 | ||
| About resource monitoring | ||
|---|---|---|
| Prev | Introducing the Disaster Recovery Orchestrator agent for Microsoft SQL Server 2012 | Next |
| ||
| About resource monitoring | ||
|---|---|---|
| Prev | Introducing the Disaster Recovery Orchestrator agent for Microsoft SQL Server 2012 | Next |
Disaster Recovery Orchestrator employs an event-based monitoring framework to determine the status of the configured application and its components. This framework is called the Intelligent Monitoring Framework (IMF), and it is implemented using custom as well as native operating system-based notification mechanisms.
IMF provides instantaneous state change notifications. Disaster Recovery Orchestrator agents detect this state change and then trigger the necessary actions.
IMF provides the following key benefits:
Instantaneous notification
Faster fault detection results in faster failover and thus less application down time.
Reduction in system resource utilization
Conventional resource monitoring occurs every 60 seconds by default. With IMF event-based monitoring there is less reliance on conventional monitoring and so this interval can be increased. Thus Disaster Recovery Orchestrator reduces CPU utilization and provides significant benefits in terms of system resource utilization.
Ability to monitor large number of components
Due to the ability to increase conventional monitor cycle intervals, IMF allows monitoring of more components with a lower system resource utilization.
IMF is enabled by default for a component if its Disaster Recovery Orchestrator agent supports IMF.
The following steps outline how IMF-based monitoring works:
A Disaster Recovery Orchestrator agent waits for the components to report the same steady state (either Online or Offline) for two consecutive monitor cycles. Then, it registers the components for IMF-based monitoring.
The agent then registers itself for receiving certain operating system-specific or custom event notifications.
If a component fails, the agent executes a monitor cycle to determine its state. If the state is Offline, Disaster Recovery Orchestrator takes the necessary corrective action, depending on the configuration.
If the component state remains the same, the agent moves to a Wait state and then waits for the next event to occur.