Re: Alert log message question

This problem happens because the LGWR and ARCH is complete for the same log file. The ARCH had not finished archiving the files but the LGWR already had to use that particular log files. Until ARCH release that file, LGWR cannot get hold of the new log file. So that is why the checkpoint is not complete.

Try to make the logfile bigger in size. It should help!

Winnie Liu

shermanej_at_eccic.com wrote in message <6p7h75$op8$1_at_nnrp1.dejanews.com>...
>I sometimes see the following in my alert log after batch jobs have been
>run:
>
>
>Sun Jul 19 15:07:14 1998
>Thread 1 cannot allocate new log, sequence 19514
>Checkpoint not complete
> Current log# 1 seq# 19513 mem# 0: /oracle4/dbs73/PROD/redoPROD07a.log
> Current log# 1 seq# 19513 mem# 1: /oracle1/dbs73/PROD/redoPROD07b.log
> .
> .
> .
>Wed Jul 22 01:06:29 1998
>Thread 1 cannot allocate new log, sequence 19590
>Checkpoint not complete
> Current log# 9 seq# 19589 mem# 0: /oracle4/dbs73/PROD/redoPROD12a.log
> Current log# 9 seq# 19589 mem# 1: /oracle1/dbs73/PROD/redoPROD12b.log
> .
> .
> .
>Thu Jul 23 05:38:55 1998
>Thread 1 cannot allocate new log, sequence 19621
>Checkpoint not complete
> Current log# 4 seq# 19620 mem# 0: /oracle2/dbs73/PROD/redoPROD1a.log
> Current log# 4 seq# 19620 mem# 1: /oracle1/dbs73/PROD/redoPROD1b.log
>
>
>
>
>Log switches occur about 1 minute apart during these batch jobs and 20-30
>minutes apart during normal database activity.
>
>We tried creating another pair of redo log files and I'm not sure that it
>really helped. I'm wondering if creating more redo log files is really the
>way to go.
>
>Any ideas on how to solve the "Checkpoint not complete" problem?
>
>Thanks in advance
>Ed Sherman
>
>
>
>-----== Posted via Deja News, The Leader in Internet Discussion ==-----
>http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Received on Thu Jul 23 1998 - 21:20:34 CDT