Closed SantoDE closed 5 years ago
Would be fixed by #246
@SantoDE @mmatur can we confirm that 246 fixed this?
If so, we can re-add GKE to our supported platform list
@dtomcej I just tried with 246 merged and still get an error. I'm using GKE 1.14 if that matters:
$ helm install --name=maesh --namespace=maesh maesh/maesh --set kubedns=true
Error: Job failed: BackoffLimitExceeded
$ k logs -f maesh-prepare-dlqw2
time="2019-09-24T15:02:49Z" level=info msg="Checking Cluster..."
2019/09/24 15:02:49 command prepare error: error during cluster check: coredns does not exist in namespace kube-system
From looking at the tiller logs, I see:
[tiller] 2019/09/24 15:06:58 preparing install for maesh
[storage] 2019/09/24 15:06:58 getting release history for "maesh"
[tiller] 2019/09/24 15:06:58 rendering maesh chart using values
[tiller] 2019/09/24 15:06:58 performing install for maesh
[tiller] 2019/09/24 15:06:58 executing 4 pre-install hooks for maesh
[tiller] 2019/09/24 15:06:58 deleting pre-install hook maesh-prepare for release maesh due to "before-hook-creation" policy
[kube] 2019/09/24 15:06:58 Starting delete for "maesh-prepare" ServiceAccount
[kube] 2019/09/24 15:06:58 building resources from manifest
[kube] 2019/09/24 15:06:58 creating 1 resource(s)
[kube] 2019/09/24 15:06:58 warning: failed to load schema: SchemaError(io.k8s.api.extensions.v1beta1.RunAsGroupStrategyOptions): invalid object doesn't have additional properties
[kube] 2019/09/24 15:06:58 Watching for changes to ServiceAccount maesh-prepare with timeout of 5m0s
[kube] 2019/09/24 15:06:58 Add/Modify event for maesh-prepare: ADDED
[tiller] 2019/09/24 15:06:58 deleting pre-install hook maesh-prepare-role for release maesh due to "before-hook-creation" policy
[kube] 2019/09/24 15:06:58 Starting delete for "maesh-prepare-role" ClusterRole
[kube] 2019/09/24 15:06:58 building resources from manifest
[kube] 2019/09/24 15:06:58 creating 1 resource(s)
[kube] 2019/09/24 15:06:58 warning: failed to load schema: SchemaError(io.k8s.api.extensions.v1beta1.RunAsGroupStrategyOptions): invalid object doesn't have additional properties
[kube] 2019/09/24 15:06:58 Watching for changes to ClusterRole maesh-prepare-role with timeout of 5m0s
[kube] 2019/09/24 15:06:58 Add/Modify event for maesh-prepare-role: ADDED
[tiller] 2019/09/24 15:06:58 deleting pre-install hook maesh-prepare for release maesh due to "before-hook-creation" policy
[kube] 2019/09/24 15:06:58 Starting delete for "maesh-prepare" ClusterRoleBinding
[kube] 2019/09/24 15:06:58 building resources from manifest
[kube] 2019/09/24 15:06:58 creating 1 resource(s)
[kube] 2019/09/24 15:06:58 warning: failed to load schema: SchemaError(io.k8s.api.extensions.v1beta1.RunAsGroupStrategyOptions): invalid object doesn't have additional properties
[kube] 2019/09/24 15:06:58 Watching for changes to ClusterRoleBinding maesh-prepare with timeout of 5m0s
[kube] 2019/09/24 15:06:58 Add/Modify event for maesh-prepare: ADDED
[tiller] 2019/09/24 15:06:58 deleting pre-install hook maesh-prepare for release maesh due to "before-hook-creation" policy
[kube] 2019/09/24 15:06:58 Starting delete for "maesh-prepare" Job
[kube] 2019/09/24 15:06:58 Using reaper for deleting "maesh-prepare"
[kube] 2019/09/24 15:06:58 building resources from manifest
[kube] 2019/09/24 15:06:58 creating 1 resource(s)
[kube] 2019/09/24 15:06:58 warning: failed to load schema: SchemaError(io.k8s.api.extensions.v1beta1.RunAsGroupStrategyOptions): invalid object doesn't have additional properties
[kube] 2019/09/24 15:06:58 Watching for changes to Job maesh-prepare with timeout of 5m0s
[kube] 2019/09/24 15:06:58 Add/Modify event for maesh-prepare: ADDED
[kube] 2019/09/24 15:06:58 maesh-prepare: Jobs active: 0, jobs failed: 0, jobs succeeded: 0
[kube] 2019/09/24 15:06:58 Add/Modify event for maesh-prepare: MODIFIED
[kube] 2019/09/24 15:06:58 maesh-prepare: Jobs active: 1, jobs failed: 0, jobs succeeded: 0
[kube] 2019/09/24 15:07:00 Add/Modify event for maesh-prepare: MODIFIED
[tiller] 2019/09/24 15:07:00 warning: Release maesh pre-install maesh/templates/hooks/prepare-hook.yaml could not complete: Job failed: BackoffLimitExceeded
[tiller] 2019/09/24 15:07:00 failed install perform step: Job failed: BackoffLimitExceeded
@donnyv12, the code was merged, but it is not in a public release yet. You would have to use a custom built image to test.
@dtomcej yes we can confirm. When we developed that, I tested it on GKE with a private image. :)
@donnyv12 would you be willing to test? I could provide you with a private image to test.
@SantoDE I would be happy to test it. Let me know where I can grab the private image.
It's working from my side without custom image, since the PR is merged. :)
Fixed by #246
Proposal
There are multiple options I like to discuss:
Other options?
Background
At the moment, GKE does nut support running Core-DNS. As Core-DNS is a requirement for Maesh in order to run, we're currently not supporting GKE. We want to change that.