Reputation: 2077
I have a 185GB data dump in 600 IXF files ranging in size from 4K to 50GB. The import command has so far done 30GB in 10hrs. I'm trying to speed up the import process and it seems the way to go is:
db2move newdb load
The problem is that I do not have the original schema so I cannot create the tables first before doing the load. Is there a way to hack the load command into creating the tables?
Upvotes: 0
Views: 2226
Reputation: 11042
You can't do this using the db2move
command; however, you can use the IMPORT
command with the ROWCOUNT
option to create each table and load only a single row into each table. Use the db2move.lst
file to determine which IXF file corresponds to each table.
Once you have created the tables with the IMPORT commands, then use db2move newdb load
to then replace all data you've already inserted via db2move import
:
For example, assuming tab1.ixf
corresponds to the table DB2INST1.TAB1
:
db2 "import from tab1.ixf of ixf rowcount 1 create into db2inst1.tab1"
# repeat for additional tables
db2move newdb load -lo replace
Upvotes: 1