Skip to content

Latest commit

 

History

History
134 lines (102 loc) · 6.82 KB

Backlog.md

File metadata and controls

134 lines (102 loc) · 6.82 KB
Raw

Backlog

Next

  • BUGFIX: a9s create cluster a8s, Im Summary des Befehls wird vergessen, dass minio installiert wurde

  • BUGFIX: a9s create cluster a8s -p kind funktioniert nicht, CLI versucht beim Start immer zu prüfen ob Minikube installiert ist, egal welcher Provider gewählt wurde

  • FEATURE: CLI sollte automatisch erkennen, ob Minikube oder Kind installiert ist. Alternativ: In der Doku den Abschnitt „Setting Kubernetes Provider“ nach oben ziehen, vor den „Creating a Local a8s Postgres Cluster“-Abschnitt

  • Change: When backup-store params are supplied, they should cause existing configuration to be altered. Currently, the existing config is used regardless of parameters supplied to the CLI. This is not intuitive.

    • Plan: demo/config.go#establishBackupStoreConfigYaml * When back-store params are supplied, this file must be regenerated * The file may not be regenerated if no or insufficient arguments are supplied * Or params need to be merged?!

  • BUG: There can be an error message when trying to create a service instance instantly after installing the a8s system.

    • Error message: +Error from server (InternalError): error when creating "/Users/jfischer/a9s/usermanifests/a8s-pg-instance-clustered.yaml": Internal error occurred: failed calling webhook "mpostgresql.kb.io": failed to call webhook: Post "https://postgresql-webhook-service.a8s-system.svc:443/mutate-postgresql-anynines-com-v1beta3-postgresql?timeout=10s": dial tcp 10.104.231.248:443: connect: connection refused
    • Reproduction: Remove sleep(10) from a9s_cli_spec.rb and run e2e-tests

  • Change: Currently, the CLI uses the latest, non-tagged version the a8s-demo repository. However, to ensure the integrity of the a9s CLI it would be better to use the latest release.

  • [ARCHITECTURE] Install a8s PG on an existing cluster

    • Decide which command/verb to use
      • [DONE] a9s create stack
        • Applies the a8s stack to the current k8s cluster
          • The following cluster/context/namespace is selected: Do you want to apply the a8s stack to this cluster?
    • Make the context/namespaces configurable
      • Streamline the UX for both create cluster and create stack
        * Point out that -c can be used to specify a context / clustername
      • Update the readme
        • Add create stack documentation
    • Write a tutorial on how to apply a stack to an existing AWS cluster
      • Use -c option to point to the right context

  • Make the default namespace/context not contain the word demo.

  • [Question] Remove ?

    • Should there be a remove option?
      • Yes
    • Removes the a8s stack from the current k8s cluster

  • [POSTPONED] Run test suite on windows * Perform cold test run * Run test suite

  • CHORE: Harmonize variable declaration for params. Use package config. Maybe use viper.

  • BUGFIX: a9s create demo a8s fails if minikube has never started before.

    minikube profile list -o json
{"error":{"Op":"open","Path":"/home/ubuntu/.minikube/profiles","Err":2}}

  • Feature: In order to script the usage of a9s I want to pass all information to a9s create cluster a8s to avoid any user dialog including the creation of a work direction as well as bucket credentials.

  • Evaluate to use viper to handle config options: https://github.com/spf13/viper

  • Backup: A failed backup should be indicated to the user.

  • Restore: A failed restore should be indicated to the user.

  • Epic: Add AWS-EKS as a provider to create cluster-stack

    • a9s create cluster a8s -p aws
    • Rerequisites
      • aws command and aws configure been run
        • Darwin: brew install awscli
      • eksctl command
        • Darwin: brew install eksctl
    • Steps
      • aws configure
      • eksctl create cluster --name my-eks-cluster --region us-west-2 --node-type t2.medium --nodes 3
      • eksctl utils write-kubeconfig --cluster='my-eks-cluster' --region=us-west-2
      • Activate EBS for dynamic volume provisioning
        • eksctl utils associate-iam-oidc-provider --region=us-west-2 --cluster=my-eks-cluster --approve
        • eksctl create iamserviceaccount
          --name ebs-csi-controller-sa
          --namespace kube-system
          --cluster my-eks-cluster
          --role-name AmazonEKS_EBS_CSI_DriverRole
          --role-only
          --attach-policy-arn arn:aws:iam::aws:policy/service-role/AmazonEBSCSIDriverPolicy
          --approve
          --region=us-west-2
        • Then add the add-on "aws-ebs-csi-driver"
          • eksctl create addon --name aws-ebs-csi-driver --cluster my-eks-cluster --region=us-west-2 --service-account-role-arn arn:aws:iam::${ACCOUNT_ID}:role/AmazonEKS_EBS_CSI_DriverRole --force

  • a9s delete cluster -p aws eksctl delete cluster --name my-eks-cluster --region us-west-2

  • Epic: Use case: Deploy the demo app

    • Feature: a9s create app -k directoryWithKustomize.yaml
    • Demo App
      • The demo consists of an app and a service + kustomize file.
      • kubectl apply -k ...

  • Epic: Verbosity

    • For all commands ensure that the -v flag is respected and without it there's a clean output

  • Feature: Create service instance from backup

    • Combines:
      • Create service instance
      • Restore backup

  • Observability: Backup/Restore: The WaitForKubernetesResource function should indicate the current status cycling through all states with Status = true (scheduled, complete, ...)

  • CHORE: Check if makeup.WaitForUser(demo.UnattendedMode) is used consistently for all new commands instance & backup

  • FEATURE: Backups: Deleting Backups and Restore CRs

    • For completeness: Create command a9s delete pg backup ...
    • For completeness: Create command a9s delete pg restore ...

  • Create binaries in a release matrix, e.g. using Go Release Binaries with Gihub Action Matrix Strategy

  • Create S3 bucket with configs

    • Alternatively: Install a local storage provider, e.g. minio.
      • Costly dependency: add the local storage provider to the backup agent.

  • CHORE: Switch to https://docs.github.com/de/get-started/using-github/github-flow over git flow