EKS changes can trigger a deletion of the nodeAccess ConfigMap

[ribren]

What happened?

While running pulumi up to push some EKS changes (node capacity and type, and new tags for the cluster), pulumi deleted the $environment-$project-nodeAccess ConfigMap, which appears to be generated as part of an eks.Cluster object.

This problem appears intermittently. Sometimes a preview will give a replace, and sometimes a delete Deletion: Replacement:

Steps to reproduce

Attempt a preview or up with changes to an EKS cluster

Expected Behavior

The nodeAccess ConfigMap should remain the same between runs, or be updated when necessary.

Actual Behavior

The nodeAccess ConfigMap is either replaced or deleted.

Output of pulumi about

CLI          
Version      3.38.0
Go Version   go1.19
Go Compiler  gc

Plugins
NAME        VERSION
aws         4.38.1
docker      3.0.0
eks         0.31.0
kubernetes  3.5.2
nodejs      unknown
random      4.2.0

Host     
OS       darwin
Version  12.5.1
Arch     x86_64

This project is written in nodejs: executable='/usr/local/bin/node' version='v18.7.0'

Dependencies:
NAME                              VERSION
@pulumi/aws                       4.38.1
@pulumi/awsx                      0.30.0
@pulumi/eks                       0.31.0
@pulumi/pulumi                    3.9.1
@pulumi/random                    4.2.0
aws-sdk                           2.958.0
dotenv                            10.0.0
uuid                              8.3.2
@types/node                       16.4.8
@typescript-eslint/eslint-plugin  4.28.5
@typescript-eslint/parser         4.28.5
eslint                            7.32.0
eslint-config-airbnb-typescript   12.3.1
typescript                        4.3.5

Additional context

grep of the pulumi up command for the ConfigMap in question returned three occurrences, though only one in the final output. -- kubernetes:core/v1:ConfigMap live-actimize-nodeAccess delete original +- kubernetes:core/v1:ConfigMap live-actimize-nodeAccess replace [diff: ~data] ++ kubernetes:core/v1:ConfigMap live-actimize-nodeAccess create replacement [diff: ~data]

Contributing

Vote on this issue by adding a 👍 reaction. To contribute a fix for this issue, leave a comment (and link to your pull request, if you've opened one already).

[dixler]

Hi, I'm under the impression that this is a pulumi/pulumi-eks specific question around providers and I'll transfer you over there.

[guineveresaenger]

Hi @ribren - thank you for reporting - we'll take a look as soon as we can.

[danielrbradley]

Hi @ribren sorry it's been a while without response on this issue.

Are you still running into this issue? Would it be possible for you to provide a minimal code reproduction of the issue? E.g. an initial and updated pulumi program which triggers the replacement (even if only intermittantly).

[mikhailshilkov]

Since we haven't heard back, I'll go ahead and close this issue as stale. Also, the linked issue was closed already. Please open a new issue if the problem still persists and provide the details for us to reproduce it. Thank you!