TNS-12541，TNS-12560，TNS-00511，TNS-12542，TNS-12560，TNS-00512数据库启动监听报错

第 1章   数据库server监听错误

1.1.1数据库监听错误

1.1.1.1 问题及现象

server环境为ORACLE11G RAC环境，系统启动后，监听没起来。

[oracle@RAC4 ~]$ lsnrctl status

LSNRCTL for Linux: Version 11.2.0.3.0 - Production on 29-JUN-2015 10:44:45

Copyright (c) 1991, 2011, Oracle.  All rights reserved.



Connecting to (ADDRESS=(PROTOCOL=tcp)(HOST=)(PORT=1521))

TNS-12541: TNS:no listener

 TNS-12560: TNS:protocol adapter error

  TNS-00511: No listener

   Linux Error: 111: Connection refused

1.1.1.2 检查思路

当时同事已经检查过CRS、也尝试过使用命令启动监听等步骤。可是启动监听的时候会有下面报错。

[oracle@RAC4 ~]$ lsnrctl start

LSNRCTL for Linux: Version 11.2.0.3.0 - Production on 29-JUN-2015 10:51:46

Copyright (c) 1991, 2011, Oracle.  All rights reserved.

Starting /u01/app/oracle/product/11.2.0/db_1/bin/tnslsnr: please wait...

TNSLSNR for Linux: Version 11.2.0.3.0 - Production

Log messages written to /u01/app/oracle/diag/tnslsnr/RAC4/listener/alert/log.xml

Error listening on: (ADDRESS=(PROTOCOL=tcp)(HOST=)(PORT=1521))

TNS-12542: TNS:address already in use

 TNS-12560: TNS:protocol adapter error

  TNS-00512: Address already in use

   Linux Error: 98: Address already in use



Listener failed to start. See the error message(s) above...

竟然说系统已使用，那么理所当然的就去查看1521port和lsn进程

 

[root@RAC4 ~]# ps-ef | grep lsn

grid     52181     1  0 10:04?        00:00:11/u01/*/*/bin/tnslsnrLISTENER_SCAN1 -inherit

root    101600 100759  0 22:30 pts/0    00:00:00 grep lsn

[root@RAC4 ~]#netstat-anp | grep 1521

tcp       0     0xxx.xxx.xxx.xxx:1521          0.0.0.0:*                  LISTEN      52181/tnslsnr     （有些内容不便列出）

发现1521port和/u01/11.2.0/grid/bin/tnslsnrLISTENER_SCAN1-inherit进程存在。所以尝试杀掉此进程

 

[root@RAC4 ~]#kill -9 52181

-bash: kill:(52181) - No suchprocess

杀掉后发现系统提示没有这个进程,再次查询

[root@RAC4 ~]# ps-ef |grep lsn

grid    104147     1  0 23:23?        00:00:00/u01/*/*/bin/tnslsnrLISTENER_SCAN1 -inherit

root    104379 104353  0 23:28 pts/0    00:00:00 grep lsn

又有了。应该是父进程自己主动又给他启动了。查看其父进程是init进程。

[grid@RAC4 ~]$pstree 1

init─┬─MBADataMover───MBADataMover───MBADataMover

├─2*[tnslsnr───2*[{tnslsnr}]]

 

1.1.1.3 出现故障的原因

由于我记得在书上看到过，11G的RAC中SRVCTL能够管理监听，于是尝试使用此命令来启动监听。

1.1.1.4 解决方法

[grid@RAC4 ~]$srvctl start listener -n RAC4

 

[grid@RAC4~]$ exit

logout

[root@RAC4~]# su -oracle

[oracle@RAC4~]$lsnrctl status;

LSNRCTLfor Linux:Version 11.2.0.3.0 - Production on 01-JUL-2015 11:32:45

Copyright(c) 1991,2011, Oracle.  All rightsreserved.

Connectingto(ADDRESS=(PROTOCOL=tcp)(HOST=)(PORT=1521))

STATUSof theLISTENER

------------------------

Alias                    LISTENER

Version                  TNSLSNR for Linux: Version11.2.0.3.0 - Production

StartDate               01-JUL-2015 10:38:57

Uptime                   0 days 0 hr. 53 min. 47 sec

TraceLevel              off

Security                 ON: Local OS Authentication

SNMP                     OFF

ListenerParameterFile  /u01/11.2.0/grid/network/admin/listener.ora

ListenerLogFile        /u01/app/oracle/diag/tnslsnr/RAC4/listener/alert/log.xml

ListeningEndpointsSummary...

(DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER)))

(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=11.10.110.23)(PORT=1521)))

(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=11.10.110.25)(PORT=1521)))

ServicesSummary...

Service"+ASM"has 1 instance(s).

Instance"+ASM2", status READY, has1 handler(s) for this service...

Service"RAC3"has 1 instance(s).

Instance"RAC32", status READY, has1 handler(s) for this service...

Service"RAC3XDB"has 1 instance(s).

Instance"RAC32", status READY, has1 handler(s) for this service...

Service"RACDB007"has 1 instance(s).

Instance "RACDB0072",status READY, has1 handler(s) for this service...

Service"RACDB007JL_BJ"has 1 instance(s).

Instance"RACDB007JL2", status READY,has 1 handler(s) for this service...

Service"RACDB007JL_BJXDB"has 1 instance(s).

Instance"RACDB007JL2", status READY,has 1 handler(s) for this service...

Service"RACDB007XDB"has 1 instance(s).

Instance"RACDB0072", status READY, has1 handler(s) for this service...

The commandcompleted successfully

1.1.1.5 知识点

经过这次问题的解决，发现oracle 11G的RAC中有非常多功能都通过集群命令来控制，而不能单一的通过单机管理的方法来控制集群的某些功能