The plugin allows working with bundles. It is possible to create a bundle containing up to 10 Tekton resources and push it to a public/private registry and retrieve/download any existing one.
The Getting Started page shows all the features provided by the plugin and explain how to invoke and use them
The Tekton Hub was integrated through a search dialog that allows users to search to tasks and pipelines from the Tekton Hub to be added to the cluster. The content from the Tekton Hub is now integrated into a new view that is displayed below the cluster Tekton resources and is automatically synced with the current IntelliJ project: if you working on a Java project, then only tasks and pipelines relevant for Java will be displayed:
The Tekton plugin leverages the Kubernetes plugin's toolbar to push/pull resources
When creating/opening a pipeline in the editor, it analyzes the yaml content to verify that all tasks/clustertasks used actually exists on cluster. If not, the user is notified.
A Task and/or a TaskRun can now be started in debug mode. The container where the taskrun will be executed will be kept running even when the run fails. A terminal connected to the container will be opened so you can interact with it and identify any possible misbehavior/error. To use the debug feature, you must have access to a Kubernetes or OpenShift cluster with Tekton installed with a version greater than 0.26.0 and enable alpha mode.
To enable alpha mode you can run this command
kubectl patch configmap/feature-flags \
-n <namespace> \
--type merge \
-p '{"data":{"enable-api-fields":"alpha"}}'
N.B: By activating alpha mode, you will enable all features that are still in alpha and are not considered stable (e.g debug). Be aware that you could meet some errors during commands execution.
When a Kubernetes resource is retrieved from a cluster, its YAML content stored some internal metadata that users may not need to care about. So there is now a new command called XXX to remove it. There is also a setting so that this command is automatically executed when a Tekton resource is displayed in the editor.
When start a task or pipeline, it is now possible to specifiy a VolumeClaimTemplate. Also, if a PersistentVolumeClaim is chosen and needs to be created, it can be done from within the wizard.
It is now possible to save an inline task from a pipeline as a task or cluster task from within the pipeline editor. The inline task is saved to the cluster and replaced in the pipeline definition by a task reference.
It is possible to find all references of a task within the cluster. During a deletion the user is also informed if a task is used somewhere to prevent possible errors.
Logs have been extended to work with EventListeners and they can be now redirected into the editor
It is possible to set preferences to change the default plugin behaviour, such as delete all related resources automatically or avoid showing the wizard if a pipeline/task has no inputs
When adding a task to a pipeline definition, code completion gets also activated on single inputs (param, input/output resource and workspace). If the input contains a default value, then this value is also inserted into the pipeline definition.
When a new pipeline is started, its execution status is new reported asynchronously as a notification.
This can be turned off through a setting under Settings -> Tools -> Tekton Pipelines by Red Hat:
Task ordering can be managed through the runAfter field. However, uses of this field is not allowed when specifying finally tasks. Such wrong usages will be reported when editing your pipeline through an error marker:
When starting a task or a pipeline, it is possible to see the generated taskrun or pipelinerun payload with the Show Preview option:
When a pipeline or task is defining parameters, resources or workspaces that are not being used in the pipeline or task definition, then this parameter, resource or workspace is grey highlighted just like an unused variable is highlighted in a classic code editor.
When a name of a task is to be specified in a pipeline definition, then the list of the normal tasks was provided. The list of existing cluster tasks is now available, with the (clusterTask) suffix to distinguish from normal tasks.
When a task is selected inside a pipeline definition during code completion, the list of parameters to be passed to this task is inserted into the pipeline definition. If the definition of a parameter contains a default value, then this value is also inserted into the pipeline definiton. On the same topic, if the task definition contains a list of workspaces, then a list of workspaces will be inserted into the pipeline definition as well.
Sometimes, it may be hard to understand why a pipeline failed or why it is stuck. We added a new troubleshooting tool that will inspect low level Kubernetes resources related to your task or pipeline execution and extract information that may be helpful.
It is now much simple to add a trigger to an existing pipeline. We added a new wizard that allows you to associate one or more trigger bindings to a pipeline (and if you don't have bindings already, you can create it while adding the trigger).
Like the pipeline editor, the pipelinerun editor has been enhanced to include a visual representation of the pipeline workflow and status. Conditions and finally clauses are also supported.
It is now possible to specify a run prefix when starting a task or pipeline. This allows to better identify the execution in the set of runs
The dynamic input of a when expression can be easily specified as code completion can now retrieve the results from other tasks of the pipeline.
Any task using input or output resources can be specified in the pipeline workflow and values for the resource field will be managed through code completion
The pipeline editor has been enhanced to include a visual representation of the pipeline workflow. Conditions and finally clauses are also supported.
As starting a task or pipeline may require a lot of inputs, it is also possible to start a task or pipeline using the settings from a previous run. The start wizard will have these input values initialized from the values used in the previous run instead of getting the default value.
When a parameter/workspace/resource is to be set on a task, code assist will present the candidate values.
Tasks inside a pipeline can be sequence through the runAfter field. When code completion is activate on such a field, the list of available tasks in the pipeline is proposed:
When a task is to be conditionnaly executed through conditions (using the conditionRef field), the list of available conditions is proposed:
When a task or pipeline resource is to be deleted, you now have the choice to delete associated resources (task runs / pipeline runs) through an option:
The taskruns/pipelineruns are now displayed ordered by the starting time (most recent started run is displayed first), allowing users to keep focusing on recent work
A task or pipeline run still active (in the running state) can be canceled through a new Cancel action:
When a task or pipeline is started, the logs view is automatically displayed
It is now possible to delete several different elements, even if they are of the same nature (ie task runs and pipeline runs)
Communication with the cluster is now supported automatically. The plugin will honor proxy settings from IntelliJ so proxy is configured by default or if proxy is associated with the cluster API server hostname, then communication from this plugin to the cluster will go through the configured proxy.
This plugin now supports the Tekton Triggers resources (if Tekton Triggers runtime is installed on the cluster):
- TriggerTemplates
- TriggerBindings
- ClusterTriggerBindings
- EventListeners
The following commands have been added:
- Start Last Run (for pipelines and tasks)
- Edit any Tekton resource by double clicking on the resource in the tree
The Tekton tree now displays:
- TaskRuns
- PipelineRuns
- Conditions
as separate nodes.
Code assist and syntax validation based on JSON schemas has been added to the Tekton resources editing experience in addition to the code snippets.
To use the plugin, developers can deploy Tekton Pipelines into a OpenShift or Kubernetes cluster. OpenShift clusters can be provisioned using the following ways:
- OpenShift Local
- Developer Sandbox for Red Hat OpenShift
- Red Hat OpenShift Cluster Manager - This 60-day, self-supported trial lets you install and run Red Hat OpenShift Container Platform on infrastructure you manage.
- Red Hat OpenShift Dedicated - Red Hat OpenShift Dedicated is a fully managed service of Red Hat OpenShift on Amazon Web Services (AWS) and Google Cloud.
- Azure Red Hat OpenShift - Azure Red Hat OpenShift is a fully-managed service of Red Hat OpenShift on Azure, jointly engineered, managed and supported by Microsoft and Red Hat.
- Red Hat OpenShift Service on AWS (ROSA) - Build, deploy, and manage Kubernetes applications with Red Hat OpenShift running natively on AWS.
To get started with Tekton and learn how to create and run a simple pipeline for building and deploying containerized app on Kubernetes/OpenShift, you can follow one of these tutorials:
Development of the Tekton Pipelines Plugin is largely following development of the tkn CLI as well as Tekton Pipelines which are both under active development and potentially subject to drastic changes.
Below you can find a list of the current actions supported by this plugin to interact with Tekton Pipelines. These are accessible via the context menu when right-clicking an item in the tree.
New- Open an editor with a configuration sample to create a new pipeline/task/clusterTask/resource.Open in Editor- Open the existing pipeline/task/clusterTask/resource configuration in an editor.Start- Start a pipeline/task with user indicated input/output resources and parameters.Delete- Delete the selected pipeline/task/clusterTask/resource.Refresh- Refresh the selected itemShow Logs- Show a dialog to choose the pipelineRun/taskRun to print logs for (only available for Pipeline/Task)Follow Logs- Show a dialog to choose the pipelineRun/taskRun to follow logs for (only available for Pipeline/Task)
Show Logs- Print logs for the selected PipelineRun/TaskRunFollow Logs- Follow logs for the selected PipelineRun/TaskRun
The plugin takes advantage of the normal saving workflow of the IntelliJ IDE. When you finish to edit your configuration
and want to push the changes to the cluster, click on Save All (CTRL + S). A prompt will be shown asking if you want to push the changes.