Hi, I would like to know what are our existing shutdown procedure for IoTManager in openshift pod.
When openshift draining a node, it will send SIGTERM to pod, which will initiate shutdown.
For example, if IoT Controller send a REST POST command and waiting for reply. While waiting, SIGTERM received, will the IoTManager wait for the reply to be received before terminating driver or the controller? Or it will initiate the shutdown immediately?
Thanks
Hi, no, Sigterm is a process level operation and the rest driver is an application level one, currently there’s no mechanism to broadcast the sigterm to the components to initiate a graceful shutdown. I think this is a valuable input and would suggest you open a ticket to express this concern.
Hi Joao Roque, thanks for your answer.
In another partially related question, is IoTManager currently unsupported to configure and use liveness, readiness, and startup probes in OpenShift as well?
No, as far as I know, right now there’s none configured