If you have an old AIX IBM Server you will know, there is a limit of the size of file you can have, what do you do if the database is much bigger than the limit. In the example below we will split the file to 1gb chunks.
———– Export Section
L1:/home5/dump > mknod /tmp/split_pipe p
L1:/home5/dump > mknod /tmp/exp.pipe p
mknod: /tmp/exp.pipe: Do not specify an existing file.
L1:/home5/dump > nohup split -b1000m < /tmp/split_pipe > /home5/dump/L1_dmp.dmp.Z &
L1:/home5/dump > nohup compress < /tmp/exp.pipe > /tmp/split_pipe &
L1:/home5/dump > nohup exp system/12345 file=/tmp/exp.pipe full=Y log=/home5/dump/exp_live.txt &
L1:/home5/dump > Sending nohup output to nohup.out.
—————— IMPORT Section
# The import script assumes in this example that the above export script created 2 split files
# called expfileaa and expfileab. The order of the file for the cat command is very important.
Copy the files over to the Linux machine.
mknod /tmp/split_pipe p
cat /cima1/backup/export/infobase/xaa /cima1/backup/export/infobase/xab > /cima1/backup/export/infobase/L1.dmp.Z
SQL> CREATE DIRECTORY aix AS ‘/backup/export/aix’;
SQL> GRANT read, write ON DIRECTORY aix TO starsky;
imp startsky/12345 full=y file=L1.dmp show=y log=import.sql
Well you have exported a database in AIX by splitting the file, thereafter moving the files to linux and importing them into a Oracle 11g Database.