Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Resolve #101 - Support K8s 1.22 - 1.27 #102

Open
wants to merge 73 commits into
base: main
Choose a base branch
from

Conversation

rlratcliffe
Copy link
Contributor

@rlratcliffe rlratcliffe commented Sep 3, 2023

This PR is to allow support for Tavros up to Kubernetes 1.27. Tavros was last built with K8s 1.21 as far as I can tell. K8s 1.27 is now available, and my client needed to update to at least 1.25. Not all Kubernetes (K8s) versions have removals of APIs, but there were removals of APIs in K8s 1.22 and 1.25 that affected a number of Tavros components. K8s maintains a guide of these changes here.

Used Fairwinds' Pluto to generate list of what needed to be updated.

These changes include general maintenance but are primarily targeting these K8s API removals.

  • Updated Keycloak operator 12.0.4 -> 19.0.2. Updated various KeycloakClients in order to support the changes. Roles and scopes appear to not be created the same and can be created from the CRD instead, so these changes fix those issues.
  • Updated Kong chart from 2.3.0 -> 2.15.0 and Kong gateway from 2.5.0.0-alpine -> 3.0.0.0-alpine
  • Updated ECK from 1.6.0 -> 2.5.0
  • Updated kube-state-metrics from v2.0.0-rc.1 -> v2.7.0
  • Updated cert manager from 1.3.0 -> 1.11.0
  • Switched postgres helm repo due to the index changing.
  • Updated Jenkins from 2.277.4-lts-alpine -> 2.416-jdk11 and Jenkins operator from v0.6.0 0 -> v0.8.0-beta2. This PR will also need to be merged for pipeline related changes and a release will need to be made with v0.5. If no changes on the Jenkins PR, for testing this branch the library can be just changed in the configmap-jenkins-config to use the pre-release library version: v0.5-alpha.4
  • Updated Jaeger from v1.25.0 -> v1.47.0 and fixed the initialization of certs
  • Updated sealed secrets from v0.15.0 -> v0.17.5.
  • Updated various YAMLs to replace API endpoints

Some additional changes/additions, that either facilitated testing of these changes or are features we use at my client. Can move them to a different PR if that would be better.

  • Additional AKS ansible commands related to resource groups, storage, and DNS zones related to AKS provisioning
  • Added acceptance tests and other docs. Tests should eventually be automated but are a start.
  • Added tavros-repo helm repo to dev/test/prod to facilitate ease of deploying camel apps with less confusion
  • Git shell in Gitea namespace just in case it's necessary to push to git while kong is down
  • Troubleshooting shells inside of dev/test/prod namespaces. Image contains curl to be able to curl to the APIs with ease.
  • Fixed issue with occasionally failing to install Kibana assets

Thus far, this has only been tested on AKS.


  • Requires Jenkins Tavros library merge/release - Yes
  • Requires Tavros chart release - No
  • Requires archetype release - No

fix deprecated version + issue with missing quay image
old chart no longer existed, and values didn't match and needed to be reconfigured. old values are here: https://github.com/bitnami/charts/blob/4c72712f8d0c79ca3b573c87b993bfe80fc99b01/bitnami/postgresql/values.yaml
maybe not a good idea if it deletes IP stuff but trying it out
https://cert-manager.io/docs/installation/supported-releases/

prod-kong-ingress-certs & sandbox-kong-ingress-certs weren't being created, but are now
NOTE: this keycloak-operator is deprecated and no more fixes are available
was suddenly failing after many repeat runs
need to fix:
- erroring out deployment in ansible script
- jaeger user not being created properly in elastic (tavros-es-creds not working to access elastic node out of the box)
- tavros-es-http-certs (-n jaeger) secret needs ca.crt from tavros-es-http-certs-public (-n elastic-system)
- fix init script that creates jaeger role and user so that it doesn't silently fail, and therefore will keep retrying
- move init-runner wait before wait for elastic in ansible script
the storage account gets deleted when the resource group does
the default for when there's no cert is really just for the test, which I'm not sure is ideal
@rlratcliffe rlratcliffe marked this pull request as ready for review September 4, 2023 14:20
@rlratcliffe rlratcliffe mentioned this pull request Sep 4, 2023
14 tasks
@rlratcliffe rlratcliffe changed the title Resolve #101 Resolve #101 - Support K8s 1.22 - 1.27 Sep 4, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant