I’ve got my main meteor app including its mongoDB running on a big DO droplet - all good!
Additionally, there is a “node-worker”, running on its own, small droplet, which is connected to the same mongoDB.
From time to time, this worker is running cronjobs which can be very mongoDB-heavy - sometimes even several minutes long! While these are running, have the worker’s nodejs process at 100% cpu is just fine, but it obviously is a big problem that the mongo instance on the main droplet is also at 100% cpu for several minutes, as during this time, the whole app becomes completely unresponsive for every user!
–> What is the proper way to solve this? Somehow limit the mongo-operations per seconds for the worker? If so, how can I do that? Is there any other way?
This has to be a common problem! I hope someone can help!
Thanks a bunch, best regards, Patrick