Multiple Cluster pipeline for Conformitron #160
Conversation
|
I run this with |
There was a problem hiding this comment.
@Howlla As discussed, I recommend the following:
- A Blueprints Construct holding the cluster setup with addons
- Using K8s version map to render versions with Construct with stages for pipeline for multiple clusters.
- Remove Argo out of the equation. Use FluxConfig
There was a problem hiding this comment.
@Howlla Overall looks great, I have some minor feedback.
- Documentation is missing
- Reference doc to mkdocs for the pattern
- Commented code cleanup, remove account numbers and personal owner reference
- Think about accommodating BottleRocket as one of the cluster type
- I recommend creating a specific Construct or reusing Observability Builder class in this as Construct which can be rendered multi times to create many clusters
- Multi Account Observability should be part of this. You can refer to our Multi Account Mon pattern in CDK Accelerator.
- Centralized Logging with ADOT to CW
- Centralized tracing with ADOT to X-RAY
- Centralized Monitoring with ADOT to AMP
- Grafana Operator and Grafana Dashboarding for multi cluster should be a part.
- We should bring in Hardening to Clusters using tools like Trivy, Kyverno (Mike should help you with this).
| blueprints.CodePipelineStack.builder() | ||
| .application('npx ts-node bin/multi-cluster-conformitron.ts') | ||
| .name('multi-cluster-central-pipeline') | ||
| .owner('Howlla') |
There was a problem hiding this comment.
This will have to change when you merge!
My branch
|
Hi @elamaran11 Current status 1.Documentation is missing - Added |
merging changes
elamaran11
changed the title
|
New PR Created Here #176 |
Issue #, if available:
Description of changes: Testing pipeline for Kube 1.24 to 1.27 with x86 and ARM architectures
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.