eksctl v0.10.1 - Failure in creating nodegroup from config file

[jchoi926]

What happened?
eksctl create nodegroup --config-file=<config-file> results in [✖] getting nodegroup stack summaries: mapping stack to nodegroup summary: error collecting Cloudformation outputs for stack <nodegroup>: no output "InstanceRoleARN" in stack "<nodegroup>"

What you expected to happen?
Successfully create nodegroup

How to reproduce it?
Run the above command

Anything else we need to know?
config file:
```nodeGroups:

• name:
  ami:
  instanceType:
  desiredCapacity: 1
  minSize: 1
  maxSize: 1
  volumeSize: 100
  volumeType: gp2
  privateNetworking: true
  labels:
  environment:
  nodegroup-type:
  ssh:
  publicKeyName: ```

Able to succesfully create the nodegroup with eksctl version 0.9.0

Versions
Please paste in the output of these commands:

$ eksctl version `[ℹ]  version.Info{BuiltAt:"", GitCommit:"", GitTag:"0.10.1"}`
$ kubectl version `Client Version: version.Info{Major:"1", Minor:"13+", GitVersion:"v1.13.7-eks-fa4c70", GitCommit:"fa4c703fa1706fac99d00d7f1f2080e9d4cc7eed", GitTreeState:"clean", BuildDate:"2019-07-23T03:14:24Z", GoVersion:"go1.11.5", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"14+", GitVersion:"v1.14.6-eks-5047ed", GitCommit:"5047edce664593832e9b889e447ac75ab104f527", GitTreeState:"clean", BuildDate:"2019-08-21T22:32:40Z", GoVersion:"go1.12.9", Compiler:"gc", Platform:"linux/amd64"}`

Logs
Include the output of the command line when running eksctl. If possible, eksctl should be run with debug logs. For example:

eksctl create nodegroup --config-file <config-file> [ℹ] eksctl version 0.10.1 [ℹ] using region <region> [ℹ] will use version 1.14 for new nodegroup(s) based on control plane version [ℹ] 1 nodegroup(s) that already exist (nodegroup) will be excluded [ℹ] nodegroup "<nodegroup>" will use "<ami>" [AmazonLinux2/1.14] [ℹ] using EC2 key pair "<keypair>" [ℹ] 1 nodegroup (<nodegroup>) was included (based on the include/exclude rules) [ℹ] combined exclude rules: <nodegroup> [ℹ] no nodegroups present in the current set were excluded by the filter [ℹ] will create a CloudFormation stack for each of 1 nodegroups in cluster "<cluster>" [ℹ] 2 parallel tasks: { create nodegroup "<nodegroup>", no tasks } [ℹ] building nodegroup stack "<stack>" [ℹ] deploying stack "<stack>" [✖] unexpected status "ROLLBACK_IN_PROGRESS" while waiting for CloudFormation stack "<stack>" [ℹ] fetching stack events in attempt to troubleshoot the root cause of the failure [ℹ] 1 error(s) occurred and nodegroups haven't been created properly, you may wish to check CloudFormation console [ℹ] to cleanup resources, run 'eksctl delete nodegroup --region=<region> --cluster=<cluster> --name=<name>' for each of the failed nodegroup [✖] waiting for CloudFormation stack "<stack>": ResourceNotReady: failed waiting for successful resource state [✖] failed to create nodegroups for cluster "<cluster>"

[manumatico]

Got the exact same issue here, using the same eksctl version (0.10.1), with and additional problem: now I can't manage my eks cluster at all. Any idea of how to fix this, other than creating a new cluster?

eksctl get nodegroups --cluster=<clustername>

[✖]  getting nodegroup stack summaries: mapping stack to nodegroup summary: error collecting Cloudformation outputs for stack <stackname>: no output "InstanceRoleARN" in stack "<stackname>"

[moonnoon]

Same issue.
Using 0.9.0 is fine.

[jchoi926]

Got the exact same issue here, using the same eksctl version (0.10.1), with and additional problem: now I can't manage my eks cluster at all. Any idea of how to fix this, other than creating a new cluster?

eksctl get nodegroups --cluster=<clustername>

[✖]  getting nodegroup stack summaries: mapping stack to nodegroup summary: error collecting Cloudformation outputs for stack <stackname>: no output "InstanceRoleARN" in stack "<stackname>"

Install eksctl version 0.9.0. That should resolve your issue.

[treacher]

Might be worth seeing if 0.10.2 fixes your issue, it fixed it for me.

[forsberg]

Unfortunately I just got this on 0.10.2. Nodegroup was created with 0.10.20.

[PierreBeucher]

Same issue here with commands such as eksctl --cluster my-cluster get ng gives output:

Error: getting nodegroup stack summaries: mapping stack to nodegroup summary: error collecting Cloudformation outputs for stack [stack-name]: no output "InstanceRoleARN" in stack "[stack-name]"

Happens while a stack is being created by another instance of eksctl and the related NodeInstanceRole is not yet created in CloudFormation. This actually prevents operations on the cluster via eksctl while a nodegroup is being created

Reproduced on both 0.10.2 and 0.14.0

[tyg03485]

eksctl version: 0.13.0 (also, 0.15.0)

In my case, the attachPolicyARNs setting was wrong And After this problem, i cant manage the cluster like get all node group, update cluster.

Error: getting nodegroup stack summaries: mapping stack to nodegroup summary: error collecting Cloudformation outputs for stack [stack-name]: no output "InstanceRoleARN" in stack "[stack-name]"
or
[✖]  found 1 nodegroup(s) (stack name) without shared security group, cluster networking maybe be broken
[✖]  it's recommended to create new nodegroups, then delete old ones

[Solved]
Just Go AWS CloudFormation Console and Delete stack of node group which has problem.
FYI) eksctl delete nodegroup command not work for me.

[mellonbrook]

I built a 1.14 EKS cluster with eksctl version 0.18.0.

I am now using eksctl version 0.20.0.

This command is failing:
➜ ~ eksctl get nodegroups --cluster=staging-kube
Error: getting nodegroup stack summaries: getting nodegroup stacks: no eksctl-managed CloudFormation stacks found for "staging-kube"

The cluster was created with eksctl:

Do I need to downgrade my eksctl?

Update: Downgrading to 0.18.0 did not work. Maybe I created my cluster with another version. Or is the eksctl version even the problem here?

[michaelbeaumont]

@mellonbrook that issue and error message don't seem to be related to this one, did you have the error mentioned in the description

[michaelbeaumont]

@tyg03485 do you happen to recall what the problem with deleting the nodegroup was?

[michaelbeaumont]

@PierreBeucher Can you confirm that you're trying to call get ng while that nodegroup is still being created? Feel free to create another issue for your specific use case.

[michaelbeaumont]

As it stands, this issue is not reproducible. If anyone encounters the _specific message_ (error collecting Cloudformation outputs for stack <nodegroup>: no output "InstanceRoleARN" in stack "<nodegroup>") in the OP, please include the _entire config_, the log output _including_ the error as well as the _exact steps_ leading to the error.

For other errors, create a new issue with the _entire config_, the log output _including_ the error as well as the _exact steps_ leading to the error.

[relgames]

Hi, having the same issue on eksctl 0.21.0

 eksctl delete nodegroup --region=us-east-2 --cluster=test1 --name=test-workers-spot
[ℹ]  eksctl version 0.21.0
[ℹ]  using region us-east-2
[ℹ]  combined include rules: test-workers-spot
[ℹ]  1 nodegroup (test-workers-spot) was included (based on the include/exclude rules)
[ℹ]  will delete 1 nodegroups from auth ConfigMap in cluster "test1"
[!]  error getting instance role ARN for nodegroup "test-workers-spot": couldn't get iam configuration for nodegroup "test-workers-spot" (perhaps state "ROLLBACK_COMPLETE" is transitional): no output "InstanceRoleARN" in stack "eksctl-test1-nodegroup-test-workers-spot"

Update: found the reason, it happened when tried to create new nodegroup and there is no public network in the cluster, so the cloudformation stack failed. Deleted the stack and eksctl works again.

https://github.com/weaveworks/eksctl/issues/543