Hi

I'm moving from WIN db2 7.x to Linux db2 8.x.

I thought I had a great idea for moving the data from the win server to the linux server in a quick and dirty manner.

Export all the tables as ixf on the win server, create the database on the Linux server, connect from the win server to the linux server, and import the ixf files.

This works fine with one exception :

I have in some databases a few tables containig lobs, and they can have up to 2.5 million rows.

The first table I got stuck on had only 230.000 rows, and the problem is when I use the following statement :

db2 export to ARTICLE.ixf of ixf lobs to .\lobsARTICLE0\ modified by lobsinfile select * from ARTICLE

The export stops after 1000 rows because the 1000 first rows have exhausted the unique filename scheme in the .\lobsARTICLE0 folder.

I can change the command to

db2 export to ARTICLE.ixf of ixf lobs to .\lobsARTICLE0\ lobfile file0, file1
modified by lobsinfile select * from ARTICLE

Now the export stops after 2000 rows because the 2000 first rows have exhausted the unique filename scheme in the .\lobsARTICLE0 folder. Now I have given two seeds to the filename scheme.

BUT I have 230.000 rows, so in theory I have to add 231 seeds to the lobfile directive. And what with my 2.5 million rows table?

There must be a way to run the export command so that the export figures out how to generate filenames/storeage for the lobfiles regardless of number of rows, but I can't figure it out.

Any help?


DB2 & UDB email list listserv db2-l LazyDBA home page