DBMS_SCHEDULER.create_job 自动触发的BUG

Scheduler Job Runs Immediately After Enable When Using PL/SQL-expression (Doc ID 2420041.1)

APPLIES TO:

Oracle Database - Enterprise Edition - Version 11.2.0.4 to 18.1.0.0.0 Release 11.2 to 18

Oracle Database Cloud Schema Service - Version N/A and later

Oracle Database Exadata Cloud Machine - Version N/A and later

Oracle Cloud Infrastructure - Database Service - Version N/A and later

Oracle Database Cloud Exadata Service - Version N/A and later

Information in this document applies to any platform.

SYMPTOMS

A job created using DBMS_SCHEDULER, runs automatically after it is enabled, when REPEAT_INTERVAL is calculated by a function (repeat_interval =>'nextExecution',).

The job runs after it is enabled even if the NEXT_RUN_DATE is in the future.

  1. Create a table named testJobTab

SQL> create table testJobTab ( insertdate date);

  1. Create a procedure to insert into the table

SQL> create or replace procedure testJobinsert

is

BEGIN

insert into testJobTab (insertdate) values ( sysdate);

END;

/

  1. Create a function to calculate the date

SQL> create or replace function nextExecution

return date

is

BEGIN

return trunc(sysdate + 10);

END;

/

  1. make sure the table is empty

SQL> truncate table testJobTab;

  1. Create a job (disabled)

SQL> BEGIN

DBMS_SCHEDULER.create_job (

job_name => 'TESTJOB',

job_type => 'PLSQL_BLOCK',

job_action => 'BEGIN testJobinsert; commit; END;',

start_date => systimestamp at TIME zone 'EUROPE/BERLIN',

repeat_interval => 'nextExecution',

end_date => NULL,

enabled => FALSE,

comments => 'TESTJOB'

);

END;

/

  1. check the table

SQL> select * from testJobTab;

no rows selected

  1. check user_scheduler_jobs

SQL> set lines 300

SQL> col start_date for a50

SQL> col last_start_date for a50

SQL> col next_run_date for a50

SQL> col repeat_interval for a25

SQL> select start_date,run_count,last_start_date,next_run_date,enabled,schedule_type,repeat_interval from user_scheduler_jobs where job_name = 'TESTJOB';

START_DATE RUN_COUNT LAST_START_DATE

NEXT_RUN_DATE

ENABL SCHEDULE_TYP REPEAT_INTERVAL





13-MAR-18 08.45.52.057030 AM EUROPE/BERLIN 0

FALSE PLSQL nextExecution

  1. enable the job

SQL> exec DBMS_SCHEDULER.ENABLE('TESTJOB');

  1. check the table (job started after it was ENABLED)

SQL> select * from testJobTab;

INSERTDATE


13-MAR-2018 07:47:01

  1. check user_scheduler_jobs

There is a difference of 10 days between START_DATE/LAST_START_DATE and NEXT_RUN_DATE. LAST_START_DATE and RUN_COUNT confirms that the job has run.

SQL> select start_date,run_count,last_start_date,next_run_date,enabled,schedule_type,repeat_interval from user_scheduler_jobs where job_name = 'TESTJOB';

START_DATE RUN_COUNT LAST_START_DATE

NEXT_RUN_DATE

ENABL SCHEDULE_TYP REPEAT_INTERVAL





13-MAR-18 08.45.52.057030 AM EUROPE/BERLIN 1 13-MAR-18

08.47.01.342411 AM EUROPE/BERLIN 23-MAR-18 12.00.00.000000 AM +00:00

TRUE PLSQL nextExecution

  1. disable the job

SQL> exec DBMS_SCHEDULER.DISABLE('TESTJOB');

  1. check the table (the job did not run after it was DISABLED)

SQL> select * from testJobTab;

INSERTDATE


13-MAR-2018 07:47:01

  1. check user_scheduler_jobs

No modifications to RUN_COUNT, LAST_START_DATE and NEXT_RUN_DATE.

SQL> select start_date,run_count,last_start_date,next_run_date,enabled,schedule_type,repeat_interval from user_scheduler_jobs where job_name = 'TESTJOB';

START_DATE RUN_COUNT LAST_START_DATE

NEXT_RUN_DATE

ENABL SCHEDULE_TYP REPEAT_INTERVAL





13-MAR-18 08.45.52.057030 AM EUROPE/BERLIN 1 13-MAR-18

08.47.01.342411 AM EUROPE/BERLIN 23-MAR-18 12.00.00.000000 AM +00:00

FALSE PLSQL nextExecution

  1. enable the job again

SQL> exec DBMS_SCHEDULER.ENABLE('TESTJOB');

  1. check the table (notice that the job has run again)

SQL> select * from testJobTab;

INSERTDATE


13-MAR-2018 07:47:01

13-MAR-2018 07:53:32

  1. check user_scheduler_jobs

SQL> select start_date,run_count,last_start_date,next_run_date,enabled,schedule_type,repeat_interval from user_scheduler_jobs where job_name = 'TESTJOB';

START_DATE RUN_COUNT LAST_START_DATE

NEXT_RUN_DATE

ENABL SCHEDULE_TYP REPEAT_INTERVAL





13-MAR-18 08.45.52.057030 AM EUROPE/BERLIN 1 13-MAR-18

08.53.32.442546 AM EUROPE/BERLIN 23-MAR-18 12.00.00.000000 AM +00:00

TRUE PLSQL nextExecution

  1. drop the job

SQL> exec DBMS_SCHEDULER.drop_job ('TESTJOB');

SQL> drop table testJobTab purge;

SQL> drop function nextExecution;

SQL> drop procedure testJobinsert;

If REPEAT_INTERVAL is used with FREQ and INTERVAL (repeat_interval => 'FREQ=Minutely;INTERVAL=30'), the job does not run when is enabled which is the expected behavior.

CHANGES

CAUSE

This is caused by

BUG 27697734 - SCHEDULER JOB RUNS IMMEDIATELY AFTER DISABLE/ENABLE WHEN USING PL/SQL-EXPRESSION

SOLUTION

The BUG 27697734 - is regressed and tracking in Bug 28174090

Bug 28174090 - IMPDP FAILS WITH ORA-12005 WHEN USING JOBS HAVING PL/SQL REPEAT_INTERVAL

Download the Patch 28174090 from MOS and apply the Patch 28174090 to your current release and platform.

相关推荐
隔窗听雨眠3 天前
Oracle日志体系分层结构与故障定位路径
数据库·oracle·故障排查
曲径海盗4 天前
Try-Except怎么用,3个追问解答在自动化下的应用
java·jvm·oracle
这个DBA有点耶7 天前
数据同步的三个工程底线:DDL同步、断点续传、一致性校验——开源CDC的边界与KFS的路径
数据库·dba·代码规范
爱吃奥利奥_wen7 天前
SQL Server 基础语法详解:从建库建表到约束与增删改查
服务器·数据库·oracle
x_SpiderMan7 天前
搭建Oracle RAC 集群
数据库·oracle
测试界的小码农7 天前
【图数据库实战进阶(03):Neo4j 图数据建模实战——节点、关系与属性应该怎么设计?】
数据库·oracle·neo4j
.痞老板.7 天前
mysql初阶--数据库设计
数据库·mysql·oracle
一个天蝎座 白勺 程序猿8 天前
KingbaseES数据库归档重做日志文件管理:从配置到运维的实战指南
运维·数据库·oracle·kingbasees·金仓数据库
腾科IT教育10 天前
Oracle考试全解析:OCP、OCM认证体系与核心考点备考指南
数据库·oracle·开闭原则
这个DBA有点耶11 天前
以为自增ID只是有空洞?聊聊双主架构下无声的“数据覆盖”
数据库·mysql·dba