johannes-b / keptn-user-stories

Keptn User Stories

This repository defines user stories for Keptn. For the user stories a shipyard/uniform describe the Keptn installation. Based on that, the following personas want executed their tasks:

:person_with_blond_hair: Tim: Developer

:person_with_blond_hair: Tom: Developer

👨 Arnold: DevOps engineer (DevSecOps)

👱‍♀️ Sue: Acceptance engineer and member of performance engineering team

👩 Alice: TBD

Categorization of User Stories

User Stories

Standard Keptn 0.6.0 Scenario

• Keptn 0.6.0 Flow: This use story shows the standard flow as implemented in Keptn 0.6.0.

Continuous Delivery

• Canary Release on K8s: As a user, I want to do a canary release on Kubernetes.

• Release Brackets: As a user, I want to queue services in staging and finally release them together (even in a canary-way).

• As a user, I want to activate an approval step for any Keptn task (switch from an automatic to a manual mode)

  • As a user, I want to override (force/skip) an approval step

• As a user, I want to involve an existing CD pipeline, e.g., to execute tests.

  • I want to involve external services into the process (e.g., trigger a testing server running outside of the Keptn-cluster)​
  • I want to register a webhook for a Keptn project, stage, service, and action: keptn register webhook http://abc.com/deploy --project --stage --service --action=test
  • I want to tell Keptn that the execution of a task has finished.

Automated Operations

• As a user, I want to define a custom remediation workflow with multiple actions.

• As a user, I want to get a notification for a problem including data from multiple sources.

• Disable Self-healing: As a user, I want to disable self-healing for a certain time period; when updates are conducted which have an impact on a service's SLOs.

• As a member of DevSecOps, I’d like the offending process restarted or new instance spawned after killing the bad process; when CPU utilization reached a certain threshold and process crashed affecting performance or failure rates increased

• As a member of DevSecOps, I’d like the offending process restarted or new instance spawned after killing the bad process; when memory utilization reached a certain threshold and process crashed affecting performance or failure rates increased

• As a member of DevSecOps, I’d like disk space increased or a ticket filed to get them increased; when disk space is full and server crashed

Platform - Execution Plane:

• As a user, I want to deploy my artifacts on different target systems (multiple clusters, AWS Lambda, …) using blue/green, primary/canary, etc.

• As a user, I want to modify/control the used service mesh.

Platform - Control Plane:

• As a user, I want to run Keptn (core) standalone (e.g., using Docker compose)​

Shipyard (Definition of WHAT, Process)

• As a user, I want to define custom processes in the shipyard (i.e., approval before deploying to production, tests of dark-deployments, etc.)​

Uniform (Definition of HOW, Tooling):

• Web/Docker Hooks:

  • As a user, I want to register a webhook for a Keptn project, stage, service and action

  • As a user, I want to provide a Docker image which is used for tests/remediations​

• Custom Keptn Services: As a user, I want to extend my Keptn installation with additional (custom) services

• As a user, I want to extend my workflow/process:

  • keptn update shipyard "sli_provider+=prometheus,neoload" // add prometheus and neoload as sli prov.​
  • keptn update uniform --stage=hardening "tests+=selenium" // add selenium tests for stage hardening​

GitOps - GitFlow

• (GitOps) As a user, I want to update the configuration in my Git repo which is afterwards applied.​

  • I want to be able to edit my deployments in the Git repo (adding a new environment variable, changing cpu/memory limits), and Keptn should apply it​

• (GitOps) As a user, I want to onboard already deployed services/deployments [on another cluster]​

  • I want to be able to re-create a deployment that Keptn stores in a Git Repo on another cluster (potentially even without needing to install Keptn, just with kubectl apply -f https://github.com/some-user/my-project.git)​

• (GitFlow) Sync Stages As a user, I want to sync a stage with the last successfully evaluated configuration of its preceding stage.

DevOps - User Experience (Bridge)

• As a member of DevSecOps or performance engineering team, I’d like application performance graphs automatically generated as artifacts.

• As a member of DevSecOps or performance engineering team, I’d like the ability compare application performance build by build, stage by stage to identify areas of opportunity through deep dive analysis using the artifacts.