mTLS Support via Delegates

Configure mutual TLS authentication between delegates and Harness to secure communication with client certificates.

Mutual TLS (mTLS) secures communication between Harness delegates and the Harness platform by requiring both the client and server to verify each other's identity. This page explains how to generate the required certificates, configure mTLS on your Harness account, and enable mTLS on Kubernetes, Helm, and Docker delegates.

---

What you will learn in this topic

By the end of this topic, you will be able to:

• Understand how mTLS works with Harness delegates.

• Create CA and client certificates for mTLS authentication.

• Configure mTLS on Kubernetes, Helm, and Docker delegates.

---

Before you begin

Before you configure mTLS on a delegate, ensure you have the following:

• Harness account with mTLS enabled: Contact Harness Support to enable mTLS for your account. This is an advanced feature that requires configuration by Harness.

• Delegate with modern image format: The delegate must use an image tag in yy.mm.xxxxx format. Legacy delegates are not supported. Go to Delegate image types to understand delegate image formats.

• Access to delegate YAML: Ability to edit the delegate manifest (Kubernetes), Helm values, or Docker run command.

• OpenSSL installed: Generated CA and client certificates.

• Account vanity URL: Your Harness account's custom vanity URL (for example, customer-acme.harness.io). Contact Harness Support if you do not know your vanity URL.

---

Understand mTLS

mTLS is part of the TLS (Transport Layer Security) protocol, which allows the server to verify the authenticity of the client. To achieve that, the client has to provide a client certificate during the TLS handshake, which is verified by the server using a previously configured CA certificate.

Due to security reasons, every customer must create their own CA certificate and signed client certificates. Harness verifies the client certificate at your account's mTLS endpoint, which is your account's vanity URL on port 8443 (for example, https://<vanity_url>.harness.io:8443).

Harness supports the following mTLS modes:

• LOOSE: Both non-mTLS and mTLS delegates are accepted.

• STRICT: Only mTLS delegates are accepted. Any non-mTLS delegates are blocked.

LEGACY *.AGENT.HARNESS.IO ENDPOINTS

Accounts onboarded before the current rollout use a dedicated per-account endpoint of the form <subdomain>.agent.harness.io (for example, customer1.agent.harness.io), without the :8443 port. These endpoints continue to work, so existing setups do not need to change. However, all new mTLS setups use your account's vanity URL on port 8443 (https://<vanity_url>.harness.io:8443) and no longer require a dedicated *.agent.harness.io subdomain. If you have a legacy .agent.harness.io setup, keep using it unless Harness Support advises you to migrate.

---

Create a CA certificate and a client certificate

Harness does not create or distribute the CA and client certificates that are used for mTLS. You must set up the certificates and configure your delegate to use them as part of an mTLS connection.

Project-level certificates are not supported for mTLS delegates. Only one certificate is supported per account. This means you cannot have multiple project or organization level certificates.

In the following examples, OpenSSL is used to create the required certificates. For the Subject, use the text of your choice. It does not have to match the mTLS DNS name or contain harness.io.

Create a CA certificate

Use the following OpenSSL command to create a test CA certificate with no password and 25+ years of validity. You must provide the public portion of the CA certificate (ca.crt) to Harness to enable mTLS.

openssl req -x509 -sha256 -nodes -days 9999 -newkey rsa:2048 \
-subj "/O=Example ORG/CN=CA Cert" -keyout "ca.key" -out "ca.crt"

Create a client certificate

Follow these steps to create a client certificate signed by your CA certificate.

1. Create the configuration used to create the client certificate:

   cat << EOF > "client.cnf"
   [req]
   default_bits = 2048
   prompt = no
   default_md = sha256
   x509_extensions = v3_req
   distinguished_name = dn
   
   [dn]
   O = Example ORG
   CN = Client
   
   [v3_req]
   # empty
   EOF

2. After the configuration file has been created, create a new certificate signing request together with the key pair:

   openssl req -new -config "client.cnf" -nodes -out "client.csr" -keyout "client.key"

3. Using the previously created CA certificate with the certificate signing request, create the final signed client certificate:

   openssl x509 -req -sha256 -days 9999 -extfile client.cnf -extensions v3_req \
   -CAcreateserial -CA "ca.crt" -CAkey "ca.key" \
   -in "client.csr" -out "client.crt"

   Provide the client.crt and client.key to the delegate YAML when you install the delegate.

4. After you create the certificates, provide the public cert of the CA certificate to Harness Support. Harness enables mTLS on your account's vanity URL, and your delegates connect to https://<vanity_url>.harness.io:8443.

   After this, Harness will perform the steps to enable the mTLS.

---

Enable mTLS on delegate

You can enable mTLS on delegates with the image tag in yy.mm.xxxxx format. Legacy delegates are not supported. Go to Delegate image types for more information on delegate types.

Configure mTLS on Kubernetes, Helm, or Docker delegates

Select the appropriate tab below based on your delegate deployment method.

Kubernetes delegate

To enable mTLS on a Kubernetes delegate, create a secret and update the delegate YAML. For an example Kubernetes manifest, go to Sample Kubernetes manifest.

1. From the same folder where you have your client.crt and client.key files, run the following command to create the secret.

   kubectl create secret -n <delegate namespace> generic client-certificate \
     --from-file client.crt=client.crt \
     --from-file client.key=client.key

2. In the Deployment resource of manifest YAML file, make the following updates:

   1. Under spec.template.spec.containers[0].env, update the value for MANAGER_HOST_AND_PORT to https://<vanity_url>.harness.io:8443.

   2. Under spec.template.spec.containers[0].env, update the value for TI_SERVICE_URL to https://<vanity_url>.harness.io:8443/ti-service/.

   If you do not want to update the service URLs (TI_SERVICE_URL), enable the feature flag CI_OVERRIDE_SERVICE_URLS by contacting Harness Support. Additionally, enable this feature flag if you wish to use this feature with STO and SCS steps.

   1. Under spec.template.spec.containers[0].env, add the following YAML.

              - name: DELEGATE_CLIENT_CERTIFICATE_PATH
                value: "/etc/mtls/client.crt"
              - name: DELEGATE_CLIENT_CERTIFICATE_KEY_PATH
                value: "/etc/mtls/client.key"

   2. Under spec.template.spec.containers, add the following YAML.

              volumeMounts:
                - mountPath: /etc/mtls
                  name: client-certificate
                  readOnly: true

   3. Under spec.template.spec, add the following YAML.

            volumes:
              - name: client-certificate
                secret:
                  secretName: client-certificate
                  defaultMode: 400

3. In the ConfigMap resource of manifest YAML file, make the following updates:

   1. Under data.config.yaml, update the value for managerHost to https://<vanity_url>.harness.io:8443.

   2. Under data.config.yaml, add the following YAML.

          clientCertificateFilePath: /etc/mtls/client.crt
          clientCertificateKeyFilePath: /etc/mtls/client.key

4. In the CronJob resource of manifest YAML file, make the following updates:

   1. Under jobTemplate.spec.template.spec.containers.volumeMounts, add the following YAML.

                    - name: client-certificate
                      mountPath: /etc/mtls
                      readOnly: true

   2. Under jobTemplate.spec.template.spec.volumes, add the following YAML.

                  - name: client-certificate
                    secret:
                      secretName: client-certificate

5. Save and apply the manifest.

Helm delegate

To enable mTLS on a Kubernetes delegate, create a secret that will store client.crt and client.key files.

1. From the same folder where you have your client.crt and client.key files, run the following command to create the secret.

   kubectl create secret -n <delegate namespace> generic client-certificate \
     --from-file client.crt=client.crt \
     --from-file client.key=client.key

2. In your helm command add the mTLS.secretName flag to enable the mTLS feature. Setting this flag will mount the secret as a volume to both Delegate and Upgrader components and configure appropriate config options. For example:

   helm upgrade -i helm-delegate --namespace <delegate namespace> --create-namespace \
       harness-delegate/harness-delegate-ng \
     --set delegateName=<delegate name> \
     --set accountId=<account ID> \
     --set delegateToken=<delegate token> \
     --set managerEndpoint=https://<vanity_url>.harness.io:8443 \
     --set delegateDockerImage=harness/delegate:yy.mm.verno \
     --set mTLS.secretName=client-certificate \
     --set replicas=1 \
     --set upgrader.enabled=true

Docker delegate

To enable mTLS on a Docker delegate, configure the Docker run command with certificate volume mounts and environment variables.

1. Copy the following example command.

   docker run -d --cpus=1 --memory=2g -u root -v <path to client certificate>:/etc/mtls/client.crt -v <path to client key>:/etc/mtls/client.key \
     -e DELEGATE_NAME=docker-delegate \
     -e NEXT_GEN="true" \
     -e DELEGATE_TYPE="DOCKER" \
     -e ACCOUNT_ID=<account ID> \
     -e DELEGATE_CLIENT_CERTIFICATE_PATH=/etc/mtls/client.crt \
     -e DELEGATE_CLIENT_CERTIFICATE_KEY_PATH=/etc/mtls/client.key \
     -e DELEGATE_TOKEN=<delegate token> \
     -e MANAGER_HOST_AND_PORT=https://<vanity_url>.harness.io:8443 harness/delegate:yy.mm.verno

2. Update all of the placeholders: <path to client certificate>, <path to client key>, <account ID>, <delegate token>, and <vanity_url>.

3. Run the command.

---

Next steps

• Install delegates: Learn how to install delegates in different environments.

• Delegate metrics and auto-scaling: Monitor delegate health and configure auto-scaling.

• Secure delegates with tokens: Understand how to rotate delegate tokens for enhanced security.