its very slow now.

[spicysudhi]

Hi

oracle 8i on win2k server. everything was fine upto last week. no changes made in the week just passed. but database suddenly runnning like a ..., very very slow. diskspace is there, server cpu usage is looks normal (well below 2%).

every insert/update/delete stateemnts taking time, select is fine.

please suggest me, how to resolve this. I am using free Toad, in that no RED buttons, all looks ok.

i know u may need some more info, but i dont know wht u may require to give some answer.

thanks & regards,
sudhi

 

[aryeh1010]

Have you had any large data loads between the time the database was running well and now?

Aryeh Keefe

 

[carp]

And you're sure no changes have been made? No indexes disabled or changed and no triggers added?

 

[spicysudhi]

no index changes/triggers.

data load - its same as a week before. everyday nearly 200MB csv file refreshes some tables.

regards,
sudhi

 

[JtheRipper]

Hi there,

Just an idea....if you use the cost based optimizer you might have to update your database statistics.

How many redo logs do you have and what are their sizes? Is the load not getting too big (deletes and updates) so that the lgwr is looking for the next group but it has not been archived yet (if archivelog mode)?

J.

 

[sem]

Haven't you installed/launched antivirus?

Regards, Dima

http://seminihin.narod.ru

 

[spicysudhi]

I have this message in the alert log.

Thu Apr 15 14:29:01 2004
Current log# 1 seq# 1596682 mem# 0: C:\ORACLE\ORADATA\SCM\REDO01.LOG
Thread 1 cannot allocate new log, sequence 1596683
Checkpoint not complete
Current log# 1 seq# 1596682 mem# 0: C:\ORACLE\ORADATA\SCM\REDO01.LOG
Thread 1 advanced to log sequence 1596683
Thu Apr 15 14:29:03 2004
ARC0: media recovery disabled
Thu Apr 15 14:29:03 2004
Current log# 2 seq# 1596683 mem# 0: C:\ORACLE\ORADATA\SCM\REDO02.LOG
Thu Apr 15 14:29:03 2004
Errors in file c:\oracle\admin\SCM\udump\ORA03956.TRC:


antivirus installed on the server.
have 3 redo logs.

regards,
sudhi

 

[JtheRipper]

Hi there,

You receive the error because Oracle wants to use a redo log group that has not been checkpointed yet. Best is to add a few more redo groups to your database or to increase the size of your redo logs to make sure that you do not run into this problem. If I were you I would increase the size of my redo files and add a few more redo groups. If you are in archivelog mode, make sure automatic archival s enabled.

Hope it helps,
J.