Original works, from the "Blue Blog" blog, when reproduced please be sure to indicate the source, otherwise have the right to pursue copyright legal responsibility.
The blog:http://blog.csdn.net/huangyanlong/article/details/46863813 of Blue
Environment:
Operating system: CentOS 6.4 64bit
Middleware: Was 7
Database: Oracle 11G RAC 11.2.0.3
Problem:
Build Library was unable to connect to the database and reported that the database could not be archived incorrectly.
Solution Ideas:
Address Oracle's inability to archive issues.
Operation Record:
Sql> select instance_name from V$instance;
Instance_name
----------------
Hyl1
Sql> select * from V$log;
sql> alter system switch logfile;
--Manual switch log Group, long time tamping, manual stop
^calter System Switch LogFile
*
ERROR at line 1:
Ora-01013:user requested cancel of current operation
--View the alarm log and find an archive failure error, which points to a ora-00333 error, possibly corrupted log group
Incomplete read from log member ' +data/hyl/redo01.log '. Trying next member.
Arc0:all Archive Destinations made inactive due to error 333
Arc0:closing Local archive Destination log_archive_dest_1: ' +arch/hyl/archivelog/2015_07_06/thread_1_seq_ 3935.542.884340013 ' (Error 333) (HYL1)
Arch:archival stopped, error occurred. would continue retrying
ORACLE Instance Hyl1-archival Error
Ora-16038:log 1 sequence# 3935 cannot be archived
Ora-00333:redo Log read error block count
Ora-00312:online Log 1 Thread 1: ' +data/hyl/redo01.log '
Mon Jul 06 10:00:18 2015
Arch:archival stopped, error occurred. would continue retrying
ORACLE Instance Hyl1-archival Error
Ora-16014:log 1 sequence# 3935 not archived, no available destinations
Ora-00312:online Log 1 Thread 1: ' +data/hyl/redo01.log '
--Verify the archive log directory space, excluding the lack of space caused
Sql> select Group_number,name,total_mb,free_mb from V$asm_diskgroup;
Group_number NAME TOTAL_MB FREE_MB
------------ ------------------------------ ---------- ----------
1 CRS 449952 449023
2 DATA 14500000 13677212
3 ARCH 1000000 999167
4 FLASH 1000000 999895
Sql> select * from V$log;
--View the archive status and discover that log Group 1 (not the current log group) is not archived
Sql> Col MEMBER for A50
Sql> SELECT * from V$logfile order by group#;
group# STATUS TYPE MEMBER is_
---------- ------- ------- -------------------------------------------------- ---
1 ONLINE +data/xzxt/redo01.log NO
2 ONLINE +data/xzxt/redo02.log NO
3 ONLINE +data/xzxt/redo03.log NO
4 ONLINE +data/xzxt/redo04.log NO
The logs are not diversified.
Sql> ALTER DATABASE Clear unarchived logfile Group 1;
--Log Group 1 is not the current log group, try to empty it
Sql> select * from V$log;
--Archive restored to normal
sql> alter system switch logfile;
Summary:
This failure is caused by corruption of the log group, and an error occurs when the current log group is full for log switching and needs to be archived. Because this corruption log group is not the current log group, processing is simpler. However, if the current log group is corrupted due to power outages, hardware failures, and so on, the database will need to be backed up for incomplete recovery and data loss will occur. It is recommended to establish a diversity of log groups.
After the failure, follow-up maintenance advice made a brief summary, feedback to the resident engineer.
Suggestions:
1, to find out whether there is a power outage in the near future, to determine the cause of the log group damage;
2, confirm the hardware equipment without any warning, whether it is a hardware level failure;
3, if the exclusion of power, hardware failure and other factors, may be due to Oracle logic errors, need to continue tracking;
4, it is recommended that after this failure, a full preparation as soon as possible;
5, it is recommended to complete a plurality of log groups;
6, performance considerations, it is recommended to post a single node log group from 2 groups to 3 groups or more;
RAC 11g_ Production Library log group corruption handling