環境11gR2 ,centOS64bit
問題描述：
Tue May 07 18:40:01 2013
Thread 1 cannot allocate new log, sequence 78
Private strand flush not complete
Current log# 4 seq# 77 mem# 0: /u01/app/oracle/oradata/upload/redo04.log
Thread 1 advanced to log sequence 78 (LGWR switch)
Current log# 5 seq# 78 mem# 0: /u01/app/oracle/oradata/upload/redo05.log
Tue May 07 18:40:01 2013
Archived Log entry 36 added for thread 1 sequence 77 ID 0x30de06b8 dest 1:1：增加redo group
2：設置參數db_writer_processes
未能解決該問題！
以下文章摘錄自谷歌：
ODM FINDING:
Alert Log Messages: Private Strand Flush Not Complete
Applies to:
Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 11.2.0.3 - Release: 10.2 to 11.2
Information in this document applies to any platform.
Private strand flush not complete
Symptoms
"Private strand flush not complete" messages are being populated to the alert log, example:
Mon Jan 23 16:09:36 2012
Thread 1 cannot allocate new log, sequence 18358
Private strand flush not complete
Current log# 7 seq# 18357 mem# 0: /u03/oradata/bitst/redo07.log
Thread 1 advanced to log sequence 18358
Current log# 8 seq# 18358 mem# 0: /u03/oradata/bitst/redo08.log
>>
>>
Changes
When you switch logs all private strands have to be flushed to the current log before the switch is allowed to proceed.
Cause
The message means that we haven't completed writing all the redo information to the log when we are trying to switch. It is similar in nature to a "checkpoint not complete" except that is only involves the redo being written to the log. The log switch can not occur until all of the redo has been written.
A "strand" is new terminology for 10g and it deals with latches for redo .
Strands are a mechanism to allow multiple allocation latches for processes to write redo more efficiently in the redo buffer and is related to the log_parallelism parameter present in 9i.
The concept of a strand is to ensure that the redo generation rate for an instance is optimal and that when there is some kind of redo contention then the number of strands is dynamically adjusted to compensate.
The initial allocation for the number of strands depends on the number of CPU's and is started with 2 strands with one strand for active redo generation.
For large scale enterprise systems the amount of redo generation is large and hence these strands are *made active* as and when the foregrounds encounter this redo contention (allocated latch related contention) when this concept of dynamic strands comes into play.
There is always shared strands and a number of private strands .
Oracle 10g has some major changes in the mechanisms for redo (and undo), which seem to be aimed at reducing contention.
Instead of redo being recorded in real time, it can be recorded 'privately' and pumped into the redo log buffer on commit.
Similarly the undo can be generated as 'in memory undo' and applied in bulk. This affect the memory used for redo management and the possibility to flush it in pieces. The message you get is related to internal Cache Redo File management.
...You can disregard these messages as normal messages.
Solution
These messages are not a cause for concern unless there is a significant time gap between the "cannot allocate new log" message and the "advanced to log sequence" message.
Increasing the value for db_writer_processes can in some situations avoid the message from being generated.
Hdr: 7005709 10.2.0.3 RDBMS 10.2.0.3 TXN MGMT LOCAL PRODID-5 PORTID-59
Abstract: LOG FILE SWITCH (PRIVATE STRAND FLUSH INCOMPLETE) WHEN _IN_MEMORY_UNDO=TRUE
PROBLEM:
--------
Oracle 10.2.0.3.0
This is a large database for Oracle Applications plus a lot of custom code.
Last night a user noticed slowness between 9 and 10 pm. When we now run AWR
and ADDM reports, we see high waits for "log file switch (private strand
flush
incomplete)".
This looks similar to Bug 6806770 except that we're HPUX 11.11, non-Itanium.
As per the bug, we currently have _in_memory_undo=TRUE.
We had this setting set to false for a while as a workaround for another bug
(5028099), but re-enabled it after applying a patch (6733906).
$ >grep memory spfilecfap.ora
*._in_memory_undo=TRUE
DIAGNOSTIC ANALYSIS:
--------------------
One night a user noticed slowness between 9 and 10 pm. When we now
run AWR and ADDM reports, we see high waits for "log file switch (private
strand flush incomplete)".
We had _in_memory_undo=FALSE from Jan 15 until March 16, when we applied
patch 6733906. Before Jan 15, we had numerous alert log messages like what's
below. Since Mar 16, we've had approx. 25 more of these messages. In
between Jan 15 and Mar 16 we had none of the messages.
Thread 1 cannot allocate new log, sequence 92288
Private strand flush not complete
WORKAROUND:
-----------
The only workaround is to set parameter _in_memory_undo = FALSE to disable
IMU
RELATED BUGS:
-------------
6806770 - still at status 11
Slow Running User Process And Top Database Wait Event Is 'log file switch (private strand flush incomplete)'
Applies to:
Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 10.2.0.3 - Release: 10.2 to 10.2
Information in this document applies to any platform.
Symptoms
Checked for relevance on 31-Oct-2011
A user process is running very slow and the top database wait event is always 'log file switch (private strand flush incomplete)'.
AWR Report shows the following Top 5 Timed Events.
log file switch (private strand flush incomplete)
buffer busy waits
log file sync
db file sequential read
log file switch (checkpoint incomplete)
Cause
Unpublished Bug 5605290 which causes deadlocks between the CKPT and LGWR or DBWR processes, and can manifest itself in many ways.
Solution
Unpublished Bug 5605290 is fixed in the 10.2.0.4 patchset, so applying it will resolve this issue. In addition there is a workaround available to prevent the problem until able to get the database patched up.
sqlplus / as sysdba
alter system set "_in_memory_undo" = FALSE scope=both;