RMAN duplicate from active 时遭遇 ORA-17627 ORA-12154

最近在从活动数据库进行异机克隆时碰到了ORA-17629，ORA-17627，ORA-12154的错误，起初以为是一个Bug呢。Oracle Bug着实太多了，已经成了习惯性思维了。汗！错误提示是无法连接到连接到远程数据库，连接字符串无法解析。咦，配置了从auxiliary DB到target DB的tnsnames，且都是连通的阿......

1、故障现象
    --下面的操作在auxiliary DB所在的机器上完成
    [oracle@linux4 ~]$ export ORACLE_SID=sybo3
    [oracle@linux4 ~]$ sqlplus / as sysdba
    SQL> startup nomount;
    [oracle@linux4 ~]$ rman target sys/oracle@TAR auxiliary sys/oracle@AUX

Recovery Manager: Release 11.2.0.1.0 - Production on Wed Jul 31 16:00:59 2013

Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.

connected to target database: SYBO3 (DBID=2347733014)
    connected to auxiliary database: SYBO3 (not mounted)

RMAN> duplicate database to sybo3 from active database spfile nofilenamecheck;

Starting Duplicate Db at 31-JUL-13
    using target database control file instead of recovery catalog
    allocated channel: ORA_AUX_DISK_1
    channel ORA_AUX_DISK_1: SID=125 device type=DISK

contents of Memory Script:
    {
       backup as copy reuse
       targetfile  '/u01/oracle/db_1/dbs/spfilesybo3.ora' auxiliary format
     '/u01/oracle/db_1/dbs/spfilesybo3.ora'   ;
       sql clone "alter system set spfile= ''/u01/oracle/db_1/dbs/spfilesybo3.ora''";
    }
    executing Memory Script

Starting backup at 31-JUL-13
    allocated channel: ORA_DISK_1
    channel ORA_DISK_1: SID=23 device type=DISK
    RMAN-00571: ===========================================================
    RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
    RMAN-00571: ===========================================================
    RMAN-03002: failure of Duplicate Db command at 07/31/2013 16:01:07
    RMAN-03015: error occurred in stored script Memory Script
    RMAN-03009: failure of backup command on ORA_DISK_1 channel at 07/31/2013 16:01:07
    ORA-17629: Cannot connect to the remote database server                       
    ORA-17627: ORA-12154: TNS:could not resolve the connect identifier specified
    ORA-17629: Cannot connect to the remote database server

2、故障分析与解决
    上面是一堆的RMAN与ORA相关错误号了
    在执行脚本backup as copy reuse时出现了错误，也就是说备份的时候出现错误RMAN-03009
    其次是ORA-17629，不能连接到远程数据库服务器，ORA-17627后跟着ORA-12154，说明是由于无法解析字符串
    下面来看看ORA-17629到底是什么问题
    [oracle@linux4 ~]$ oerr ora 17629
    17629, 00000, "Cannot connect to the remote database server"
    // *Cause:  Connecting to the remote server specified by database connect
    //          string for netowrk file transfer failed.
    // *Action: Check additional error messages

重要的提示信息:for netowrk file transfer failed，网络文件传输失败。
    由于我们使用的是从活动数据库进行克隆，因此活动数据库的数据文件等等应当会通过网络复制到辅助数据库，更确切地是说是通过Oracle Net。
    也就是说尽管我们在辅助数据库端配置了到target DB以及到Auxiliary DB的tnsnames的连接是不够的，target DB端也要连接到Auxiliary DB传送文件。
    上面只是一个初步的推测，从Oracle Metalink找到了关于这个问题的描述。就是需要在两个服务器之间都配置到target DB与Auxiliary DB的tnsnames。

检查一下两个主机的tnsnames.ora的配置
    [oracle@linux3 admin]$ more tnsnames.ora
    # tnsnames.ora Network Configuration File: /u01/oracle/db_1/network/admin/tnsnames.ora
    # Generated by Oracle configuration tools.

TAR =                               #Target Server只有到Target DB的tnsnames.ora
      (DESCRIPTION =
        (ADDRESS_LIST =
          (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.7.25)(PORT = 1531))
        )
        (CONNECT_DATA =
          (SERVICE_NAME = SYBO3.ORASRV.COM)
        )
      )
 
    [oracle@linux4 admin]$ more tnsnames.ora  #Auxiliary Server有到Target DB和Auxiliary DB的tnsnames.ora
    TAR =
      (DESCRIPTION =
        (ADDRESS_LIST =
          (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.7.25)(PORT = 1531))
        )
        (CONNECT_DATA =
          (SERVICE_NAME = SYBO3.ORASRV.COM)
        )
      )

AUX =
      (DESCRIPTION =
        (ADDRESS_LIST =
          (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.7.26)(PORT = 1531))
        )
        (CONNECT_DATA =
          (SERVICE_NAME = SYBO3.ORASRV.COM)
        )
      )

-- Author : Robinson Cheng
    -- Blog   : http://blog.csdn.net/robinson_0612

解决方案
        在Target Server端添加到Auxiliary DB的tnsnames.ora，即与Auxiliary server端使用相同的tnsnames entry

下面测试tnsnames的连通性
    [oracle@linux3 admin]$ tnsping aux 
    [oracle@linux3 admin]$ tnsping tar
    [oracle@linux4 admin]$ tnsping aux
    [oracle@linux4 admin]$ tnsping tar

再次duplicate时，成功鸟，喝茶哟！

3、MetaLink的描述(Doc ID 1144273.1)
    Applies to:
    Oracle Database - Enterprise Edition - Version 11.2.0.1 and later
    Information in this document applies to any platform.

Symptoms

The duplicate database from active failed with the following errors :

ORA-17629: Cannot connect to the remote database server
    ORA-17627: ORA-12154: TNS:could not resolve the connect identifier specified
    ORA-17629: Cannot connect to the remote database server

Cause
        Cause 1:

For Active Duplication you need to connect to the target and Auxiliary DB using net service name ,even if you are running RMAN at the Auxiliary DB.
        The net service name to connect to the Auxiliary DB should also be configured at the target Database.
  
        In the tnsnames.ora on the machine of the TARGET database, there are 2 services created to connect to auxiliary database that have the same "service name" but "different" configuration.

Possible Cause 2:
  
        Mismatch of Parameters DEFAULT_DOMAIN and DB_DOMAIN
  
        db_domain=PROD.world         # init<SID>.ora
        default_domain=PROD.WORLD    # sqlnet.ora
  
        # DEFAULT_DOMAIN defined on sqlnet.ora is 'case' sensitive and when defined must match
           DB_DOMAIN initialization parameter

Solution

1. To be sure that the tnsnames.ora file of each machine (machine where is the source database and machine where is the auxiliary  database) has the service name to connect to auxiliary database.
  
        2. Verify that there are not duplicate service_name with different configuration in the tnsnames.ora file.
  
        As per the Step 2 of the below RMAN doc :
        http://download.oracle.com/docs/cd/B28359_01/backup.111/b28270/rcmdupdb.htm#i1008564
  
          Step 2: Establish Oracle Net Connectivity to the Auxiliary Instance
  
        When duplicating from an active database, you must first have connected as SYSDBA to the auxiliary instance by means of a net service name. This net service name must also be available on the source database instance.
        The source database instance, to which RMAN is connected as TARGET, uses this net service name to connect directly to the auxiliary database instance.

Double check using the following commands on the TARGET and AUXILIARY

% tnsping <target_db>
        % tnsping <auxiliary_db>
      Once this is executed for the TARGET and AUXILIARY, it should return the same 'connect' information.

Additional Note:

If the listener is started from a different environment to rman client
        Then server processes will try to resolve the AUXILIARY service name using $TNS_ADMIN/tnsnames.ora where the value of TNS_ADMIN is as set in the 'listener runtime environment'.

相关参考
    RMAN 数据库克隆文件位置转换方法

基于RMAN的异机数据库克隆(rman duplicate)

基于 RMAN 的同机数据库克隆

基于用户管理的同机数据库克隆

Oracle 冷备份

Oracle 热备份

Oracle 备份恢复概念

Oracle 实例恢复

Oracle 基于用户管理恢复的处理

SYSTEM 表空间管理及备份恢复

SYSAUX表空间管理及恢复

Oracle 基于备份控制文件的恢复(unsing backup controlfile)

RMAN 概述及其体系结构

RMAN 配置、监控与管理

RMAN 备份详解

RMAN 还原与恢复

RMAN catalog 的创建和使用

基于catalog 创建RMAN存储脚本

基于catalog 的RMAN 备份与恢复

RMAN 备份路径困惑

自定义 RMAN 显示的日期时间格式

只读表空间的备份与恢复

Oracle 基于用户管理的不完全恢复

理解 using backup controlfile

使用RMAN实现异机备份恢复(WIN平台)

使用RMAN迁移文件系统数据库到ASM

基于Linux下 Oracle 备份策略(RMAN)

Linux 下RMAN备份shell脚本

使用RMAN迁移数据库到异机

RMAN 提示符下执行SQL语句

Oracle 基于 RMAN 的不完全恢复(incomplete recovery by RMAN)

rman 还原归档日志(restore archivelog)