Scaling druid with auto-scaling group

So is it possible to use aws auto-scaling group to scale up and down: broker, historical node, and middle manager, based on cpu or memory

Hi mostafa,

You should be able to spin up any number of nodes for broker, historical and middle manager using auto-scaling.



Hey prathamesh,

scaling up and down will not break anything, what if aws ASG decided to scale down and removed an instance that was handling a query or something, that might make it fall right, have you used aws auto-scaling group with druid?