> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/continuous-delivery/3.0/new-to-deployments/cd-integrations.md).
# What's Supported
Supported deployment targets, manifest stores, and artifact sources for Harness Deployments.
This page lists the deployment targets, manifest stores, and artifact sources supported in Harness Deployments.
{% tabs %}
{% tab title="Deployments" %}
### Kubernetes
**Supported deployment strategies**
| Strategy | Description |
| ------------ | --------------------------------------------------------------------------------------- |
| Rolling | Incrementally replaces pods batch by batch with automatic rollback |
| Canary | Deploys a subset of pods alongside stable pods, then promotes with a rolling deploy |
| Blue-green | Maintains two pod sets and swaps service selectors at cutover |
| Blank canvas | No managed steps — build your own sequence for Jobs, CronJobs, and supporting resources |
**Supported manifest types**
K8s Manifest, Values YAML, Helm Chart, Kustomize, Kustomize Patches, OpenShift Template
**Supported infrastructure providers**
Kubernetes (direct), Google Kubernetes Engine (GKE), Microsoft Azure (AKS), Amazon Elastic Kubernetes Service (EKS), Rancher
**Other**
| Capability | Supported |
| ----------------------------------------- | --------- |
| OpenShift (DeploymentConfig, `oc` client) | ✅ |
| Server-side apply | ✅ |
| Manifest pruning | ✅ |
| Automatic rollback on failure | ✅ |
| Kubernetes Steady State Check | ✅ |
| Harness release history tracking | ✅ |
| Delegate 1.0 and Delegate 3.x | ✅ |
***
### Helm
**Supported deployment strategies**
| Strategy | Description |
| ------------ | ----------------------------------------------------------------------------------------- |
| Basic deploy | Runs `helm upgrade --install` in a single phase |
| Canary | Deploys a canary Helm release at a specified instance count, then deletes after promotion |
| Blue-green | Deploys to a stage Helm release, swaps traffic, then cleans up the old release |
**Supported Helm versions**
| Version | Support |
| ------- | ---------- |
| Helm V3 | ✅ |
| Helm V2 | Deprecated |
**Supported chart store types**
Harness Code, GitHub, Git, GitLab, Bitbucket, Azure Repos, Amazon S3, Google Cloud Storage, HTTP Helm
**Other**
| Capability | Supported |
| ----------------------------- | --------------------------------------- |
| `helm test` after deploy | ✅ |
| Server-side chart rendering | ✅ |
| Automatic rollback on failure | ✅ (via Helm Rollback step) |
| Ignore failed release history | ✅ |
| Manifest pruning | ❌ (Helm manages release state natively) |
| Delegate 1.0 and Delegate 3.x | ✅ |
***
### Google Cloud Run
**Supported deployment strategies**
| Strategy | Description |
| ------------ | ------------------------------------------------------------------------------------------------------------------------------ |
| Basic | Deploys a new Cloud Run revision and routes 100% of traffic to it immediately with automatic rollback on failure |
| Canary | Deploys a new revision and progressively shifts traffic via explicit Traffic Shift steps, with rollback to the pre-shift state |
| Blank canvas | No managed step sequence — compose Deploy, Traffic Shift, Rollback, and Job steps in any order |
**Supported artifact types**
| Artifact type | Supported |
| ------------------------------------------ | --------- |
| Google Artifact Registry (container image) | ✅ |
| Docker Hub (container image) | ✅ |
**Supported manifest store types**
Service manifests (Cloud Run YAML / Knative serving spec) can be stored in: Harness Code, GitHub, Git, GitLab, Bitbucket, Azure Repos
**Other**
| Capability | Supported |
| --------------------------------------------------- | --------- |
| Automatic rollback on failure | ✅ |
| Traffic routing to named revisions | ✅ |
| Cloud Run Jobs (batch / scheduled workloads) | ✅ |
| Instance sync via Google Cloud Monitoring | ✅ |
| Service Account key authentication | ✅ |
| OIDC / Workload Identity authentication | ✅ |
| Inherit from Delegate authentication | ✅ |
| Containerized step execution (`runtime.kubernetes`) | ✅ |
| Harness Cloud runtime | ✅ |
***
### AWS Lambda
**Supported deployment strategies**
| Strategy | Description |
| -------- | ------------------------------------------------------------------------------------------------------------------ |
| Rolling | Publishes a new function version and routes 100% of traffic to it immediately with automatic rollback on failure |
| Canary | Publishes a new function version and progressively shifts traffic via a weighted alias (10% → 100%), with rollback |
**Supported artifact types**
| Artifact type | Rolling | Canary |
| ---------------------------- | ------- | ------------------------ |
| Amazon S3 (ZIP) | ✅ | ✅ |
| Amazon ECR (container image) | ✅ | ✅ |
| Artifactory (generic ZIP) | ✅ | ✅ |
| Nexus 3 (raw ZIP) | ✅ | ✅ (runtime-input matrix) |
Nexus 3 is supported for canary only as a ZIP artifact. Container-image canary deployments require ECR.
**Supported manifest store types**
Function definition and alias definition manifests can be stored in: Harness Code, GitHub, Git, GitLab, Bitbucket, Azure Repos, AWS S3
**Other**
| Capability | Supported |
| ----------------------------------------------------------- | ------------------------------------------------------------------------------ |
| Automatic rollback on failure (rolling) | ✅ |
| Automatic rollback on failure (canary) | ✅ |
| Function alias management | ✅ (optional for rolling; required for ZIP canary; auto-managed for ECR canary) |
| Secrets in function manifest | ✅ |
| Pipeline-level input expressions | ✅ |
| Multi-service × multi-infra stage matrix (2 × 2 = 4 stages) | ✅ |
| Service runtime inputs (V1 `inputs_yaml` overlay) | ✅ |
| Containerized step execution (`runtime.kubernetes`) | ✅ (required for all Lambda steps) |
| {% endtab %} | |
{% tab title="Manifest Sources" %}
Manifests and config files can be stored in the following sources. The table shows which sources are supported per manifest type.
| Manifest type | Harness Code | GitHub | Git | GitLab | Bitbucket | Azure Repos | AWS S3 | Google Cloud Storage | HTTP Helm | Custom Remote |
| --------------------------------------------- | :----------: | :----: | :-: | :----: | :-------: | :---------: | :----: | :------------------: | :-------: | :-----------: |
| K8s Manifest | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | | ✅ |
| Values YAML | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | ✅ |
| Kustomize | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | | |
| Kustomize Patches | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | | ✅ |
| OpenShift Template | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | | ✅ |
| Helm Chart | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | |
| Cloud Run Service Manifest | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | | |
| Lambda Function Definition / Alias Definition | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | | |
| {% endtab %} | | | | | | | | | | |
{% tab title="Artifacts" %}
The table shows which artifact sources are supported per deployment type.
| Deployment type | Harness Artifact Registry | Docker Hub | Amazon ECR | GCR\* | ACR | Artifactory | Nexus 3 | Google Artifact Registry | GitHub Package Registry | Custom |
| ---------------- | :-----------------------: | :--------: | :--------: | :---: | :-: | :---------: | :-----: | :----------------------: | :---------------------: | :----: |
| Kubernetes | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| Helm | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | ✅ | ✅ |
| Google Cloud Run | | ✅ | | | | | | ✅ | | |
| AWS Lambda | | | ✅ | | | ✅ | ✅ | | | |
**GCR** is deprecated. Migrate to Google Artifact Registry.
AWS Lambda also supports **Amazon S3** as a ZIP artifact source (not shown above as S3 is a file store, not a container registry). The full Lambda artifact matrix: S3 ✅, ECR ✅, Artifactory ✅, Nexus 3 ✅.
{% endtab %}
{% endtabs %}
{% @harness-feedback/feedback %}