The network’s monitor routine verifies child NdioBoard component(s)—the “pingable” device equivalent in Ndio. For general information, see “About Monitor” in the Drivers Guide.
The Ndio ping monitor provides added functionality that allows the Ndio driver to reboot the JACE if serious problems are detected with I/O functionality. This Ndio reboot logic is configurable in several properties under the Monitor slot of the NdioNetwork, as shown in Figure 3.
The reboot-related Monitor properties include the following:
-
Reboot Enabled
If set to true (default), and the NdioNetwork is enabled, the JACE reboots whenever Ndio reboot logic detects a serious issue. (If NdioNetwork is not running, only the Ndio processor is reset).
-
Reboot Period
Default is 30 minutes. Defines the amount of time in which the maximum number of reboots can occur. If more than the configured number of reboots occurs within this period, the network’s Enabled property is set to false, and the JACE is rebooted.
-
Max Reboot Period Count
Default is 3. Specifies the maximum number of reboots that can occur within the reboot period (above) before Ndio is disabled.
-
Total Reboot Request
(Read-only status) Specifies how many times Ndio reboot logic has rebooted the JACE.
-
Last Reboot Request
(Read-only status) Provides a timestamp of when Ndio reboot logic last rebooted the JACE.
As of build 3.0.88, the Ndio driver includes added “reboot logic” to automatically correct conditions where the JACE or Ndio components stop communicating effectively. This logic allows the Ndio processor(s) to behave in a predictable fashion in the case of a problem.
The following scenarios are addressed:
-
Station stops communicating with Ndio daemon.
By design, once the station starts communicating with the Ndio daemon, it must continue to communicate. If communication fails or is halted, the Ndio daemon stops talking to the Ndio processor, and the Ndio processor is reset. If the NdioNetwork is still running, this is detected, and the JACE is rebooted. If the NdioNetwork is not running, the Ndio processor resets itself and stays in that state until the next reboot cycle.
-
Ndio daemon stops communicating with the Ndio processor.
If the Ndio daemon fails to communicate with the Ndio process for any reason after talking to it at least once, the Ndio processor will reset. If an NdioNetwork is in the running station, this reset may result in a JACE reboot.
-
The Ndio processor stops responding to requests from the Ndio daemon.
If the Ndio processor resets or stops responding, the results in an error condition reported to the NdioNetwork, which may choose to reboot the JACE.
Whenever any Ndio-requested reboot is about to occur, the station is automatically saved, and messages are sent to the event
log prior to rebooting. Additionally, an alert is generated.
Under the NdioNetwork’s Monitor slot, several properties are available to configure this Ndio reboot logic (see Figure 3).