Skip to content

Latest commit

 

History

History
798 lines (659 loc) · 27.2 KB

taskruns.md

File metadata and controls

798 lines (659 loc) · 27.2 KB
Raw

TaskRuns

Overview

A TaskRun allows you to instantiate and execute a Task on-cluster. A Task specifies one or more Steps that execute container images and each container image performs a specific piece of build work. A TaskRun executes the Steps in the Task in the order they are specified until all Steps have executed successfully or a failure occurs.

Configuring a TaskRun

A TaskRun definition supports the following fields:

  • Required:
    • apiVersion - Specifies the API version, for example tekton.dev/v1beta1.
    • kind - Identifies this resource object as a TaskRun object.
    • metadata - Specifies the metadata that uniquely identifies the TaskRun, such as a name.
    • spec - Specifies the configuration for the TaskRun.
  • Optional:
    • serviceAccountName - Specifies a ServiceAccount object that provides custom credentials for executing the TaskRun.
    • params - Specifies the desired execution parameters for the Task.
    • resources - Specifies the desired PipelineResource values.
      • inputs - Specifies the input resources.
      • outputs - Specifies the output resources.
    • timeout - Specifies the timeout before the TaskRun fails.
    • podTemplate - Specifies a Pod template to use as the starting point for configuring the Pods for the Task.
    • workspaces - Specifies the physical volumes to use for the Workspaces declared by a Task.

Specifying the target Task

To specify the Task you want to execute in your TaskRun, use the taskRef field as shown below:

spec:
  taskRef:
    name: read-task

You can also embed the desired Task definition directly in the TaskRun using the taskSpec field:

spec:
  taskSpec:
    resources:
      inputs:
        - name: workspace
          type: git
    steps:
      - name: build-and-push
        image: gcr.io/kaniko-project/executor:v0.17.1
        # specifying DOCKER_CONFIG is required to allow kaniko to detect docker credential
        env:
          - name: "DOCKER_CONFIG"
            value: "/tekton/home/.docker/"
        command:
          - /kaniko/executor
        args:
          - --destination=gcr.io/my-project/gohelloworld

Tekton Bundles

Note: This is only allowed if enable-tekton-oci-bundles is set to "true" or enable-api-fields is set to "alpha" in the feature-flags configmap, see install.md

You may also reference Tasks that are defined outside of your cluster using Tekton Bundles. A Tekton Bundle is an OCI artifact that contains Tekton resources like Tasks which can be referenced within a taskRef.

spec:
taskRef:
  name: echo-task
  bundle: docker.io/myrepo/mycatalog

Here, the bundle field is the full reference url to the artifact. The name is the metadata.name field of the Task.

You may also specify a tag as you would with a Docker image which will give you a repeatable reference to a Task.

spec:
taskRef:
  name: echo-task
  bundle: docker.io/myrepo/mycatalog:v1.0.1

You may also specify a fixed digest instead of a tag which ensures the referenced task is constant.

spec:
taskRef:
  name: echo-task
  bundle: docker.io/myrepo/mycatalog@sha256:abc123

A working example can be found here.

Any of the above options will fetch the image using the ImagePullSecrets attached to the ServiceAccount specified in the TaskRun. See the Service Account section for details on how to configure a ServiceAccount on a TaskRun. The TaskRun will then run that Task without registering it in the cluster allowing multiple versions of the same named Task to be run at once.

Tekton Bundles may be constructed with any toolsets that produces valid OCI image artifacts so long as the artifact adheres to the contract. Additionally, you may also use the tkn cli (coming soon).

Specifying Parameters

If a Task has parameters, you can use the params field to specify their values:

spec:
  params:
    - name: flags
      value: -someflag

Note: If a parameter does not have an implicit default value, you must explicitly set its value.

Specifying Resources

If a Task requires Resources (that is, inputs and outputs) you must specify them in your TaskRun definition. You can specify Resources by reference to existing PipelineResource objects or embed their definitions directly in the TaskRun.

Note: A TaskRun can use either a referenced or an embedded Resource but not both simultaneously.

Below is an example of specifying Resources by reference:

spec:
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: java-git-resource
    outputs:
      - name: image
        resourceRef:
          name: my-app-image

And here is an example of specifying Resources by embedding their definitions:

spec:
  resources:
    inputs:
      - name: workspace
        resourceSpec:
          type: git
          params:
            - name: url
              value: https://github.com/pivotal-nader-ziada/gohelloworld

Note: You can use the paths field to override the paths to a Resource.

Specifying Resource limits

Each Step in a Task can specify its resource requirements. See Defining Steps

Specifying a Pod template

You can specify a Pod template configuration that will serve as the configuration starting point for the Pod in which the container images specified in your Task will execute. This allows you to customize the Pod configuration specifically for that TaskRun.

In the following example, the Task specifies a volumeMount (my-cache) object, also provided by the TaskRun, using a PersistentVolumeClaim volume. A specific scheduler is also configured in the SchedulerName field. The Pod executes with regular (non-root) user permissions.

apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
  name: mytask
  namespace: default
spec:
  steps:
    - name: writesomething
      image: ubuntu
      command: ["bash", "-c"]
      args: ["echo 'foo' > /my-cache/bar"]
      volumeMounts:
        - name: my-cache
          mountPath: /my-cache
---
apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  name: mytaskrun
  namespace: default
spec:
  taskRef:
    name: mytask
  podTemplate:
    schedulerName: volcano
    securityContext:
      runAsNonRoot: true
      runAsUser: 1001
    volumes:
      - name: my-cache
        persistentVolumeClaim:
          claimName: my-volume-claim

Specifying Workspaces

If a Task specifies one or more Workspaces, you must map those Workspaces to the corresponding physical volumes in your TaskRun definition. For example, you can map a PersistentVolumeClaim volume to a Workspace as follows:

workspaces:
  - name: myworkspace # must match workspace name in the Task
    persistentVolumeClaim:
      claimName: mypvc # this PVC must already exist
    subPath: my-subdir

For more information, see the following topics:

Specifying Sidecars

A Sidecar is a container that runs alongside the containers specified in the Steps of a task to provide auxiliary support to the execution of those Steps. For example, a Sidecar can run a logging daemon, a service that updates files on a shared volume, or a network proxy.

Tekton supports the injection of Sidecars into a Pod belonging to a TaskRun with the condition that each Sidecar running inside the Pod are terminated as soon as all Steps in the Task complete execution. This might result in the Pod including each affected Sidecar with a retry count of 1 and a different container image than expected.

We are aware of the following issues affecting Tekton's implementation of Sidecars:

  • The configured nop image must not provide the command that the Sidecar is expected to run, otherwise it will not exit, resulting in the Sidecar running forever and the Task eventually timing out. For more information, see the associated issue.

  • The kubectl get pods command returns the status of the Pod as "Completed" if a Sidecar exits successfully and as "Error" if a Sidecar exits with an error, disregarding the exit codes of the container images that actually executed the Steps inside the Pod. Only the above command is affected. The Pod's description correctly denotes a "Failed" status and the container statuses correctly denote their exit codes and reasons.

Specifying LimitRange values

In order to only consume the bare minimum amount of resources needed to execute one Step at a time from the invoked Task, Tekton only requests the maximum values for CPU, memory, and ephemeral storage from within each Step. This is sufficient as Steps only execute one at a time in the Pod. Requests other than the maximum values are set to zero.

When a LimitRange parameter is present in the namespace in which TaskRuns are executing and minimum values are specified for container resource requests, Tekton searches through all LimitRange values present in the namespace and uses the minimums instead of 0.

For more information, see the LimitRange code example.

Configuring the failure timeout

You can use the timeout field to set the TaskRun's desired timeout value. If you do not specify this value for the TaskRun, the global default timeout value applies. If you set the timeout to 0, the TaskRun will have no timeout and will run until it completes successfully or fails from an error.

The global default timeout is set to 60 minutes when you first install Tekton. You can set a different global default timeout value using the default-timeout-minutes field in config/config-defaults.yaml. If you set the global timeout to 0, all TaskRuns that do not have a timeout set will have no timeout and will run until it completes successfully or fails from an error.

The timeout value is a duration conforming to Go's ParseDuration format. For example, valid values are 1h30m, 1h, 1m, 60s, and 0.

If a TaskRun runs longer than its timeout value, the pod associated with the TaskRun will be deleted. This means that the logs of the TaskRun are not preserved. The deletion of the TaskRun pod is necessary in order to stop TaskRun step containers from running.

Specifying `ServiceAccount' credentials

You can execute the Task in your TaskRun with a specific set of credentials by specifying a ServiceAccount object name in the serviceAccountName field in your TaskRun definition. If you do not explicitly specify this, the TaskRun executes with the credentials specified in the configmap-defaults ConfigMap. If this default is not specified, TaskRuns will execute with the default service account set for the target namespace.

For more information, see ServiceAccount.

Monitoring execution status

As your TaskRun executes, its status field accumulates information on the execution of each Step as well as the TaskRun as a whole. This information includes start and stop times, exit codes, the fully-qualified name of the container image, and the corresponding digest.

Note: If any Pods have been OOMKilled by Kubernetes, the TaskRun is marked as failed even if its exit code is 0.

The following example shows the status field of a TaskRun that has executed successfully:

completionTime: "2019-08-12T18:22:57Z"
conditions:
  - lastTransitionTime: "2019-08-12T18:22:57Z"
    message: All Steps have completed executing
    reason: Succeeded
    status: "True"
    type: Succeeded
podName: status-taskrun-pod-6488ef
startTime: "2019-08-12T18:22:51Z"
steps:
  - container: step-hello
    imageID: docker-pullable://busybox@sha256:895ab622e92e18d6b461d671081757af7dbaa3b00e3e28e12505af7817f73649
    name: hello
    terminated:
      containerID: docker://d5a54f5bbb8e7a6fd3bc7761b78410403244cf4c9c5822087fb0209bf59e3621
      exitCode: 0
      finishedAt: "2019-08-12T18:22:56Z"
      reason: Completed
      startedAt: "2019-08-12T18:22:54Z"

The following tables shows how to read the overall status of a TaskRun:

status reason completionTime is set Description
Unknown Started No The TaskRun has just been picked up by the controller.
Unknown Pending No The TaskRun is waiting on a Pod in status Pending.
Unknown Running No The TaskRun has been validate and started to perform its work.
Unknown TaskRunCancelled No The user requested the TaskRun to be cancelled. Cancellation has not be done yet.
True Succeeded Yes The TaskRun completed successfully.
False Failed Yes The TaskRun failed because one of the steps failed.
False [Error message] No The TaskRun encountered a non-permanent error, and it's still running. It may ultimately succeed.
False [Error message] Yes The TaskRun failed with a permanent error (usually validation).
False TaskRunCancelled Yes The TaskRun was cancelled successfully.
False TaskRunTimeout Yes The TaskRun timed out.

When a TaskRun changes status, events are triggered accordingly.

Monitoring Steps

If multiple Steps are defined in the Task invoked by the TaskRun, you can monitor their execution status in the status.steps field using the following command, where <name> is the name of the target TaskRun:

kubectl get taskrun <name> -o yaml

The exact Task Spec used to instantiate the TaskRun is also included in the Status for full auditability.

Steps

The corresponding statuses appear in the status.steps list in the order in which the Steps have been specified in the Task definition.

Monitoring Results

If one or more results fields have been specified in the invoked Task, the TaskRun's execution status will include a Task Results section, in which the Results appear verbatim, including original line returns and whitespace. For example:

Status:
  # […]
  Steps:
  # […]
  Task Results:
    Name:   current-date-human-readable
    Value:  Thu Jan 23 16:29:06 UTC 2020

    Name:   current-date-unix-timestamp
    Value:  1579796946

Cancelling a TaskRun

To cancel a TaskRun that's currently executing, update its status to mark it as cancelled.

When you cancel a TaskRun, the running pod associated with that TaskRun is deleted. This means that the logs of the TaskRun are not preserved. The deletion of the TaskRun pod is necessary in order to stop TaskRun step containers from running.

Example of cancelling a TaskRun:

apiVersion: tekton.dev/v1alpha1
kind: TaskRun
metadata:
  name: go-example-git
spec:
  # […]
  status: "TaskRunCancelled"

Debugging a TaskRun

Breakpoint on Failure

TaskRuns can be halted on failure for troubleshooting by providing the following spec patch as seen below.

spec:
  debug:
    breakpoint: ["onFailure"]

Upon failure of a step, the TaskRun Pod execution is halted. If ths TaskRun Pod continues to run without any lifecycle change done by the user (running the debug-continue or debug-fail-continue script) the TaskRun would be subject to TaskRunTimeout. During this time, the user/client can get remote shell access to the step container with a command such as the following.

kubectl exec -it print-date-d7tj5-pod-w5qrn -c step-print-date-human-readable

Debug Environment

After the user/client has access to the container environment, they can scour for any missing parts because of which their step might have failed.

To control the lifecycle of the step to mark it as a success or a failure or close the breakpoint, there are scripts provided in the /tekton/debug/scripts directory in the container. The following are the scripts and the tasks they perform :-

debug-continue: Mark the step as a success and exit the breakpoint.

debug-fail-continue: Mark the step as a failure and exit the breakpoint.

More information on the inner workings of debug can be found in the Debug documentation

Code examples

To better understand TaskRuns, study the following code examples:

Example TaskRun with a referenced Task

In this example, a TaskRun named read-repo-run invokes and executes an existing Task named read-task. This Task uses a git input resource that the TaskRun references as go-example-git.

apiVersion: tekton.dev/v1alpha1
kind: PipelineResource
metadata:
  name: go-example-git
spec:
  type: git
  params:
    - name: url
      value: https://github.com/pivotal-nader-ziada/gohelloworld
---
apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
  name: read-task
spec:
  resources:
    inputs:
      - name: workspace
        type: git
  steps:
    - name: readme
      image: ubuntu
      script: cat workspace/README.md
---
apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  name: read-repo-run
spec:
  taskRef:
    name: read-task
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: go-example-git

Example TaskRun with an embedded Task

In this example, a TaskRun named build-push-task-run-2 directly executes a Task from its definition embedded in the TaskRun's taskSpec field:

apiVersion: tekton.dev/v1alpha1
kind: PipelineResource
metadata:
  name: go-example-git
spec:
  type: git
  params:
    - name: url
      value: https://github.com/pivotal-nader-ziada/gohelloworld
---
apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  name: build-push-task-run-2
spec:
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: go-example-git
  taskSpec:
    resources:
      inputs:
        - name: workspace
          type: git
    steps:
      - name: build-and-push
        image: gcr.io/kaniko-project/executor:v0.17.1
        # specifying DOCKER_CONFIG is required to allow kaniko to detect docker credential
        env:
          - name: "DOCKER_CONFIG"
            value: "/tekton/home/.docker/"
        command:
          - /kaniko/executor
        args:
          - --destination=gcr.io/my-project/gohelloworld

You can also embed resource definitions in your TaskRun. In the example below, a git resource definition provides input for the TaskRun named read-repo:

apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  name: read-repo
spec:
  taskRef:
    name: read-task
  resources:
    inputs:
      - name: workspace
        resourceSpec:
          type: git
          params:
            - name: url
              value: https://github.com/pivotal-nader-ziada/gohelloworld

Reusing a Task

The following example illustrates the reuse of the same Task. Below, you can see several TaskRuns that instantiate a Task named dockerfile-build-and-push. The TaskRuns reference different Resources as their inputs. See Building and pushing a Docker image for the full definition of this example Task.

This TaskRun builds mchmarny/rester-tester:

# This is the referenced PipelineResource
metadata:
  name: mchmarny-repo
spec:
  type: git
  params:
    - name: url
      value: https://github.com/mchmarny/rester-tester.git
# This is the TaskRun
spec:
  taskRef:
    name: dockerfile-build-and-push
  params:
    - name: IMAGE
      value: gcr.io/my-project/rester-tester
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: mchmarny-repo

This TaskRun builds the wget builder from googlecloudplatform/cloud-builder:

# This is the referenced PipelineResource
metadata:
  name: cloud-builder-repo
spec:
  type: git
  params:
    - name: url
      value: https://github.com/googlecloudplatform/cloud-builders.git
# This is the TaskRun
spec:
  taskRef:
    name: dockerfile-build-and-push
  params:
    - name: IMAGE
      value: gcr.io/my-project/wget
    # Optional override to specify the subdirectory containing the Dockerfile
    - name: DIRECTORY
      value: /workspace/wget
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: cloud-builder-repo

This TaskRun builds the docker builder from googlecloudplatform/cloud-builder with 17.06.1:

# This is the referenced PipelineResource
metadata:
  name: cloud-builder-repo
spec:
  type: git
  params:
    - name: url
      value: https://github.com/googlecloudplatform/cloud-builders.git
# This is the TaskRun
spec:
  taskRef:
    name: dockerfile-build-and-push
  params:
    - name: IMAGE
      value: gcr.io/my-project/docker
    # Optional overrides
    - name: DIRECTORY
      value: /workspace/docker
    - name: DOCKERFILE_NAME
      value: Dockerfile-17.06.1
  resources:
    inputs:
      - name: workspace
        resourceRef:
          name: cloud-builder-repo

Using custom ServiceAccount credentials

The example below illustrates how to specify a ServiceAccount to access a private git repository:

apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  name: test-task-with-serviceaccount-git-ssh
spec:
  serviceAccountName: test-task-robot-git-ssh
  resources:
    inputs:
      - name: workspace
        type: git
  steps:
    - name: config
      image: ubuntu
      command: ["/bin/bash"]
      args: ["-c", "cat README.md"]

In the above code snippet, serviceAccountName: test-build-robot-git-ssh references the following ServiceAccount:

apiVersion: v1
kind: ServiceAccount
metadata:
  name: test-task-robot-git-ssh
secrets:
  - name: test-git-ssh

And name: test-git-ssh references the following Secret:

apiVersion: v1
kind: Secret
metadata:
  name: test-git-ssh
  annotations:
    tekton.dev/git-0: github.com
type: kubernetes.io/ssh-auth
data:
  # Generated by:
  # cat id_rsa | base64 -w 0
  ssh-privatekey: LS0tLS1CRUdJTiBSU0EgUFJJVk.....[example]
  # Generated by:
  # ssh-keyscan github.com | base64 -w 0
  known_hosts: Z2l0aHViLmNvbSBzc2g.....[example]

Running Step Containers as a Non Root User

All steps that do not require to be run as a root user should make use of TaskRun features to designate the container for a step runs as a user without root permissions. As a best practice, running containers as non root should be built into the container image to avoid any possibility of the container being run as root. However, as a further measure of enforcing this practice, TaskRun pod templates can be used to specify how containers should be run within a TaskRun pod.

An example of using a TaskRun pod template is shown below to specify that containers running via this TaskRun's pod should run as non root and run as user 1001 if the container itself does not specify what user to run as:

apiVersion: tekton.dev/v1beta1
kind: TaskRun
metadata:
  generateName: show-non-root-steps-run-
spec:
  taskRef:
    name: show-non-root-steps
  podTemplate:
    securityContext:
      runAsNonRoot: true
      runAsUser: 1001

If a Task step specifies that it is to run as a different user than what is specified in the pod template, the step's securityContext will be applied instead of what is specified at the pod level. An example of this is available as a TaskRun example.

More information about Pod and Container Security Contexts can be found via the Kubernetes website.

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License.