e2e

README

Build Deploy and E2E test Automation Framework

This is the folder in which we can invoke e2e tests for tce addons packages.

Prerequisites

1. Dependent packages or tools (like ytt) would be handled in respective addons package suites.

Supported flags

1. --kubeconfig : Use to set the kube config path
2. --kube-context : Use to set cluster context
3. --packages : Use to filter out the addon package tests to run
4. --version: Provide respective version for packages mentioned.
5. --provider : The environment in which test should run (eg: docker, aws, vsphere etc)
6. --cluster-type : Set type of cluster (eg: management)
7. --create-cluster : Provide true if cluster has to provision. Provide false if the cluster is already exist. If true automation creates cluster based on the provider and cluster type.
8. --tce-version : Provide tce release version to install(eg: "v0.7.0"). If not provided then build it from source code.
9. --guest-cluster-name: Provide cluster name (for workload cluster incase of managed)
10. --management-cluster-name: Provide cluster name for management cluster.
11. --cluster-plan: Provide Cluster Plan (eg: dev, prod etc). By default it will be set to "dev".
12. --cleanup-cluster: Provide true for tearing down the cluster.

How to run framework to install TCE release from github page, provision cluster and test

ginkgo -v -- --kubeconfig=$KUBECONFIG --packages="external-dns" --version="0.8.0" --provider="docker" --cluster-type="management" --guest-cluster-name="tce-mycluster" --create-cluster --tce-version="v0.7.0"

How to create management cluster on docker

ginkgo -v -- --kubeconfig=$KUBECONFIG --packages="external-dns" --version="0.8.0" --provider="docker" --cluster-type="managed" --tce-version="v0.7.0" --management-cluster-name="tce-management-cluster" --create-cluster --guest-cluster-name="tce-workload-cluster"

How to run framework to install TCE by building release from source code, provision cluster and test

ginkgo -v -- --kubeconfig=$KUBECONFIG --packages="all" --provider="docker" --cluster-type="management" --guest-cluster-name="tce-mycluster" --create-cluster

How to run addons package test if cluster is already available

1. to run individual packages test

   ginkgo -v -- --packages="calico,contour" --version="3.11.3,1.17.1" --guest-cluster-name="tce-mycluster" or ginkgo -v -- --packages="contour" --version="1.17.1" --guest-cluster-name="tce-mycluster"

2. to run all addons package tests

   ginkgo -v -- --kubeconfig=$KUBECONFIG --packages="all" --guest-cluster-name="tce-mycluster"

3. how to pass cluster context

   ginkgo -v -- --kubeconfig=$KUBECONFIG --packages="antrea" --version="0.11.3" --kube-context="cluster-dev-test-admin@cluster-dev-test"

Notes

1. If all packages tests need to run then --version is not required along with --packages="all"
2. By default --packages="all" will run the package testing for the latest version of respective packages. One can modify the required package version in addons_config.yaml if they wanted to test lower version of supported packages.
3. Default cluster plan is dev.
4. Provide release version of TCE in --tce-version to fetch it from github and install. If value is not provided then automation code checks if the tanzu cli already available in the environment, if not it will create a release from source code and installs tce.
5. In case of any failures, created cluster will be tear down.
6. Cluster tearing down is optional. One can delete cluster passing --cleanup-cluster.

Documentation

Index

• Constants
• Variables
• func CheckClusterHealth(contextName string) error
• func DeleteCluster() error
• func DeployTanzuCluster() error
• func GetKubeConfig(clusterContext string) (*rest.Config, error)
• func GetResourceObj(clusterContext string) error
• func Initialize()
• func RunAddonsTests() error
• func RunCommand(commandName, args string) error
• type AddonsConfig
• type Config
  • func New() *Config
• type PackageConfiguration

Constants

const (
	MANAGED = "managed"
	DOCKER  = "docker"
	TIMEOUT = 120
)

Variables

var (
	ConfigVal        *Config
	MetallbInstalled bool
	VeleroInstalled  bool
)

var (
	Clientset *kubernetes.Clientset
	Res       *resources.Resources
)

Functions

func CheckClusterHealth

func CheckClusterHealth(contextName string) error

func DeleteCluster

func DeleteCluster() error

func DeployTanzuCluster

func DeployTanzuCluster() error

func GetKubeConfig

func GetKubeConfig(clusterContext string) (*rest.Config, error)

GetKubeConfig will set cluster context passed using flag -kube-context and gives you the *rest.Config to access the cluster objects

func GetResourceObj

func GetResourceObj(clusterContext string) error

GetResourceObj sets resource object to communicate with apiserver

func Initialize

func Initialize()

func RunAddonsTests

func RunAddonsTests() error

func RunCommand

func RunCommand(commandName, args string) error

Types

type AddonsConfig

type AddonsConfig struct {
	Name    string `yaml:"name"`
	Version string `yaml:"version"`
}

type Config

type Config struct {
	Kubeconfig             string
	Kubecontext            string
	Provider               string
	ClusterType            string
	Packages               string
	Version                string
	GuestClusterName       string
	ClusterPlan            string
	MgmtClusterName        string
	ClusterInstallRequired bool
	ClusterCleanupRequired bool
	TceVersion             string
}

func New

func New() *Config

type PackageConfiguration

type PackageConfiguration struct {
	Packages []AddonsConfig `yaml:"packages"`
}