stage then its better to use API stage. If time is critical factor to your processing then extract and ftp the data is the best method. How much time do you have to pull this data?

[quote="nvuradi"]Hi All,



We need to extract the data from OLTP system which is on Mainframe DB2 host

We have two approaches to extract the data from DB2 tables:

1. Using DB2 API Stage which directly fetches the data from DB2 tables.
2. Using JCL scripts to fetch the data from Mainframe and ftp the files to UNIX box. and process those files using a sequential stage.

can anyone suggest which approach is the best one taking into consideration the performance aspect. our jobs will be run daily insert/updating around million records.


Any help will be highly appreciated. :)

Thank you
Narendar[/quote]