Check point

[sureshgane]

I am doing an import.But when I am checking the alert log files I am getting the following message.But the import completes successfully.Has anyone come across like this. what is the reason for this message & how we can rectify this.All suggestions are appreciated
Thanks
G.Suresh
Thread 1 cannot allocate new log, sequence 68335
Checkpoint not complete
Current log# 3 seq# 68334 mem# 0: O:\ORACLE816\ORADATA\TEST\REDO01.LOG
Thread 1 advanced to log sequence 68335
Current log# 1 seq# 68335 mem# 0: O:\ORACLE816\ORADATA\TEST\REDO03.LOG

[sureshgane]

Actually I am having 3 redolog files of 1024K each.If I add another file or if i resize to 10M can I expect any change.Can we have different size of redolog files.Which is the optimum size to create & how many
Thanks
G.Suresh

[Halo]

This problem takes place when the redo log files are small in size. Personally I think that 1M is too small, you should increase the size. Remember that a lot of redo is generated during an import, causing a lot of log switches. There is no correct size of these files, typically, you would size them in a way that there are not more than 3-4 log switches per hour at peak times. You can have different sizes for these files, but this is somethng that is not done usually. I have 4 groups of redo log files with 2 members each of size 250M each. If your db is for OLTP, you need not have large redo log files, on the other hand for Batch processing you probably would need large files. My db is mainly OLTP with a batch that runs for about 10-15 min everyday, generating about 200M+ of redo.

[Edited by Halo on 11-27-2000 at 05:59 PM]

[sureshgane]

Thanks!!!!!!!!!! Halo