Skip to main content
Version: 1.3.x

Management Plane Installation

This page will show you how to install the Tetrate Service Bridge management plane in a production environment.

Before you start:

✓ Make sure that you’ve checked the requirements
✓ You’ve configured all the external dependencies
✓ Downloaded Tetrate Service Bridge CLI (tctl)
✓ Synced the Tetrate Service Bridge images

TSB Management Plane

To keep installation simple but still allow a lot of custom configuration options we have created a management plane operator. The operator will run in the cluster and bootstraps the management plane as described in a ManagementPlane Custom Resource. It watches for changes and enacts them. To help in creating the right Custom Resource Document (CRD) we have added the ability to our tctl client to create the base manifests which you can then modify according to your required set-up. After this you can either apply the manifests directly to the appropriate clusters or use in your source control operated clusters.

Operators

If you would like to know more about the inner workings of Operators, and the Operator Pattern, review the Kubernetes documentation

required certificates

For the TSB management plane to communicate with TSB control planes, you need to set up TLS secrets. These secrets need to be created from the same trust chain and have the tls.crt, tls.key, and ca.crt fields set.

In the management plane namespace (default is tsb), this secret must be named xcp-central-cert. In the control plane namespaces (default is istio-system), this secret must be named xcp-edge-cert.

If you have installed cert manager in your TSB management plane cluster, we provide some convenience methods in tctl to help bootstrap these certificates.

Operator Installation

First, create the manifest allowing you to install the management plane operator from your private Docker registry:

tctl install manifest management-plane-operator \
--registry <registry-location> > managementplaneoperator.yaml

The managementplaneoperator.yaml file created by the install manifest command can be applied directly to the appropriate cluster by using the kubectl client:

kubectl apply -f managementplaneoperator.yaml

After applying the manifest you will see the operator running in the tsb namespace:

kubectl get pod -n tsb

Example output:

NAME                                            READY   STATUS    RESTARTS   AGE
tsb-operator-management-plane-d4c86f5c8-b2zb5 1/1 Running 0 8s

Management Plane Installation

The management plane components need some secrets for external communication purposes. The required secrets are split into five categories represented by the flag’s prefix: tsb, xcp, postgres, elastic and ldap.

These can be generated in the correct format by passing them as command-line flags to the management-plane manifest command.

The management plane communicates with a cluster control plane over mTLS. You will need to set up a TLS certificate and store it in a secret named xcp-central-cert in the management plane namespace including the standard tls.crt, tls.key and ca.crt fields. The CA certificate must be able to verify the certificate presented from the control plane that will be set up in a similar fashion when onboarding a cluster i.e. this xcp-central-cert will need to be created from the same chain of trust as the control plane certificates. For the CA, we recommend plugging into your existing PKI infrastructure. The leaf certificate must have a trust domain of xcp.tetrate.io and set in its URI SAN. e.g. spiffe://xcp.tetrate.io/central.

cert-manager

In case you have installed cert-manager in the management plane cluster, you can have tctl automatically install certificates for secure communication with control planes. To do this, add an --xcp-certs flag to the install manifest command listed below.

The below command represents the minimum required configuration for creating secrets. See the CLI reference documentation for all available options such as providing CA certificates for Elasticsearch, PostgreSQL and LDAP.

tctl install manifest management-plane-secrets \
--elastic-password <elastic-password> \
--elastic-username <elastic-username> \
--ldap-bind-dn <ldap-bind-dn> \
--ldap-bind-password <ldap-bind-password> \
--postgres-password <postgres-password> \
--postgres-username <postgres-username> \
--tsb-admin-password <tsb-admin-password> \
--tsb-server-certificate "$(cat foo.cert)" \
--tsb-server-key "$(cat foo.key)" > managementplane-secrets.yaml

You can check the bundled explanation from tctl by running this help command:

tctl install manifest management-plane-secrets --help

Once you've created your secrets manifest, apply it to your cluster.

Vault Injection

If you’re using Vault injection for certain components, remove the applicable secrets from the manifest that you’ve created before applying it to your cluster.

kubectl apply -f managementplane-secrets.yaml

Installation

Now we’re ready to deploy the management plane.

To deploy the management plane we need to create a ManagementPlane custom resource in the Kubernetes cluster that describes the management plane.

Below is a ManagementPlane custom resource that describes a basic management plane. Save this managementplane.yaml and adjust it according to your needs:

apiVersion: install.tetrate.io/v1alpha1
kind: ManagementPlane
metadata:
name: managementplane
namespace: tsb
spec:
hub: <registry-location>
dataStore:
postgres:
host: <postgres-hostname-or-ip>
port: <postgres-port>
name: <database-name>
telemetryStore:
elastic:
host: <elastic-hostname-or-ip>
port: <elastic-port>
version: <elastic-version>
identityProvider:
ldap:
host: <ldap-hostname-or-ip>
port: <ldap-port>
search:
baseDN: dc=tetrate,dc=io
iam:
matchDN: "cn=%s,ou=People,dc=tetrate,dc=io"
matchFilter: "(&(objectClass=person)(uid=%s))"
sync:
usersFilter: "(objectClass=person)"
groupsFilter: "(objectClass=groupOfUniqueNames)"
membershipAttribute: uniqueMember
tokenIssuer:
jwt:
expiration: 1h
issuers:
- name: https://jwt.tetrate.io
algorithm: RS256
signingKey: tls.key

For more information on what each of these sections describes and how to configure them, please check out the following links:

Edit the relevant sections, save your configured custom resource to a file and apply it to your Kubernetes cluster.

kubectl apply -f managementplane.yaml

Once applied, ensure that TSB has created a default tenant and onboarded your identity provider.

kubectl create job -n tsb teamsync-bootstrap --from=cronjob/teamsync

Note: TSB will automatically do this every hour, so this command only needs to be run once after the initial installation.

Verifying Installation

To verify your installation succeeded, log in as the admin user. Try to connect to the TSB UI or login with the tctl CLI tool.

The TSB UI is reachable on port 8443 of the external IP as returned by the following command:

kubectl get svc -n tsb envoy

To configure tctl’s default config profile to point to your new TSB cluster do the following:

tctl config clusters set default --bridge-address $(kubectl get svc -n tsb envoy --output jsonpath='{.status.loadBalancer.ingress[0].ip}'):8443

Now you can log in with tctl and provide the tenant (which will default to tetrate) and admin account credentials.

tctl login
Tenant: tetrate
Username: admin
Password: *****
Login Successful!