job

Documentation

Index

• Constants
• func CreateJob(ctx context.Context, c clientset.Interface, ns string, job *batchv1.Job) (*batchv1.Job, error)
• func GetAllRunningJobPods(ctx context.Context, c clientset.Interface, ns, jobName string) ([]v1.Pod, error)
• func GetJob(ctx context.Context, c clientset.Interface, ns, name string) (*batchv1.Job, error)
• func GetJobPods(ctx context.Context, c clientset.Interface, ns, jobName string) (*v1.PodList, error)
• func NewTestJob(behavior, name string, rPol v1.RestartPolicy, parallelism, completions int32, ...) *batchv1.Job
• func NewTestJobOnNode(behavior, name string, rPol v1.RestartPolicy, parallelism, completions int32, ...) *batchv1.Job
• func UpdateJob(ctx context.Context, c clientset.Interface, ns string, job *batchv1.Job) (*batchv1.Job, error)
• func WaitForAllJobPodsGone(ctx context.Context, c clientset.Interface, ns, jobName string) error
• func WaitForJobComplete(ctx context.Context, c clientset.Interface, ns, jobName string, reason *string, ...) error
• func WaitForJobCondition(ctx context.Context, c clientset.Interface, ns, jobName string, ...) error
• func WaitForJobFailed(c clientset.Interface, ns, jobName string) error
• func WaitForJobFinish(ctx context.Context, c clientset.Interface, ns, jobName string) error
• func WaitForJobGone(ctx context.Context, c clientset.Interface, ns, jobName string, ...) error
• func WaitForJobPodsRunning(ctx context.Context, c clientset.Interface, ns, jobName string, ...) error
• func WaitForJobPodsSucceeded(ctx context.Context, c clientset.Interface, ns, jobName string, ...) error
• func WaitForJobReady(ctx context.Context, c clientset.Interface, ns, jobName string, ready *int32) error
• func WaitForJobState(ctx context.Context, c clientset.Interface, ns, jobName string, ...) error
• func WaitForJobSuspend(ctx context.Context, c clientset.Interface, ns, jobName string) error
• type JobState

Constants

const (
	// JobTimeout is how long to wait for a job to finish.
	JobTimeout = 15 * time.Minute

	// JobSelectorKey is a job selector name
	JobSelectorKey = "job"
)

Functions

func CreateJob

func CreateJob(ctx context.Context, c clientset.Interface, ns string, job *batchv1.Job) (*batchv1.Job, error)

CreateJob uses c to create job in namespace ns. If the returned error is nil, the returned Job is valid and has been created.

func GetAllRunningJobPods

func GetAllRunningJobPods(ctx context.Context, c clientset.Interface, ns, jobName string) ([]v1.Pod, error)

GetAllRunningJobPods returns a list of all running Pods belonging to a Job.

func GetJob

func GetJob(ctx context.Context, c clientset.Interface, ns, name string) (*batchv1.Job, error)

GetJob uses c to get the Job in namespace ns named name. If the returned error is nil, the returned Job is valid.

func GetJobPods

func GetJobPods(ctx context.Context, c clientset.Interface, ns, jobName string) (*v1.PodList, error)

GetJobPods returns a list of Pods belonging to a Job.

func NewTestJob

func NewTestJob(behavior, name string, rPol v1.RestartPolicy, parallelism, completions int32, activeDeadlineSeconds *int64, backoffLimit int32) *batchv1.Job

NewTestJob returns a Job which does one of several testing behaviors. notTerminate starts a Job that will run effectively forever. fail starts a Job that will fail immediately. succeed starts a Job that will succeed immediately. randomlySucceedOrFail starts a Job that will succeed or fail randomly. failOnce fails the Job the first time it is run and succeeds subsequently. name is the Name of the Job. RestartPolicy indicates the restart policy of the containers in which the Pod is running. Parallelism is the Job's parallelism, and completions is the Job's required number of completions.

func NewTestJobOnNode

func NewTestJobOnNode(behavior, name string, rPol v1.RestartPolicy, parallelism, completions int32, activeDeadlineSeconds *int64, backoffLimit int32, nodeName string) *batchv1.Job

NewTestJobOnNode is similar to NewTestJob but supports specifying a Node on which the Job's Pods will run. Empty nodeName means no node selection constraints.

func UpdateJob

func UpdateJob(ctx context.Context, c clientset.Interface, ns string, job *batchv1.Job) (*batchv1.Job, error)

UpdateJob uses c to update a job in namespace ns. If the returned error is nil, the returned Job is valid and has been updated.

func WaitForAllJobPodsGone

func WaitForAllJobPodsGone(ctx context.Context, c clientset.Interface, ns, jobName string) error

WaitForAllJobPodsGone waits for all pods for the Job named jobName in namespace ns to be deleted.

func WaitForJobComplete

func WaitForJobComplete(ctx context.Context, c clientset.Interface, ns, jobName string, reason *string, completions int32) error

WaitForJobComplete uses c to wait for completions to complete for the Job jobName in namespace ns. This function checks if the number of succeeded Job Pods reached expected completions and the Job has a "Complete" condition with the expected reason. The pointer "reason" argument allows us to skip "Complete" condition reason verifications. The conformance test cases have the different expected "Complete" condition reason ("CompletionsReached" vs "") between conformance CI jobs and e2e CI jobs since the e2e conformance test cases are performed in both conformance CI jobs with GA-only features and e2e CI jobs with all default-enabled features. So, we need to skip "Complete" condition reason verifications in the e2e conformance test cases.

func WaitForJobCondition

func WaitForJobCondition(ctx context.Context, c clientset.Interface, ns, jobName string, cType batchv1.JobConditionType, reason *string) error

WaitForJobCondition waits for the specified Job to have the expected condition with the specific reason. When the nil reason is passed, the "reason" string in the condition is not checked.

func WaitForJobFailed

func WaitForJobFailed(c clientset.Interface, ns, jobName string) error

WaitForJobFailed uses c to wait for the Job jobName in namespace ns to fail

func WaitForJobFinish

func WaitForJobFinish(ctx context.Context, c clientset.Interface, ns, jobName string) error

WaitForJobFinish uses c to wait for the Job jobName in namespace ns to finish (either Failed or Complete).

func WaitForJobGone

func WaitForJobGone(ctx context.Context, c clientset.Interface, ns, jobName string, timeout time.Duration) error

WaitForJobGone uses c to wait for up to timeout for the Job named jobName in namespace ns to be removed.

func WaitForJobPodsRunning

func WaitForJobPodsRunning(ctx context.Context, c clientset.Interface, ns, jobName string, expectedCount int32) error

WaitForJobPodsRunning wait for all pods for the Job named JobName in namespace ns to become Running. Only use when pods will run for a long time, or it will be racy.

func WaitForJobPodsSucceeded

func WaitForJobPodsSucceeded(ctx context.Context, c clientset.Interface, ns, jobName string, expectedCount int32) error

WaitForJobPodsSucceeded wait for all pods for the Job named JobName in namespace ns to become Succeeded.

func WaitForJobReady

func WaitForJobReady(ctx context.Context, c clientset.Interface, ns, jobName string, ready *int32) error

WaitForJobReady waits for particular value of the Job .status.ready field

func WaitForJobState

func WaitForJobState(ctx context.Context, c clientset.Interface, ns, jobName string, timeout time.Duration, state JobState) error

WaitForJobState waits for a job to be matched to the given condition. The condition callback may use gomega.StopTrying to abort early.

func WaitForJobSuspend

func WaitForJobSuspend(ctx context.Context, c clientset.Interface, ns, jobName string) error

WaitForJobSuspend uses c to wait for suspend condition for the Job jobName in namespace ns.

Types

type JobState

type JobState func(job *batchv1.Job) string

JobState is used to verify if Job matches a particular condition. If it matches, an empty string is returned. Otherwise, the string explains why the condition is not matched. This should be a short string. A dump of the job object will get added by the caller.