Skip to main content

Executions Management

Last updated on

The Executions Management page gives you account-level visibility into all queued and running pipeline executions. You can see where queued executions sit in the queue, monitor executions that are currently running, and abort executions that are no longer needed.

Feature flag

This feature is behind the feature flag PIPE_QUEUED_PIPELINE_OBSERVABILITY. Contact Harness Support to enable it.

What will you learn in this topic?​

Access Executions Management​

Go to Account Settings, select Security and Governance, and then select Executions Management.

Executions Management location in Account Settings under Security and Governance
Account Admin required

Only users with Account Admin permissions can access the Executions Management page. This is an account-level view showing queued and running pipelines across all organizations and projects.

The Executions Management page lists two kinds of executions across your entire account:

  • Queued: Executions that are waiting to run.
  • Running: Executions that are currently running.

When you trigger multiple pipelines simultaneously, Harness queues some executions based on the resource constraints and concurrency limits you have configured. Instead of failing or rejecting new executions, Harness holds them in a queue and runs them in order when resources become available.

Executions sit in the queue for three main reasons: the pipeline has resource constraints configured, the maximum concurrent executions limit is reached, or the pipeline is waiting for another pipeline to release a lock it needs.

Queue behavior​

Each row represents a queued or running execution. The table shows the following information:

  • Queue No: The execution's position in the global account queue, where position 1 is next to run. Running executions show a dash (-) because they are no longer in the queue.
  • Pipeline Name: The pipeline name and its execution ID.
  • Status: The execution state, such as RUNNING or a queued reason.
  • Project and Organization: The project and organization the execution belongs to.
  • Trigger Summary: The way the execution started, such as a manual run, a cron schedule, or a webhook.
  • Executed By: The user or system that initiated the run.
Executions Management table showing running executions with the Status column
Queue position

Queue positions are calculated globally across your entire account. When you apply filters to narrow the list, the queue numbers stay the same. You might see positions such as 5, 240, and 1320 with gaps in between. Those gaps represent executions that are still queued but hidden by your current filters.

Filter executions​

Use the Filter option to narrow the list. The available filters include:

  • Search: Find executions by pipeline name or identifier.
  • Organization and Project: Show executions from specific organizations and projects. The project filter updates to show only projects within the selected organizations.
  • Status: Filter by execution state, including Queued Execution Concurrency Reached, Queued Plan Creation, and Running.
  • Priority: Focus on high, medium, or low priority executions.
  • Trigger Type: Filter by how the execution was triggered, such as a manual run, a cron schedule, or a webhook.
  • Pipeline Tags: Filter by the tags applied to pipelines.
  • Timeframe: Show executions from a specific period, such as the last 7, 30, or 90 days, or a custom range.

Select Add Filter to apply more filters. Select Save to keep a filter set, or Reset to clear all filters.

Abort executions​

You can abort both queued and running executions. Aborting a queued execution removes it from the queue before it starts. Aborting a running execution stops it. You might abort executions when you triggered the wrong pipeline, or when newer changes make an older execution unnecessary.

Perform the following steps to abort executions:

  1. Select one or more executions using the checkboxes.
  2. Select Abort at the top of the list. The button shows how many executions are selected.
  3. Review your selection and confirm.

Harness attempts to abort each selected execution. Some aborts might fail if an execution completed while you were making your selection. The results show which aborts succeeded and which could not be completed.

Abort is permanent

Once you abort an execution, there is no way to resume it. Trigger the pipeline again if you want it to run.

Common scenarios​

Abort outdated deployments​

When multiple deployment pipelines queue up for the same service during active development, you often need only the latest one. Filter by the relevant project, select the older executions, and abort them to keep just the most recent one.

Clean up test executions​

If you are developing or testing pipelines and have accumulated many test runs, filter by your test project and abort them. This frees up capacity for production pipelines that need to run.

Identify bottlenecks​

When you notice executions with very high queue positions (100+), it indicates a bottleneck. Look at the types of pipelines queued and their resource constraints. You might need to adjust concurrency limits, optimize pipeline execution time, or review resource constraint settings to reduce queue buildup.

Limitations​

The Executions Management page shows queued and running executions. For completed pipelines, use the Pipeline Execution History view instead.

You cannot reorder the execution queue or change priorities from this page. The queue order is determined by when executions were created and their configured priority settings.

Queue positions are calculated when the page loads. As executions complete and new ones are added, positions change. Refresh the page to see the current queue state.

The table displays a maximum of 100 executions per page.

Next steps​