I have been looking through documentation and have not been able to locate a list of possible Critical Errors. Is there one that exists somewhere in the documentation? We are trying to handle certain critical errors differently and wanted to make sure we handle them appropriately. As of right now I am just blanket handling errors by bringing the service down but for ones such as unable to reach the Azure service due to loss of connection or error grabbing the message out of the queue i want to retry or possibly restart my service.
There’s no definitive list of errors that could happen, any exception could cause a critical failure if it repeats enough times to trigger the circuit breaker.
If the Critical Error is being triggered, restarting the service might not be enough to bring it back online. If the required infrastructure is still unavailable, restarting the service would result in the same behaviour. It sounds like the safe option is for the endpoint to shut itself down and to have the orchestration infrastructure (k8s?) decide whether or not to bring the endpoint back up. The orchestration infrastructure can decide to restart any other services that the endpoint could be dependent on.
Alternatively you could build some kind of “test” login inside the on critical error action to determine the correct response. Either way: there is no definitive list of errors that could trigger a critical error.
Thank you for getting back to me. I had pretty much come to that conclusion but still had my fingers crossed there may be a partial list.
I will look into the orchestration infrastructure a bit and some ideas we have thrown around in the office.