integration

README

Integration tests

The basics

To run all tests from the minikube root directory:

make integration

Quickly iterating on a single test

Run a single test on an active cluster:

make integration -e TEST_ARGS="-test.run TestFunctional/parallel/MountCmd --profile=minikube --cleanup=false"

WARNING: For this to work repeatedly, the test must be written so that it cleans up after itself.

See main.go for details.

Disabling parallelism

make integration -e TEST_ARGS="-test.parallel=1"

Testing philosophy

• Tests should be so simple as to be correct by inspection
• Readers should need to read only the test body to understand the test
• Top-to-bottom readability is more important than code de-duplication

Tests are typically read with a great air of skepticism, because chances are they are being read only when things are broken.

Documentation

Index

• func CanCleanup() bool
• func Cleanup(t *testing.T, profile string, cancel context.CancelFunc)
• func CleanupWithLogs(t *testing.T, profile string, cancel context.CancelFunc)
• func ExpectedDefaultDriver() string
• func HyperVDriver() bool
• func MaybeParallel(t *testing.T)
• func NoneDriver() bool
• func PodWait(ctx context.Context, t *testing.T, profile string, ns string, selector string, ...) ([]string, error)
• func ReadLineWithTimeout(b *bufio.Reader, timeout time.Duration) (string, error)
• func StartArgs() []string
• func Status(ctx context.Context, t *testing.T, path string, profile string) string
• func Target() string
• func TestMain(m *testing.M)
• func UniqueProfileName(prefix string) string
• func WaitForStartSlot(t *testing.T)
• type RunResult
  • func Run(t *testing.T, cmd *exec.Cmd) (*RunResult, error)
  • func (rr RunResult) Command() string
  • func (rr RunResult) Output() string
• type StartSession
  • func Start(t *testing.T, cmd *exec.Cmd) (*StartSession, error)
  • func (ss *StartSession) Stop(t *testing.T)

Functions

func CanCleanup

func CanCleanup() bool

CanCleanup returns if cleanup is allowed

func Cleanup

func Cleanup(t *testing.T, profile string, cancel context.CancelFunc)

Cleanup cleans up after a test run

func CleanupWithLogs

func CleanupWithLogs(t *testing.T, profile string, cancel context.CancelFunc)

CleanupWithLogs cleans up after a test run, fetching logs and deleting the profile

func ExpectedDefaultDriver

func ExpectedDefaultDriver() string

ExpectedDefaultDriver returns the expected default driver, if any

func HyperVDriver

func HyperVDriver() bool

HyperVDriver returns whether or not this test is using the Hyper-V driver

func MaybeParallel

func MaybeParallel(t *testing.T)

MaybeParallel sets that the test should run in parallel

func NoneDriver

func NoneDriver() bool

NoneDriver returns whether or not this test is using the none driver

func PodWait

func PodWait(ctx context.Context, t *testing.T, profile string, ns string, selector string, timeout time.Duration) ([]string, error)

PodWait waits for pods to achieve a running state.

func ReadLineWithTimeout

func ReadLineWithTimeout(b *bufio.Reader, timeout time.Duration) (string, error)

ReadLineWithTimeout reads a line of text from a buffer with a timeout

func StartArgs

func StartArgs() []string

StartArgs returns the arguments normally used for starting minikube

func Status

func Status(ctx context.Context, t *testing.T, path string, profile string) string

Status returns the minikube cluster status as a string

func Target

func Target() string

Target returns where the minikube binary can be found

func TestMain

func TestMain(m *testing.M)

TestMain is the test main

func UniqueProfileName

func UniqueProfileName(prefix string) string

UniqueProfileName returns a reasonably unique profile name

func WaitForStartSlot

func WaitForStartSlot(t *testing.T)

WaitForStartSlot enforces --start-offset to avoid startup race conditions

Types

type RunResult

type RunResult struct {
	Stdout   *bytes.Buffer
	Stderr   *bytes.Buffer
	ExitCode int
	Args     []string
}

RunResult stores the result of an cmd.Run call

func Run

func Run(t *testing.T, cmd *exec.Cmd) (*RunResult, error)

Run is a test helper to log a command being executed \_(ツ)_/¯

func (RunResult) Command

func (rr RunResult) Command() string

Command returns a human readable command string that does not induce eye fatigue

func (RunResult) Output

func (rr RunResult) Output() string

Output returns human-readable output for an execution result

type StartSession

type StartSession struct {
	Stdout *bufio.Reader
	Stderr *bufio.Reader
	// contains filtered or unexported fields
}

StartSession stores the result of an cmd.Start call

func Start

func Start(t *testing.T, cmd *exec.Cmd) (*StartSession, error)

Start starts a process in the background, streaming output

func (*StartSession) Stop

func (ss *StartSession) Stop(t *testing.T)

Stop stops the started process