1、查看 alert_PROD.log
【错误信息】:ORA-1652: unable to extend temp segment by 128 in tablespace TEMP1
查看临时表空间基础信息及其使用情况:
基础信息查看:
select dtf.TABLESPACE_NAME, dtf.file_name, dtf.BYTES / 1024 / 1024 as "MB" from dba_temp_files dtf order by dtf.TABLESPACE_NAME;
使用情况查看:
SELECT d.tablespace_name "Name", TO_CHAR(NVL(a.bytes / 1024 / 1024 / 1024, 0), '99,999,990.900') "Size (G)", TO_CHAR(NVL(t.hwm, 0) / 1024 / 1024 / 1024, '99999999.999') "HWM (G)", TO_CHAR(NVL(t.hwm / a.bytes * 100, 0), '990.00') "HWM % ", TO_CHAR(NVL(t.bytes / 1024 / 1024 / 1024, 0), '99999999.999') "Using (G)", TO_CHAR(NVL(t.bytes / a.bytes * 100, 0), '990.00') "Using %" FROM sys.dba_tablespaces d, (select tablespace_name, sum(bytes) bytes from dba_temp_files group by tablespace_name) a, (select tablespace_name, sum(bytes_cached) hwm, sum(bytes_used) bytes from v$temp_extent_pool group by tablespace_name) t WHERE d.tablespace_name = a.tablespace_name(+) AND d.tablespace_name = t.tablespace_name(+) AND d.extent_management like 'LOCAL' AND d.contents like 'TEMPORARY';
通过查看临时表空间组中的表空间可以判断,该表空间处于异常状态。
2、启动ORA-1652跟踪错误
在session级别启用数据库的跟踪:
ALTER SESSION SET EVENTS '1652 trace name errorstack';
在系统级别启用数据库的跟踪:
ALTER SYSTEM SET EVENTS '1652 trace name errorstack';
把该参数写入到spfile文件中:
ALTER SYSTEM SET EVENT = '1652 trace name errorstack' SCOPE = SPFILE;
3、查看alert_PROD.log中错误日志
Errors in file /ora/prod/12.1.0/admin/PROD_erpdb/diag/rdbms/prod/PROD/trace/PROD_ora_5165.trc: more /ora/prod/12.1.0/admin/PROD_erpdb/diag/rdbms/prod/PROD/trace/PROD_ora_5165.trc
----- Error Stack Dump ----- ORA-1652: unable to extend temp segment by 128 in tablespace TEMP1 ----- Current SQL Statement for this session (sql_id=cfrvgzfkpxp49) ----- 注:中间为需要优化的代码 ----- PL/SQL Stack -----
4、打印SQL执行计划
pl/sql Developer 工具按F5.
5、分析执行计划
全表扫描(Table access full)的子句。
通过SQL优化的规则分析该字句。
6、对该SQL重要部分在Test系统中进行优化测试。
(1)执行优化前:
(2)创建索引
create index ind_CPMD_TRANSACTION_ID on cux.CUX_PA_MATERIAL_A_DETAILS(nvl(CPMD.TRANSACTION_ID,0));
(3)修改后
7、检查Test系统中进行优化结果。
优化前后对比
名称 |
优化前 |
优化后 |
扫描方式 |
TABLE ACCESS FULL |
INDEX RANGE SCAN |
Cost |
12,735 |
844 |
IO cost |
12,681 |
842 |
CPU cost |
447,695,496 |
17,435,302 |
8、在Prod系统中进行部署
(1)备份 CUX_PA_MAT_A_DETS_COLLECT_PKG到本地
(2)创建索引
create index ind_CPMD_TRANSACTION_ID on cux.CUX_PA_MATERIAL_A_DETAILS(nvl(CPMD.TRANSACTION_ID,0));
(3)进行优化,并添加注释
A、WHERE CPMD.TRANSACTION_ID is not null 替换为 WHERE NVL(CPMD.TRANSACTION_ID,0) >0 B、 AND MTLN.LOT_NUMBER NOT LIKE '%|%' 替换为: AND INSTR(MTLN.LOT_NUMBER,'|') < 1 C、 AND T.ACCOUNT_NAME LIKE '%出库%'; 替换为: AND INSTR(T.ACCOUNT_NAME,'出库') > 0 ;
(4)进行代码检查
检查项:a、检查修改代码是否正确
b、检查源代码是否已经注释
C、检查是否在修改代码后添加注释(包括时间、功能、修改人员)
(5)执行代码,重新提交请求
9、执行查询操作,查看结果
整体优化前后对比:
名称 |
优化前 |
优化后 |
扫描方式 |
TABLE ACCESS FULL |
INDEX RANGE SCAN |
Cost |
51,651 |
751 |
IO cost |
51,423 |
740 |
CPU cost |
1,885,854,949 |
94,466,486 |
9、检查alert_PROD.log中是否出现ORA-1652错误
Beginning log switch checkpoint up to RBA [0x8c9.2.10], SCN: 5965544065707 Tue Mar 20 16:19:47 2018 Thread 1 advanced to log sequence 2249 (LGWR switch) Current log# 1 seq# 2249 mem# 0: /data/prod/proddata/log01a.dbf Current log# 1 seq# 2249 mem# 1: /data/prod/proddata/log01b.dbf Tue Mar 20 16:19:51 2018 Archived Log entry 2248 added for thread 1 sequence 2248 ID 0x1556bb36 dest 1: Tue Mar 20 16:24:52 2018 Completed checkpoint up to RBA [0x8c9.2.10], SCN: 5965544065707 Tue Mar 20 16:27:28 2018 Incremental checkpoint up to RBA [0x8c9.a420.0], current log tail at RBA [0x8c9.21670.0] Tue Mar 20 16:47:32 2018 Incremental checkpoint up to RBA [0x8c9.32675.0], current log tail at RBA [0x8c9.3d64f.0] Tue Mar 20 16:58:03 2018 Tue Mar 20 17:00:03 2018 Closing scheduler window Closing Resource Manager plan via scheduler window Clearing Resource Manager plan via parameter Tue Mar 20 17:07:35 2018 Incremental checkpoint up to RBA [0x8c9.547b9.0], current log tail at RBA [0x8c9.5979d.0] Tue Mar 20 17:27:37 2018 Incremental checkpoint up to RBA [0x8c9.5eece.0], current log tail at RBA [0x8c9.609cf.0]
10、关闭ORA-1652跟踪事件。对应的关闭脚本如下:
ALTER SESSION SET EVENTS '1652 trace name context off'; ALTER SYSTEM SET EVENTS '1652 trace name context off'; ALTER SYSTEM RESET EVENT SCOPE = SPFILE SID = '*';