08-31-2017 05:58 AM
I have table in Oracle whihc is haiving 120 crore records and i want to create Full extract in DB2 table and In our SAS job it takes 4/4.5 hours to load 4-5 crore records when it is running in LSF/schedular.
Kindly suggest the fastest mode by which the data can be imported to DB2. we have SAS DI studio, datafluxstudio, SAS EG amd using SAS 9.3 .
08-31-2017 06:20 AM
Do it in two steps. Pull the table into SAS, and then load it into DB2. You might suffer from (unnecessary) network congestion when you simultaneously read and write.
BTW, don't expect 1.200.000.000 records to go over in a flash. That IS a lot of data, and your network does have a limited bandwidth.
Calculate your physical data size (nobs x observation size), and relate that to your network bandwidth, to see if you are able to saturate the network. Once the connection is saturated, there's nothing you can do, with the exception of adding network cards and network paths.