Help, How should I ARCHIVE???

Situation:

 Currently, I have a table, MAIN, which contains over 3 million rows  which is over 1 GIG. I need to be able to have a monthly job which  will archive the oldest month of data, but, I must be capable of  restoring any given month's data in a timely manner (say, 2 hours).  Here are a couple of alternatives that I have come up with. Your  feedback will be greatly appreciated.

NOTE: The names of tables have been changed to protect the innocent.

Method 1:

 To archive data:
  Create a temporary ARCHIVE table and insert one month of data from   the MAIN table. Then export this ARCHIVE table and backup the   expdata.dmp file to tape. Finally, delete the correponding data   in both ARCHIVE and MAIN.

 To restore data:
  Retrieve expdata.dmp from tape. Import expdata.dmp into ARCHIVE   table then insert data back into MAIN table via SQL*Plus script.

Method 2:

 To archive data:
  Create (spool) a delimited file (myfile.dat) of one month of data from   the MAIN table via SQL*Plus script. Backup myfile.dat to tape.

 To restore data:
  Retrieve myfile.dat from tape. Using SQL*Loader, append this data   back into MAIN table.

Method 3:
  <<your suggestions go here>>
  No, buying more disk space in not an alternative.

Please provide any pro's/con's of these methods.

Thanks in advance! Received on Fri Mar 11 1994 - 22:26:34 CET