Daniel

Did you ever find a workable solution for this? It appears i'm facing the same challenge

dang this should be merged but this is unmanaged nowadays is it?

We have this issue as well if we use self hosted runners hosted in europe. At the same time usiing the Github provided runners I don't get this :/ Anybody...

using `"@google-cloud/[email protected]":` here ``` Failed to add metadata entry A: Wed, 12 Feb 2020 20:34:43 GMT. Metadata key "a" contains illegal characters. For more information see https://github.com/grpc/grpc-node/issues/1173 Failed to add...

I'd like to refer to my comment in the other thread https://github.com/actions-runner-controller/actions-runner-controller/issues/1581#issuecomment-1181618360 basically that the pod gets the termination signal is not the problem (I think). I would suspect that...

default: `terminationGracePeriodSeconds: 30` (in seconds) :trollface: Not alot of time to finish a job :confused: That is the kubernetes side of things, mumoshu is expert on the ARC side of...

I would suggest he have some folks who run into this issue to test it out. I don't know how long the ARC would wait for a termination, if there's...

I just tried it myself. And at least in my setup the docker container immediately started shutting down (which is an issue, meaning we have to prevent k8s from sending...

I'm running with a 0 pdb to avoid unnecessary evitcs. i.e.: ``` apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: runner-16-ramdisk spec: maxUnavailable: 0 selector: matchLabels: runner-deployment-name: runner-16-ramdisk ``` The runners surely...

it looks like `terminationGracePeriodSeconds` can already be configured. @GoldenMouse can you try setting this to about the longest running job that you have? then the pod should get the chance...