Hey i've got a bit of an unusual one. We're running pulumi inside Temporal workers (via automation API) and for temporal to know if the worker is dead/stuck we trigger "heartbeat" on every stdout line of pulumi. Since pulumi outputs periodic

.

as it works, this usually works but we still occasionally see heartbeat timeouts even with relatively long timeouts (3 minutes) Are there any other ways to hook into pulumi to know if it is still "alive"? some kind of healthcheck we can do on it? The coarse solution would be to just check if the pulumi process is still running at all...

Interesting question! I don't think there is anything like a healthcheck that exists. I found eventstreams in the go SDK but not sure if that is more granular than the progress dots. Are you in a container?

Yeah slowly widening the timeouts has been the strategy for now. We're in a container and using a temporal typescript worker to call typescript pulumi (via inline function)