hello,

I have tested to import around 10,000,000 delimited flat file(about 1G)
into DB2 UDB 8.2 in AIX machine. I have 4 WAY and 8GB AIX box (no
application running this time. it's a development machine)

During the import, I didn't modify the bufferpool size and I had to
cancel half way because it was taking way too long.

My question is will adding say lots of bufferpool help in this case? I
am using commitcount of 5000? Will commitcount slows it down
significantly? Nevermind the disk layout (I only have 2 x 73.4 disks),
what can possibly be faster way to import the 10 million rows into the
database? In other words, how do I maximize the import and reduce the
import time?

if anyone can help mention the parameters that affect the import, I
will apprecite much. Will the same parameters apply to DB2 Load?

Db2 Novice

DB2 Novice wrote:
> hello,
>
> I have tested to import around 10,000,000 delimited flat file(about 1G)
> into DB2 UDB 8.2 in AIX machine. I have 4 WAY and 8GB AIX box (no
> application running this time. it's a development machine)
>
> During the import, I didn't modify the bufferpool size and I had to
> cancel half way because it was taking way too long.

Read the Data Movement & Utilities Guide. This is discussed.

Short answer: Use LOAD.