High Availability Controller (HAC) is an automatic redundancy management program for solidDB HotStandby (HSB) configurations. It maintains the availability of the database service by detecting failures, performing failovers to standby units, and restarting failed processes when necessary.
A failure can be caused by a hardware problem in a database server, a database process failure, or a broken HSB link. HAC monitors the HSB states of the servers and, in the case of a failure, ensures that the server that is not affected by the failure holds the primary role and that the server is ready to accept the transaction load.
In other words, HAC plays the role of a watchdog program. In its implementation, the solidDB event mechanism is used to monitor the server states. Every time the state of the HSB server changes, it sends an event to HAC, which then deduces potential needs for actions that are executed by using the HSB admin command API.
HAC has two main purposes:
1 HAC can be used as a watchdog to automatically maintain the availability of the database service. In this mode (AUTOMATIC mode), HAC performs the following actions:
▪ starts, restarts, and terminates the database server processes (optional),
▪ monitors the state of the servers and the HSB link between them,
▪ determines and performs the required steps to maintain the database availability.
2 HAC can also play the role of a monitor for an HSB system by monitoring the state changes of the HSB servers (triggered by something else) and reporting the status of the system. In this mode (ADMINISTRATIVE mode), HAC does not execute any HSB state transitions or otherwise modify the HSB system.
A sample graphical user interface (GUI) component called High Availability Manager (HAM) is also available for HAC.
