Run an application in a cluster? then read on. Want to get thread dumps and more of your app? read on.
These features were built in response to real world needs from the high end users we have on the platform. They run things at an impressive scale and have all sorts of “good to have” challenges (when your app is popular).
Fabian recently released some new SDK commands for dealing with specific instances of an application running in a cluster. You may want this if you suspect one instance is misbehaving, and you want to find out why (or just replace that bit without replacing your whole cluster). An “instance” is a running process of your app on a specific server.
For example - to list instances of your app:
To restart JUST that instance:
Replace just an instance (if you think it is bad, good for large cluster):
You can tail the log of a specific instance. You can even get a stackdump by sending your process a SIGQUIT (and do more):
bees app:instance:invoke acmeiiiaaa cs send_sigquit
want more, read the docs:
This is a powerful feature for people running serious clusters (you know who you are !).