Re: HELP: Unable to allocate new log (?)

When Oracle gets done writing a redo log in archivelog mode, the archiver kicks in and tries to copy the redo log just written to the archive_dump_dest location. What's happening here, is your archiver process is not done copying your redo log file before he wants to use it again and your database hangs.

There are many, many reasons why this could happen. First, and foremost, you should check to see if you have I/O contention at the OS level. If the archiver is running into contention for the disk, your archived redo log takes longer to write.

If there is little contention, you may not have enough redo for your application. You can either resolve this by either adding more redo logs or increasing the size of your redo logs. If you are using the default redo log size, I would guess they are probably to small. I like to have redo logs about 10-25MB depending on the application.

Mirza Mohsin Beg wrote in message <37E55FCA.4B361B71_at_netearnings.com>...
>
>Hi,
>
>I keep getting messages in my Oracle 8.0.5.0 DB alert log file together
>with a window that pops up on the console saying there was a log-file
>'problem'. The alert log then shows as follows :-
>
>.......
>D:\LOGDIRECTORY1\LOAN_LOG41.ORA
>Thread 1 cannot allocate new log, sequence 189
>All online logs needed archiving
> Current log# 4 seq# 188 mem# 0: T:\REDOLOG\LOG_LOG42.ORA
> Current log# 4 seq# 188 mem# 1: D:\LOGDIRECTORY1\LOAN_LOG41.ORA
>Thread 1 advanced to log sequence 189
>......
>
>The archivelog (and duplex archiving) is turned on and the DB seems to
>be spooling the archive logs in both directories without a problem.
>There is several GB of available space in both directories also. Can
>anyone help?
>
>
>-M
>ps: please email me directly also.
>
Received on Sun Sep 19 1999 - 21:05:16 CDT