DRBD Server Shutdown and Restart sequence
This applies to operators who host their own in-house servers running dual-server redundancy using DRBD and Heartbeat. The main active server is the Primary and redudant server is Secondary. Depending on the operators configuration, they may have a VMware environment active with a number of servers. These servers are normally a Respax Webserver and a Remote Desktop Environment running Respax Taskmaster.
Please note: If Heartbeat is active on the secondary server, you must shutdown Secondary first and then the Primary. When powering up, start Primary first. This is to ensure that the redundant Secondary server doesn't take over as the Primary.
The shutdown task is done via Console or via a SSH Client session on port 22. Shutdown the virtual machines first and then the actual servers. You must have root privileges to perform these task either through the root user or through the sudo list. You must also make sure that all Respax users are logged out from the Respax Application.
Shutdown Process
1. If applicable, login to the virtual Microsoft Remote Desktop environment as administrator. Ensure all users are logged off and shutdown the server.
2. Login via SSH and sudo to root on the Respax Webserver, use the following commands to achieve root and to shutdown the server:
sudo su - shutdown -h now
If you are watching on the console, you should see all process gradually shutting down. If the virtual environment is running, it may take a few minutes to shutdown.
3. Repeat 2 on the Secondary server.
4. Ensure all users are logged off and Repeat 2 on the Primary server.
The hardware should power itself down.
Startup Process
1. Power up the Primary Server. Once it has completed its boot process, all relevant Virtual servers should eventually start themselves up.
2. Once Primary has booted, Power up Secondary.