kubernetes-retired/kube-aws

Config template error on init

USA-RedDragon opened this issue · 7 comments

USA-RedDragon commented

I seem to be having an issue during the init process.

Running the command (redacted for privacy):

kube-aws init \
  --cluster-name=kube-aws-test \
  --external-dns-name=kube.ourcompany.com \
  --hosted-zone-id=Z53861T2NU46E2 \
  --region=us-east-1 \
  --availability-zone=us-east-1a \
  --key-name=kubernetes \
  --kms-key-arn="arn:aws:kms:us-east-1:123456789012:key/a1111aa-11a1-1a11-111a-a1a11111aa1a" \
  --s3-uri=s3://kube-aws-test/

gives the output of:

Error: error exec-ing default config template: Error exec-ing default config template: template: cluster.yaml:1658:57: executing "cluster.yaml" at <.Config.AdminAPIEndpoint.DNSName>: can't evaluate field Config in type config.InitialConfig

in v0.15.0. Downgrading to v0.14.3 resolved the issue.

sdarwin commented

Also got this error.

Error: error exec-ing default config template: Error exec-ing default config template: template: cluster.yaml:1658:57: executing "cluster.yaml" at <.Config.AdminAPIEndpoint.DNSName>: can't evaluate field Config in type config.InitialConfig

sei-nicolas commented

Same error here. I know very little of this code structure/looked very fast so take what follows cautiously, but I'm under impression that in https://github.com/kubernetes-incubator/kube-aws/blob/master/builtin/files/cluster.yaml.tmpl at line 1658, ".Config.AdminAPIEndpoint.DNSName" should be replaced with ".ExternalDNSName"

I guess ".Config.AdminAPIEndpoint.DNSName" is valid (only) in plugins templates (e.g. https://github.com/kubernetes-incubator/kube-aws/blob/master/builtin/files/plugins/dashboard/plugin.yaml) where the templater has a full "Config" struct, while cluster.yaml.tmpl is handled with only the InitialConfig struct.

sei-nicolas commented

=> Actually, I checked that this fixed the issue/was test-passing (and got stuck with the PR, see #1824, sorry for the mess)

weisjohn commented

confirmed v0.14.3 fixes this issue

davidmccormick commented

We have another urgent fix for the v0.15.x branch so we'll put both of these out together in a 0.15.1 release.

fejta-bot commented

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

dominicgunn commented

This was fixed by @davidmccormick.