Additional Information and Tips › Manually Recover a Failed Server-Move IP Address

Manually Recover a Failed Server-Move IP Address

If IP Redirection is used, you must remove the IP address manually. You cannot use this method for scenarios that do not support Move IP redirection (Hyper-V HA, CS HA).

To recover a failed server using Move IP Address redirection method

1. Boot the Master server without a network connection, to avoid IP conflicts.
2. From the TCP/IP properties dialog, remove the additional IP address.
3. Reboot the server and reconnect to the network.
4. If it is not already running, start the scenario from the Manager. If automatic reverse replication was set to On, the scenario runs in backward mode so that the Replica server is now active and the Master server is on standby.
5. Wait for synchronization to complete.
6. Perform a manual switchover to return the active role to the Master server. It is recommended that you do so outside of normal business hours.

Troubleshooting Oracle Databases

Oracle database fails to start after switchover

Symptom:

My vCenter Server HA scenario uses an Oracle database. After switchover, the Oracle database does not restart and I get the following errors:

• ER00319, 83, Error, <HOSTNAME><DATE/TIME>, Switchover Aborted
• ER00303, 82, Error, <HOSTNAME><DATE/TIME>, Unable to start vCenter Services
• ER00360, 81, Error, <HOSTNAME><DATE/TIME>, Unable to start vCenter after switchover. Service 'vpxd' not started

Solution:

These errors occur when the Oracle database does not successfully mount the following switchover. Use the command line to solve the problem:

1. Shut down

   [ORACLE_HOME]\bin\oradim.exe -shutdown -sid orcl -usrpwd * -shutmode immediate
   

2. Start again

   [ORACLE_HOME]\bin\oradim.exe -startup -sid orcl -usrpwd * -nocheck 0
   

Cannot Connect Oracle Enterprise Manager After Switchover

Symptom

After a switchover or failover occurs, the Oracle Enterprise Manager on the Replica is not accessible.

Solution

Reconfigure the Database Console. Before you begin, stop the running scenario or suspend the Is-Alive check. The following procedure is an example. Provide the correct values for <oracle_db_instance_name>, Oracle Database Version (the example uses 11g R2), file paths, and SID for your environment.

1. set oracle_home=D:\app\administrator\product\11.2.0\dbhome_1
2. set oracle_sid=orcl
3. emca -deconfig dbcontrol db -repos drop
4. To verify the TEMP DB already exists, perform the following steps:
   1. sqlplus /nolog
   2. connect / as sysdba
   3. select * from dba_temp_files where tablespace_name = 'TEMP';

   If an error occurs, perform these steps:

   • alter database temp file 'D:\APP\ADMINISTRATOR\ORADATA\ORCL\TEMP01.DBF' drop including datafiles;
   • alter tablespace temp add tempfile 'D:\APP\ADMINISTRATOR\ORADATA\ORCL\TEMP01.DBF';
5. emca -config dbcontrol db -repos create
6. net stop OracleDBConsole<oracle_db_instance_name>
7. net start OracleDBConsole<oracle_db_instance_name>