/kubeam

Kubernetes Application and Workflow manager

Primary LanguageGoApache License 2.0Apache-2.0

kubeam-logo

Kubernetes Application and Workflow manager

Integrates directly with your CI (Jenkins BO/CircleCI/Bamboo) and allows you to have end to end CI/CD.

jenkins-bo-and-kubeam

TODO:

  • We need to Add available cluster detection to kubeAM. Redis (not fully featured yet)
  • Add samples of Jenkinsfile and circle.yaml
  • Add TLS mutual cert auth. If possible add suport for multiple certs validation (one per account)
  • Add Logic to parse application.json and return in structure. This will drive the logic of each rest endpoint.
  • convert from using kubectl to cliento-go For better error handling, retry with backoff and return information to caller rest endpoint. We have a PR for this need input on pros and cons on doing client-go way.
  • Better build system. Look at isakconf for a sample.
  • Reporting of cluster aviability.
  • getStatus should loop for each resource and generate a report. Right now is issuing a single call. If one resource is missing. Causes report to fail.
  • Health-check of kubeam
  • Add support for includes to templates. This is needed for suporting source CIDR restrictions.
  • Switch to go template from fast-template
  • Rename provision api to something with more meaning is supposed to be endpoint for resources that timeout.
  • Make /v1/provision work with generic applications using api.yaml file
  • make /v1/waitforready generic driven by api.yaml
  • Remove self-deploy endpoint. Now it can self deploy using api.yaml just like any other app.
  • /v1/waitforready should look for multiple resources and wait for all of them. This is for complex applications

API

Create a brand new cluster
curl -X POST -k https://admin:{password}@localhost:8443/v1/create/sampleapp/QA/67/rhel7-1718-02/300

Recreates (Delete then create) a resource using cluster selector; A cluster will be selected for you (Resources have TTL)
curl -X POST -k https://admin:{password}@localhost:8443/v1/provision/sampleapp/QA/rhel7-170918-02

Deploys to expecified
curl -X POST -k https://admin:{password}@localhost:8443/v1/deploy/sampleapp/qa/rhel7-170918-02

Gets status of a cluster
curl -X GET -k https://admin:{password}@localhost:8443/v1/provision/sampleapp/QA/67

List active clusters from Redis
curl -X GET -k https://admin:{password}@localhost:8443/v1/listclusters/sampleapp/QA

Wait for resource to be active
wget --no-check-certificate --user=admin --password=$SECRET https:/localhost:443/v1/waitforready/sampleapp/QA/66 -q -O - PS. wget due to limitations in how curl does timeout

Get detail on running cluster
curl -X GET -k https://admin:${SECRET}@localhost:443/v1/getclusterdetail/sampleapp/QA/66

Get detail on all running clusters
curl -X GET -k https://admin:${SECRET}@localhost:443/v1/getallclustersdetail/sampleapp/QA

Event Tracker

PAYLOAD="
{
   name: "Deployed",
   msg: "Passed all tests",
   ts: "2018-01-01 01:01:01"
}
curl -X POST -k https://admin:${SECRET}@localhost:443/v1/event/sampleapp/QA/66/{docker-TAG}

TODO: TAGS

PAYLOAD="
{
   name: "BATS-OK",
   msg: "PASSED sampleapp-BATS",
   ts: "2018-01-01 01:01:01"
}
curl -X POST -k https://admin:${SECRET}@localhost:443/v1/settag/sampleapp/QA/66/{docker-TAG}

KEYS are:  <APPLICATION>/<ENVIRONMENT>/<SHARD>/<DOCKER-TAG>

Other:

Self deployment
/v1/provision/self/{environment}/{tag}

Health-check TCP:8081/health-check