Skip to content
This repository was archived by the owner on Jan 9, 2020. It is now read-only.

Conversation

@kimoonkim
Copy link
Member

Fixes #448.

Copy link

@ash211 ash211 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Seems generally right, though I think I remember @foxish having an opinion about service accounts previously.

@mccheah is this the right step to make this change in?

.editOrNewSpec()
.withServiceAccount(account)
.withServiceAccountName(account)
.endSpec()
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

indent these in a couple spaces for each sub-item. Even though the methods are all straight chained, there's a structure we've been trying to keep clear through indentation

}.getOrElse(driverSpec.driverPod)
}.getOrElse(
driverServiceAccount.map {
account =>
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

move account onto previous line to match style in rest of this file

@erikerlandson
Copy link
Member

This logic implements a policy that kube secret credentials override service acct, and also prevent service acct from being applied. Is that the right policy?

@mccheah
Copy link

mccheah commented Aug 22, 2017

I think this is the right policy.

@ash211
Copy link

ash211 commented Aug 22, 2017

I thought we had validation that you can't apply both secret creds and a service account -- you have to pick one or the other

@mccheah
Copy link

mccheah commented Aug 22, 2017

I think either is fine really, but we should have preference for the non-service account based authentication.

@erikerlandson
Copy link
Member

I was imagining what amounts to supplementing permissions from one with the other, but that might be a corner case.

@foxish foxish self-assigned this Aug 23, 2017
@erikerlandson
Copy link
Member

It sounds like the current logic is restoring the original logic that was clobbered - LGTM

@kimoonkim
Copy link
Member Author

It seems everybody is fine with this, although it isn't approved yet. Maybe we can approve and merge this soon. I also wonder if we need to have a dot release in case more people start to use Kubernetes 1.6.

@erikerlandson
Copy link
Member

@foxish, you're assigned, do you want to merge?

@foxish
Copy link
Member

foxish commented Aug 24, 2017

LGTM

@foxish foxish merged commit e600a07 into apache-spark-on-k8s:branch-2.2-kubernetes Aug 24, 2017
@kimoonkim kimoonkim deleted the override-service-account branch August 24, 2017 20:59
@kimoonkim
Copy link
Member Author

Thanks everyone for the reviews!

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants