solidDB Help : Samples : HotStandby sample : Watchdog sample application : How the Watchdog sample application works : Dual primary servers
  
Dual primary servers
The following scenario demonstrates how a watchdog might attempt to recover an HSB system but end up with both servers in a primary role.
This issue should not occur if the HotStandby.AutoPrimaryAlone parameter is set to No and the watchdog is configured to handle only one failure at a time (that is, after a failure, the watchdog must determine that the HSB environment is working successfully again before it attempts to handle another failure).
In this example, Server #1 is initially the primary server and Server #2 is initially the secondary server.
 
Description
Illustration
A network failure occurs and Server #1 becomes inaccessible.
Server #1 cannot reach Server #2 and switches to PRIMARY UNCERTAIN state automatically.
Server #2 cannot reach Server #1 and switches to SECONDARY ALONE state automatically.
The diagram is described in the first column of the table
The watchdog concludes that Server #1 has failed and switches Server #2 state from SECONDARY ALONE to PRIMARY ALONE.
The diagram is described in the first column of the table
A second network failure occurs, and Server #2 becomes inaccessible.
The diagram is described in the first column of the table
The first network failure is repaired, and Server #1 becomes accessible again.
The watchdog, seeing that Server #1 is accessible and Server #2 is not, switches Server #1 state to PRIMARY ALONE.
The diagram is described in the first column of the table
The second network failure is fixed and Server #2 becomes accessible again.
Both Server #1 and Server #2 are in the PRIMARY ALONE state.
The diagram is described in the first column of the table
Go up to
How the Watchdog sample application works