1
CRS磁盘组使用率过高处理.txt
10墨值下载
检查发现客户某套数据库（19C）CRS 磁盘组使用率过高
检查发现是由于 19c 自带的 mgmt 实例下有一个 undo 表空间当前大小为 15G，占用了主要的空间
经过 mos 匹配发现是 bug：
Bug 24331121 - OCR_VOTE DISK GROUP IS 99% FULL, HIGH UNDO USED BY MGMTDB
可以通过更换（重建）undo 表空间来解决该问题，过程如下：
查看相关信息：
###使用以下命令均可确定 MGMT 实例所在位置
crsctl stat res -t
srvctl status mgmtdb
ps -ef| grep pmon_-MGMTDB
ps -ef| grep MGMTLSNR
###连接 mgmt 实例
export ORACLE_SID=-MGMTDB
sqlplus / as sysdba
SQL> show parameter undo_tablespace
NAME TYPE VALUE
------------------------------------ ----------------------
------------------------------
undo_tablespace string UNDOTBS1
SQL> create pfile='/home/grid/pfile_20200119.ora' from spfile;
File created.
[grid@tyhcdb1 ~]$ cat pfile_20200119.ora | grep -i undo
*.undo_tablespace='UNDOTBS1'
###连接 pbd
SQL> show pdbs
CON_ID CON_NAME OPEN MODE RESTRICTED
---------- ------------------------------ ---------- ----------
2 PDB$SEED READ ONLY NO
3 GIMR_DSCREP_10 READ WRITE NO
SQL> alter session set container=GIMR_DSCREP_10;
Session altered.
查看表空间信息：
set lines 999 pages 999
col TABLESPACE_NAME for a20
col FILE_NAME for a90
select TABLESPACE_NAME,FILE_NAME,AUTOEXTENSIBLE,bytes/1024/1024/1024
use_size_gb,MAXBYTES/1024/1024/1024 max_size_gb from dba_data_files;
TABLESPACE_NAME FILE_NAME
AUTOEX USE_SIZE_GB MAX_SIZE_GB
--------------------
--------------------------------------------------------------------------------
---------- ------ ----------- -----------
SYSTEM
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/system.274.1012732343
YES .21484375 31.9999847
SYSAUX
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysaux.273.1012732343
YES .302734375 31.9999847
UNDOTBS1
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/undotbs1.272.1012732343
YES 15.2294922 31.9999847
USERS
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/users.276.1012732375
YES .004882813 31.9999847
SYSGRIDHOMEDATA
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysgridhomedata.277.10127
32447 NO .48828125 0
SYSCALOGDATA
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/syscalogdata.278.10127324
49 NO 1 0
SYSMGMTDATA
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysmgmtdata.279.101273245
1 NO 2 0
SYSMGMTDATADB
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysmgmtdatadb.280.1012732
455 NO 8 0
SYSMGMTDATACHAFIX
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysmgmtdatachafix.281.101
2732469 NO 5 0
SYSMGMTDATAQ
+OCR/_MGMTDB/8CD2C997CC0F243DE0538802870A657F/DATAFILE/sysmgmtdataq.282.10127324
79 NO 2 0
10 rows selected.
SQL> show parameter db_create_file_dest
NAME TYPE VALUE
------------------------------------ ----------------------
------------------------------
db_create_file_dest string +OCR
1.创建一个新的 undo 表空间
export ORACLE_SID=-MGMTDB
sqlplus / as sysdba
alter session set container=GIMR_DSCREP_10;
create undo tablespace undotbs2 datafile '+OCR' size 2G;
2.更换数据库 undo 表空间
更换之前可以先修改下 undo_retention 参数，作用是可以让 undo 段尽快 offline 以便于后面的 drop 表
空间释放空间
alter system set undo_retention=300 scope=memory sid='*';
alter system set undo_tablespace=undotbs2 sid='*';
3.检查系统是否已经切换到 undotbs2 上，确认 undotbs1 已经 offline
of 3
10墨值下载
【版权声明】本文为墨天轮用户原创内容，转载时必须标注文档的来源（墨天轮），文档链接，文档作者等基本信息，否则作者和墨天轮有权追究责任。如果您发现墨天轮中有涉嫌抄袭或者侵权的内容，欢迎发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。
下载排行榜
评论