• 【Oracle】ORA-00257:archiver error. Connect internal only, until freed 错误的处理方法


    archive log 日志已满
    ORA-00257: archiver error. Connect internal only, until freed 错误的处理方法

    1. 用sys用户登录
      sqlplus sys/pass@tt as sysdba

    2. 看看archiv log所在位置
    SQL> show parameter log_archive_dest;
    NAME                                     TYPE        VALUE
    ------------------------------------ ----------- ------------------------------
    log_archive_dest                   string
    log_archive_dest_1               string
    log_archive_dest_10             string

    3. 一般VALUE为空时,可以用archive log list;检查一下归档目录和log sequence
    SQL> archive log list;
    Database log mode                 Archive Mode
    Automatic archival                   Enabled
    Archive destination                  USE_DB_RECOVERY_FILE_DEST
    Oldest online log sequence     360
    Next log sequence to archive  360
    Current log sequence              362

    4. 检查flash recovery area的使用情况,可以看见archivelog已经很大了,达到96.62
    SQL> select * from V$FLASH_RECOVERY_AREA_USAGE;

    FILE_TYPE    PERCENT_SPACE_USED PERCENT_SPACE_RECLAIMABLE NUMBER_OF_FILES
    ------------ ------------------ ------------------------- ---------------
    CONTROLFILE                 .13                        0               1
    ONLINELOG                  2.93                        0               3
    ARCHIVELOG                96.62                       0              141
    BACKUPPIECE                   0                         0               0
    IMAGECOPY                      0                         0               0
    FLASHBACKLOG                0                         0               0

    5. 计算flash recovery area已经占用的空间
    SQL> select sum(percent_space_used)*3/100 from v$flash_recovery_area_usage;

    SUM(PERCENT_SPACE_USED)*3/100
    -----------------------------
                           2.9904
                           
    6. 找到recovery目录, show parameter recover
    SQL> show parameter recover;
    NAME                                 TYPE        VALUE
    ------------------------------------ ----------- ------------------------------
    db_recovery_file_dest                string           /u01/app/oracle/flash_recovery_area
    db_recovery_file_dest_size        big integer   5G
    recovery_parallelism                   integer         0

                           
    7 上述结果告诉我们,归档位置用的是默认值,放在flash_recovery_area下(db_recovery_file_dest目录=/u01/app/oracle/flash_recovery_area)
    [root@sha3 10.2.0]# echo $ORACLE_BASE
    /u01/app/oracle

    [root@sha3 10.2.0]# cd $ORACLE_BASE/flash_recovery_area/tt/archivelog
    转移或清除对应的归档日志, 删除一些不用的日期目录的文件,注意保留最后几个文件(比如360以后的)

    ---------------------------------------------------------------------------------------
    注意: 
    在删除归档日志后,必须用RMAN维护控制文件,否则空间显示仍然不释放。
    ---------------------------------------------------------------------------------------

    8. rman target sys/pass
    [root@sha3 oracle]# rman target sys/pass

    Recovery Manager: Release 10.2.0.4.0 - Production on Tue Jan 20 01:41:26 2009

    Copyright (c) 1982, 2007, Oracle.  All rights reserved.

    connected to target database: tt (DBID=4147983671)

     

    9. 检查一些无用的archivelog
    RMAN> crosscheck archivelog all;

    10. 删除过期的归档
    RMAN> delete expired archivelog all;

    delete archivelog until time 'sysdate-1' ; 删除截止到前一天的所有archivelog

    11. 再次查询,发现使用率正常,已经降到23.03
    SQL> select * from V$FLASH_RECOVERY_AREA_USAGE;

    FILE_TYPE    PERCENT_SPACE_USED PERCENT_SPACE_RECLAIMABLE NUMBER_OF_FILES
    ------------ ------------------ ------------------------- ---------------
    CONTROLFILE                 .13                         0               1
    ONLINELOG                  2.93                         0               3
    ARCHIVELOG                23.03                         0              36
    BACKUPPIECE                   0                         0               0
    IMAGECOPY                     0                         0               0
    FLASHBACKLOG                  0                         0               0

    其它有用的Command:
    ----------------------------------
    如果archive log模式下不能正常startup,则先恢复成noarchive log,startup成功后,再shutdown;
    shutdown immediate;
    startup mount;
    alter database noarchivelog;
    alter database open;
    shutdown immediate;

    再次startup以archive log模式
    shutdown immediate;
    startup mount;
    show parameter log_archive_dest;  
    alter database archivelog;
    archive log list;                 
    alter database open;

    如果还不行,则删除一些archlog log
    SQL> select group#,sequence# from v$log;

        GROUP# SEQUENCE#
    ---------- ----------
             1         62
             3         64
             2         63

    原来是日志组一的一个日志不能归档
    SQL> alter database clear unarchived logfile group 1;
    alter database open;

    最后,也可以指定位置Arch Log, 请按照如下配置
    select name from v$datafile;
    alter system set log_archive_dest='/opt/app/oracle/oradata/usagedb/arch' scope=spfile

    或者修改大小
    SQL> alter system set db_recovery_file_dest_size=3G scope=both;

  • 相关阅读:
    【Charles】高级过滤
    【Python】遍历字典
    sso
    数据可视化大屏,屏幕多分辨率适配方案,且在任意屏幕下保持16:9的比例等比缩放
    从实战的角度谈微服务(八):Sentinel简单使用
    软件研发的成本,效率与质量
    Java获取POST请求Json格式参数(raw)
    Linux日志文件分析
    mongodb非交互式命令
    Java线程池之Executors.newSingleThreadExecutor()
  • 原文地址:https://www.cnblogs.com/uzipi/p/3483793.html
Copyright © 2020-2023  润新知