Is galaxy capable of autoscaling based on certain metrics? According to the demo request page it is “just click to instantly increase or decrease your app’s containers.” Does that mean that if usage spikes for one reason or another someone should be standing by to increase the number of containers or can you increase the container count if, say, the CPU usage of a container is above 60% for > 5 minutes?
Has there been any progress on this? I’m using modulus.io at the which does have this capability, but our app is quickly growing and I’m looking to move to galaxy soon for the enterprise support.
Not to my knowledge. I ended up setting up my own scalable infrastructure.
what is the prescribed way to autoscale with Galaxy currently, even if it isn’t through the webpanel? There doesn’t seem to be an API to to get metrics and increase the # of containers. I’m surprised given galaxy has been out for a year now. So the best you can currently do is pay for a lot more containers than you need to be sure??
Well, you don’t pay until you start them up
Yeah, but having a trigger to start up those containers would definitely be nice. I like the idea by @faceyspacey, where an API could be polled to read the metrics and used to programmatically spin up new containers. @khamoud, is your scalable solution built on top of Galaxy or have you resorted to using purely your own containers?
I went with docker swarm and consul for my particular needs.