dbca delete database hang ON AIX with 11.2.0.1 rac 2 nodes

oradbguy

     这套库反映dbca建库有问题，当时我上去发现dbca 时提示节点互信有问题，于是检查发现ssh有问题，后来确认到两个节点的 /etc/ssh/ssh_config 里面开启了 【X11Forwarding yes】。后来把两边节点的这个ssh参数注释掉后，就可以正常dbca，生成建库scripts。然后在OS上通过 scripts建库。
           建库脚本继续跑完，srvctl status database -d 检查了一下,（当时没有注意到node2上的实例没有startup），因为是测试dbca建库，于是用dbca 去delete database，但是到81%的地方dbca GUI就hang住了。仔细看上面srvctl status database -d的检查结果才发现 node2上的实例当时没有startup。

    检查$ORACLE_BASE/cfgtoollogs/dbca/trace.log_OraDb11g_home1
看到如图：大量的  [awt-eventqueue   databaselistdatasource.getColumncount start of function],搜了下这个关键字，和JAVA相关，不太明确。




关键字是AWT-EventQueue-0  DatabaseListDataSource->getColumnCount:Start of  function
抱歉我当时是用截图保存的信息。


   然后在自己的虚拟机测试环境LINUX X64 11.2.0.3上尝试dbca delete database
发现在【ClientResource Constructor:Blocking Semaphore owned by Thread-28:Acquire called by thread Thread-28 m_count=0】
下一步是【yncBufferFull:Acquire called by thread Worker 0 m_count=0】，通过对比猜测dbca delete database hang 就是在这步。但是通过google 没有找到故障原因，特此请教。

1. [Worker 0] [ 2013-05-05 01:30:23.503 CST ] [Semaphore.release:85]  ClientResource Constructor:Blocking Semaphore owned by Thread-28:Release called by thread Worker 0 m_count=1
   
2. [Worker 0] [ 2013-05-05 01:30:23.505 CST ] [ClusterConfig$ExecuteCommand.run:3069]  Released Semaphore by worker=Worker 0
   
3. [Thread-28] [ 2013-05-05 01:30:23.505 CST ] [ClusterConfig.block:608]  block acquired semnum=0
   
4. [Thread-28] [ 2013-05-05 01:30:23.506 CST ] [Semaphore.acquire:109]  ClientResource Constructor:Blocking Semaphore owned by Thread-28:Acquire called by thread Thread-28 m_count=0
   
5. [color=Red][Worker 0] [ 2013-05-05 01:30:23.506 CST ] [Semaphore.acquire:109]  SyncBufferFull:Acquire called by thread Worker 0 m_count=0[/color]
   
6. [Worker 1] [ 2013-05-05 01:30:23.813 CST ] [NativeSystem.isCmdScv:499]  isCmdScv: cmd=[/usr/bin/ssh -o FallBackToRsh=no -o PasswordAuthentication=no -o StrictHostKeyChecking=yes -o NumberOfPasswordPrompts=0 vn2 -n /bin/rm -rf /s01/orabase/diag/rdbms/vinst]
   
7. [Worker 1] [ 2013-05-05 01:30:23.815 CST ] [NativeSystem.isCmdScv:549]  isCmdScv: /usr/bin/ssh is present.
   
8. [Worker 1] [ 2013-05-05 01:30:23.815 CST ] [NativeSystem.isCmdScv:551]  isCmdScv: /usr/bin/ssh is a file.
   
9. [Worker 1] [ 2013-05-05 01:30:23.816 CST ] [NativeSystem.isCmdScv:568]  isCmdScv: returned true

复制代码

谢谢。

Maclean Liu(刘相兵

直接手动 删掉数据文件 和 /etc/oratab中的记录 即可

oradbguy

嗯，实际上也是这样删除的。

但是问题发生的诱因目前还没分析出来。