If you set maxWorkers in the config to 8, and then manually clone 10 spot workers, the backend will not create any new workers for regional analyses on new networks. It will however report to the UI with a message saying it is starting workers (while logging on the backend that it is not starting workers and that the new job will never finish).
We could keep track of the number of automatically launched workers separately from the total amount of workers - or just set the total limit relatively high and rely on AWS reporting / dashboards to detect anomalous growth in number of workers.
We should also change it so that when no workers are started, we throw a very visible exception.
If you set maxWorkers in the config to 8, and then manually clone 10 spot workers, the backend will not create any new workers for regional analyses on new networks. It will however report to the UI with a message saying it is starting workers (while logging on the backend that it is not starting workers and that the new job will never finish).
We could keep track of the number of automatically launched workers separately from the total amount of workers - or just set the total limit relatively high and rely on AWS reporting / dashboards to detect anomalous growth in number of workers.
We should also change it so that when no workers are started, we throw a very visible exception.