DCS-10001:Internal error encountered: Failed to create the database

上周连续处理了好几个ODA的奇葩问题，说是ODA的问题，其实不然，都是人为原因导致，言归正传，我们来聊聊这一个dcs-10001的问题：

客户环境为ODA X9-2-HA 加强版，imageinfo 19.16 高校使用。

通过ODA UI创建集群数据库提示DCS-10001:Internal error encountered: Failed to create the database，起初我定位是有DCS故障引发，故检查了DCS与zk等日志，但一无所获，

检查项目包括：

dcs-agent.log zk.log等

随后查询后台知识库：

1、建议检查系统参数kernel.sem ，这些参数并未被修改，保持刷机时的状态；

 2023-07-17 18:49:30,346 DEBUG [Thread-10374] [] c.o.d.c.u.CommonsUtils: Output :

Look at the log file "/u01/app/oracle/cfgtoollogs/dbca/dirxnatp/dirxnatp1.log" for further details. 

2023-07-17 18:49:28,936 DEBUG [Thread-10374] [] c.o.d.c.u.CommonsUtils: Output :

[WARNING] ORA-27154: post/wait create failed

2023-07-17 18:49:28,936 DEBUG [Thread-10374] [] c.o.d.c.u.CommonsUtils: Output :

ORA-27302: failure occurred at: sskgpbitsper 

2023-07-17 18:49:28,936 DEBUG [Thread-10374] [] c.o.d.c.u.CommonsUtils: Output : EMPTY CONTENT 

2023-07-17 18:49:28,937 DEBUG [Thread-10374] [] c.o.d.c.u.CommonsUtils: Output :

[FATAL] ORA-01034: ORACLE not available ◄▬▬

/u01/app/oracle/cfgtoollogs/dbca/dirxnatp/dirxnatp.log

/u01/app/oracle/cfgtoollogs/dbca/dirxnatp/dirxnatp1.log

DCS-10001:Internal error encountered: Failed to create the database

2、BUG？

create missing audit directory with correct permission and start the 2nd instance.

$ORACLE_BASE/admin/DGC folder on ODA node 1, 'njodap3b' along with the subfolders as shown:'Bug 33783815' : LNX64-1914-CMT:TDE RAC DATABASE CREATION FAILED IN MIGRATED DCS ENV DUE TO MISSING AUDIT_FILE_DEST LOCATION IN NODE2

imageinfo 版本已解决，故排除bug原因。

3、检查/u01/app/oracle/cfgtoollogs/dbca log

发现错误信息如下：

    

[main] [ 2023-08-02 15:03:42.478 CST ] [CommandBuffer.checkCommandStatus:83]  CommandBuffer:checkCommandStatus returning false, command number= 1 thread name= Thread[main,5,main]

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterConfig.submit:590]  status=false

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterException.setErrorMessage:145]  ClusterException.setErrorMessage: commands[0]=oracle.ops.mgmt.command.file.PathExistCommand@74971ed9

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterException.setErrorMessage:156]  status : true

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterException.setErrorMessage:145]  ClusterException.setErrorMessage: commands[1]=oracle.ops.mgmt.command.file.PathExistCommand@131fcb6f

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterException.setErrorMessage:156]  status : false

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterException.setErrorMessage:184]  ClusterException.setErrorMessage: errString is 'PRKC-1191 : Remote command execution setup check for node oda2 using shell /usr

/bin/ssh failed.

FIPS mode initializedssh: connect to host oda2 port 22: Connection refused'

[main] [ 2023-08-02 15:03:42.478 CST ] [ClusterConfig.destroy:468]  destroying resources for client thread Thread[main,5,main]

[main] [ 2023-08-02 15:03:42.478 CST ] [Utils.getString:278]  ==========Str is oda2

[main] [ 2023-08-02 15:03:42.479 CST ] [OracleHome.getNodeNames:446]   exception checking oracle home existence on nodesPRKC-1027 : Error checking existence of file /u01/app/odaorahome/oracle/product/19.0.0

.0/dbhome_1 on oda2

[main] [ 2023-08-02 15:03:42.479 CST ] [OracleHome.getNodeNames:474]   no of nodes returned1

INFO: Aug 02, 2023 3:03:42 PM oracle.install.commons.net.support.DefaultSSHSupportManager <init>

INFO: Preparing Default SSH support manager

PRKN-1038 : The command "/usr/bin/ssh -o FallBackToRsh=no  -o PasswordAuthentication=no  -o StrictHostKeyChecking=yes  -o NumberOfPasswordPrompts=0  oda2 -n /bin/true" run on node "oda1" gave an unexpected

output: "FIPS mode initializedssh: connect to host oda2 port 22: Connection refused"

错误原因居然是22端口不通，经与客户核实，端口已被修改为非默认，后修改为默认22端口后通过ODA GUI 创建数据库正常。

总结：ODA产品某些默认配置不能修改。