oracle_job进程相关学习测试

Posted 2020-11-30

Oracle cjq0进程测试
测试流程：
1.CJQ进程不存在
2.模拟问题处理
3.问题总结

一、问题现象
CJQ0进程不存在
[root@adg1 ~]# ps -ef|grep cjq
root      4741  4675  0 11:36 pts/6    00:00:00 grep cjq
[oracle@adg1 ~]$ sqlplus / as sysdba
SQL*Plus: Release 11.2.0.4.0 Production on Sun Jul 28 11:35:04 2019
SQL> show parameter job_que
NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
job_queue_processes                  integer     100


二、模拟问题处理
1.JOB创建
1)创建JOB测试
https://www.cnblogs.com/lijiasnong/p/3382578.html
--创建表
--truncate table TT;
SQL> GRANT EXECUTE ON DBMS_JOB TO yz;
create table TT(C_DATE DATE); 
create or replace procedure PRO_TT is
 begin 
 insert into TT values(sysdate); 
 commit;
 end;
 /
exec PRO_TT;
SQL> select * from tt;

declare   
job number;      
begin
dbms_job.submit(job,\'PRO_TT;\',sysdate,\'sysdate+1/1440\');
end;
/

--
begin
  dbms_job.remove(32);
  commit;
end;
/

--
SQL> select JOB,WHAT,BROKEN from user_jobs;
       JOB WHAT                                                          B
---------- ----------------------------------------------------------------- -
        34 PRO_TT;                                                       N
--dbms_job.broken(23,false,sysdate); 

第一次执行job        
begin
 dbms_job.run(34);
end;
/

DB_Alert
Sun Jul 28 11:42:21 2019
Starting background process CJQ0
Sun Jul 28 11:42:21 2019
CJQ0 started with pid=31, OS id=4851


[root@adg1 ~]# ps -ef|grep cjq
oracle    4851     1  0 11:42 ?        00:00:00 ora_cjq0_tt
root      4855  4675  0 11:42 pts/6    00:00:00 grep cjq

SQL> alter session set nls_date_format=\'yyyy-mm-dd hh24:mi:ss\';
SQL> select * from tt ;
C_DATE
-------------------
2019-07-28 13:58:59
2019-07-28 13:59:39
--上述操作是，创建一个JOB,并测试手工调用执行存储过程及手工执行JOB均有效，会自动插入一条测试数据。


正常情况下,JOB是间隔1分钟，插入一条记录
select job,
       log_user,
       to_char(last_date,\'yyyy-mm-dd hh24:mi:ss\') last_date,
       to_char(next_date,\'yyyy-mm-dd hh24:mi:ss\') next_date,
       interval,
       what,
       FAILURES
  from user_jobs where job=34
/
       JOB LOG_USER        LAST_DATE           NEXT_DATE              INTERVAL             WHAT                   FAILURES
---------- ---------- ------------------- ------------------- -------------------- -------------------- ----------
        34 YZ           2019-07-28 13:59:39 2019-07-28 14:00:39      sysdate+1/1440       PRO_TT;                  0
SQL> select * from tt ;
C_DATE
-------------------
2019-07-28 13:58:59
2019-07-28 13:59:39
SQL> select sysdate from dual;
SYSDATE
-------------------
2019-07-28 13:25:34
上述验证，发现JOB并未自动调用执行？？？


2)JOB异常,问题排查
2.1 查询日志或者进程信息
db alert无异常，job无异常
SQL> select * from dba_jobs_running;
no rows selected

SQL> SELECT USERNAME,PNAME,PROGRAM,TRACEFILE from v$process where PNAME like \'%CJQ%\';
USERNAME        PNAME PROGRAM                                          TRACEFILE
--------------- ----- -------------------------------------------------------------------------------------------------
oracle          CJQ0  oracle@adg1 (CJQ0)                               /u01/app/oracle/diag/rdbms/tt/tt/trace/tt_cjq0_4851.trc 
SQL> select * from v$bgprocess where name=\'CJQ0\';
PADDR              PSERIAL# NAME  DESCRIPTION                ERROR
---------------- ---------- ----- ---------------------------- ------
00000000F552A288          3 CJQ0  Job Queue Coordinator        0

[oracle@adg1 ~]$ ps -ef|grep cjq
oracle   10884  9146  0 21:37 pts/2    00:00:00 grep cjq
本次环境发现cjq进程不存在！！！
Jobs Not Executing Automatically (文档 ID 313102.1)    


2.2 重置参数
SQL> alter system set job_queue_processes=0;
SQL> alter system set job_queue_processes=200;
DB_Alert
Sun Jul 28 11:36:56 2019
ALTER SYSTEM SET job_queue_processes=0 SCOPE=BOTH;
Sun Jul 28 11:37:26 2019
ALTER SYSTEM SET job_queue_processes=200 SCOPE=BOTH;

[root@adg1 ~]# ps -ef|grep cjq
root      4817  4675  0 11:37 pts/6    00:00:00 grep cjq
通过重置job_queue_process参数并未达到启动cjq进程的目的

2.3 查询sga 内存区域状态，是否为1
SQL> oradebug setmypid
Statement processed.
SQL> oradebug dumpvar sga kkjsre
sword kkjsre_ [060040500, 060040504) = 00000001
       
2.4 修改参数,event
IF -Jobs Do Not Run Automatically as CJQ is Not Getting Started when Jobs are to be Run (文档 ID 2081753.1)    
 Upgraded from 10.2.0.4 to 11.2.0.2
SQL> alter system set events \'10195 trace name context off\';

2.5 手工调用执行Job无异常
begin
 dbms_job.run(34);
end;
/
但是进程还是未启动

2.6 终于发现问题
测试环境JOB，无法自动调用，根据文档排查,测试环境处于受限模式
http://blog.itpub.net/29487349/viewspace-1663945/
受限模式,配置后，已连接session不影响，新的session无restricted session权限则无法登陆db

SQL> select instance_name,logins from v$instance;
INSTANCE_NAME    LOGINS
---------------- ----------
tt               RESTRICTED
SQL> alter system disable restricted session;
System altered.
SQL> select instance_name,logins from v$instance;
INSTANCE_NAME    LOGINS
---------------- ----------
tt               ALLOWED
SQL> select * from tt ;
C_DATE
-------------------
2019-07-28 13:58:59
2019-07-28 13:59:39
2019-07-28 14:09:52
2019-07-28 14:10:52
2019-07-28 14:11:52
2019-07-28 14:12:52
6 rows selected.


SQL> truncate table tt;
Table truncated.


Alert 日志,恢复正常情况后，job自动调用执行，自动后台启动cjq进程服务。
Sun Jul 28 22:28:24 2019
Restarting dead background process CJQ0
Sun Jul 28 22:28:24 2019
CJQ0 started with pid=20, OS id=11444 

SQL> alter system set job_queue_processes=0;
Sun Jul 28 22:34:11 2019
Stopping background process CJQ0
Sun Jul 28 22:34:11 2019
ALTER SYSTEM SET job_queue_processes=0 SCOPE=BOTH;
SQL> alter system set job_queue_processes=200;
[oracle@adg1 ~]$ ps -ef|grep cjq
oracle   11545  9146  0 22:34 pts/2    00:00:00 grep cjq

总结：
1.修改job_queue_processes,从>0修改为0，并且cjq0进程存在，则会kill cjq0进程，并且后续的job无法自动调用执行
2.修改参数job_queue_processes >0,无法立即启动cjq0进程，而需要配合Job定时调用一个工作，会先启动cjq0进程(如果参数为0,则无法启动)
3.手工执行调用存储过程，job id 并非能够立即启动cjq0进程，修改events \'10195 trace name context off\'只是解决，
存在CJQ进程或者10g升级至11g job无法自动执行问题处理;
4.如果数据库连接状态为RESTRICTED 受限模式，如果没有相应权限，则JOB无法自动执行;
5.如果cjq进程存在，但是job并不能自动调用执行，并且sga内存区域不为1 00000001，则需要重启或者根据mos 
Jobs do not execute automatically (Doc ID 309945.1)进行处理
SQL> exec  dbms_ijob.set_enabled(true);