DBAsupport.com Forums - Powered by vBulletin
Results 1 to 2 of 2

Thread: Problem with restoring data using 'imp' command.

  1. #1
    Join Date
    Sep 2005
    Posts
    1

    Problem with restoring data using 'imp' command.

    Hi,
    I developed an application using VB 6.0 and Oracle 8i and implemented with Windows 2000 Advanced Server in a small hospital in 2003. I took time-to-time backup with 'exp' command in command prompt. Now suddenly the Windows 2000 Adv Server is corrupted and when I tried to restore in another hard disk, the data is not transferred to the tables in the new hard disk. Moreover, I think the buffer size in the previous one is 4096 and the new one is 8192( which is taeken by default). Can anyone help me restoring the data.

  2. #2
    Join Date
    Nov 2000
    Location
    Pittsburgh, PA
    Posts
    4,166
    I'm sure by now you know that an export is not a true backup. Going forward you should look into using RMAN, provided that you upgrade to a supported version. Such as 11gR2. In order to do the restore you need a new database, it is ok if you use an 8K block size rather than the 4k block size that you used before. You need to make sure that the schema's and tablespaces exist before you do the import. If the directory where you are putting the data files is different it helps to have the tablespaces created ahead of time. Creating the users/schema's ahead of time allow you to more carefully specify default tablespace and temporary tablespace. It will also help to have the listener configured and running ahead of time.

    When you do the import you want to specify most of the following:

    imp system/@ file= log= fromuser=,,etc touser=,,etc buffer=102400000 ignore=y commit=n statistics=none

    You may need to treak the above command and everything in <> needs to be values that you supply. You really only need to specify the buffer command if you use clob, blob, long, long raw or raw. it doesn't hurt to add it if you aren't sure. Ignore=y tells the import that it is ok if things already exist. Statistics=none might prevent you from getting warning messages about having bad statistics. commit=n means that it doesn't incrementally commit while it is importing. It is better to specify individual schemas, rather than doing full=y. Because when you do full=y you will get data dictionary information, which you don't want.

    Good luck and if you get errors post the log file.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


Click Here to Expand Forum to Full Width