Created: 2021-04-0710:32 PM. Last Modified: 2021-04-0710:49 PM
M580 hot standby initiate CPU swap-over when a primary rack Ethernet module has a fault
Swapping a hot standby primary system on the failure of a primary CPU rack NOC module.
I wanted to share code we developed locally for swapping the hot standby system if a NOC module in the primary rack was faulty as long as the standby system is OK.
I have placed extensive comments in the logic.
It is based on system which has two NOC modules in each rack. One pair of NOC modules (primary and Standby) manage an RSTP ring of drives and the other NOC module pair manage an RSTP ring of motor protection units.
If we have a primary ethernet module fault as long as there is no fault in the standby CPU rack i.e. CPU or ethernet modules we will initiate a swap-over of the primary CPU. Please note that any fault condition in the standby i.e. CPU has no connection to ERIO, a fault in the either NOC module, the CPU is not in run and in standby mode the logic will not swap over due to a NOC module fault.
The normal CPU swap over reasons i.e. failure of the primary CPU, primary CPU in stop, loss of comms to ERIO or loss of power to primary CPU will always trigger a swap-over to the standby. The ethernet logic does not impede this.
Please also note I did not use the IO scanner status as the failure of one device can cause this to fault and with 20 or 30 devices we still wish to communicate with as many as possible.