Continuous Delivery & GitOps
Blog →
Continuous Delivery & GitOps

Q3 FY27 Product Update: Harness Pipeline | Harness Blog

Discover the latest features, enhancements, and updates in the Q3 FY27 product update for Harness Pipeline

TL;DR

Q3 FY27 gives you more control over how pipelines run, with project-level concurrency partitions, pipeline identities, and a customizable run form. Policy as Code also expands: you can run policy sets on your own infrastructure and preview a change with Blast Radius Analysis before you save it.

Q3 FY27 advances Harness Pipeline with project-level concurrency partitions, pipeline identities, a customizable run form, template usage visibility, and a significant expansion of policy governance. The quarter's headline capabilities, running policy sets on your own infrastructure and Blast Radius Analysis in the OPA policy editor, let your policies reach private systems and show you what a change would pass or fail before you save it. For prior context, see the Q2 and Q1 Pipeline updates.

Pipeline Execution

More control over how many pipelines run, which stages run, and where your pipelines live.

Project-level pipeline concurrency

Split your account concurrency limit into High-Priority and Low-Priority partitions so critical projects keep reserved execution slots. High-priority executions can spill into unused Low capacity, while Low-priority executions use only the Low partition. Configured by account admins under Concurrency Management and enabled through Harness Support.
Learn more

Selective stage execution for DAG pipelines

Run selected stages of a DAG pipeline instead of the whole pipeline. Harness rebuilds the dependency chain from the stages you choose, so you no longer select every dependency by hand.
Learn more

Move a pipeline between projects

Move a pipeline to a different project or organization along with its input sets and triggers; execution history migrates in the background. A move cannot be undone, and hardcoded identifier references, such as chained pipelines and trigger webhook URLs, are not updated. In Beta; enabled through Harness Support.
Learn more

Add tags after an execution finishes

Add tags to an execution after it reaches Success, Failed, or Aborted using the execution tags API. New keys are added, and existing keys are overwritten. API-only; enabled through Harness Support.
Learn more

Pipeline Timeout column on the Executions page

The Executions list now shows a sortable Pipeline Timeout column with each execution's expiry time, calculated from its start time and the pipeline's configured timeout.
Learn more

Publish Events step

A new step reports the artifacts your pipeline publishes to Release Management so Artifact Tracker can record them and compute changesets. Add it in a Custom stage after the stage that publishes your artifact; it is not supported inside a CI stage.
Learn more

Pipeline identities

A pipeline, stage, or step can now authenticate to OIDC-capable systems such as AWS, GCP, Azure, and HashiCorp Vault without long-lived credentials. Define identities in YAML; in a Shell Script step, the identity name becomes an environment variable holding the token.
Learn more

Page, search, and filter large looping strategies

Page through, search, and filter the stage instances of a looping strategy in the execution and console views, so large matrix and repeat runs stay easy to navigate.
Learn more

Run Experience and Variables

A cleaner run form, safer inputs, and faster expressions.

Customizable pipeline run form

Customize the form shown when you run a pipeline: add labels and descriptions, show inputs only when a condition is met, and restrict allowed values based on other inputs. Allowed-value rules are enforced across the form, input sets, triggers, and API. In Beta; enabled through Harness Support.
Learn more

Split execution JSON for expressions

The execution JSON used by expressions is now stored per stage instead of as one large document, so pipeline-level and per-stage expression data are kept as separate documents. This helps pipelines with parallel stages.
Learn more

Zoom to fit by default in pipeline graphs

Pipeline graphs open zoomed to fit and remember your zoom per pipeline as you switch between views. Turn on Open Pipeline Graph Zoomed to Fit under Default Settings, Pipeline. Enabled through Harness Support.
Learn more

Save blank fields as empty strings

A new account-level setting preserves an empty string in an input variable instead of converting it to null. Turn it on when your steps or scripts need to tell a deliberately empty value from an unset one.

Pipeline breakpoints

Pause a pipeline run at a specific step to check it before it continues. The run waits at the breakpoint until you resume it.

Git branch list caching

Harness now caches the list of Git branches, so it makes fewer calls to your Git provider. This reduces load on your Git provider and helps branch lists load quickly.
Learn more

Governance and Policy

Policy as Code gets more reach, more safety, and a way to test changes before they land.

Type and Action filters on Policy Sets

Filter the Policy Sets list by entity type or trigger action, the same way you filter on the Policy Evaluations page. Use the filters to narrow a long list to the policy sets that apply to a given entity or action.
Learn more

Run policy sets on your own infrastructure

Run On Run pipeline policy sets on your own Kubernetes or VM infrastructure instead of Harness SaaS by turning on Run this policy set on your own environment. Policies can then use Harness secrets and reach internal systems such as ServiceNow, Jira, or private endpoints. Harness adds an OPA Evaluation stage automatically.
Learn more

Rego network built-ins removed on Harness SaaS

The http.send and net.lookup_ip_addr Rego built-ins are removed on Harness SaaS. Policies that use them should run on your own infrastructure, where they remain available.
Learn more

Large OPA evaluation outputs are stripped

Policy outputs larger than half the evaluation input size are replaced with a small placeholder to keep evaluation storage lean. Status, deny messages, severity, and errors are unaffected.
Learn more

Warn and Continue shows a Passed with warning status

When a Warn and Continue policy returns a warning, the step, stage, and pipeline now show "Passed with warning" instead of "Success". Hover to see the warning message; a failure still takes priority.
Learn more

Blast Radius Analysis in the OPA policy editor

Replay a policy against recent evaluations to see what would pass or fail before you save. Open Blast Radius Analysis in the policy editor and select Calculate: it runs your unsaved policy against the last 30 days of evaluations, or sample payloads if there are none.
Learn more

Templates and Input Sets

More visibility into template adoption, and more flexibility in how templates are composed.

Template usage dashboard

See which pipelines use each template version across your account in the Pipeline Template Usage Dashboard, and export the data as CSV or JSON. Useful for planning deprecations.
Learn more

Template overrides for notification rules

Template owners can allow pipelines to override notification rules from the Notify tab. An override replaces the template's rules rather than merging with them. Enabled through Harness Support.
Learn more

Two-level nested insert in templates

A stage template that contains its own step insert can now be inserted into a pipeline template's stage insert, so teams can add a stage and still add their own steps inside it.
Learn more

Notifications and Integrations

Step and stage info in WaitingForUserAction and PipelineResumed notifications

WaitingForUserAction notifications now list each waiting step, and PipelineResumed notifications list the stages that resumed. Enabled through Harness Support.
Learn more

Notify user groups from Webhook and Datadog channels

Pipeline notifications can now notify user groups through the Webhook and Datadog channels. Configure each group's channels under Notification Preferences in its Access Control settings.
Learn more

Azure Repos connector supports Azure DevOps Server

The Azure Repos connector now supports self-hosted Azure DevOps Server 2020 and newer over HTTP or SSH for Git repositories. Requires delegate 26.10.90209 or newer; enabled through Harness Support.
Learn more

What's Next

We are continuing to invest in pipeline execution, templates, and governance. To explore everything Harness Pipeline offers today, head to the Harness Developer Hub.

Get Started

Explore the documentation linked above for each feature, or reach out to your Harness account team to see these capabilities in your environment.

← Previous:
Next: →‍

Related Resources

No items found.

Get Started

Get Started with Harness AI

Try the full platform free. No module restrictions, no credit card.

Vishal Vishwaroop
Senior Developer Relations Engineer
Vishal Vishwaroop is a senior Developer Relations Engineer.
vishal-vishwaroop
Vishal Vishwaroop