Configuring for production
Configuring Red Hat Developer Hub for your IT ecosystem
Abstract
Preface
Learn how to configure Red Hat Developer Hub for production to work in your IT ecosystem.
Chapter 1. Configuring external PostgreSQL databases
As an administrator, you can configure and use external PostgreSQL databases in Red Hat Developer Hub. You can use a PostgreSQL certificate file to configure an external PostgreSQL instance using the Operator or Helm Chart.
Developer Hub supports the configuration of external PostgreSQL databases. You can perform maintenance activities, such as backing up your data or configuring high availability (HA) for the external PostgreSQL databases.
By default, the Red Hat Developer Hub operator or Helm Chart creates a local PostgreSQL database. However, this configuration is not suitable for the production environments. For production deployments, disable the creation of local database and configure Developer Hub to connect to an external PostgreSQL instance instead.
1.1. Configuring an external PostgreSQL instance using the Operator
You can configure an external PostgreSQL instance using the Red Hat Developer Hub Operator. By default, the Operator creates and manages a local instance of PostgreSQL in the same namespace where you have deployed the RHDH instance. However, you can change this default setting to configure an external PostgreSQL database server, for example, Amazon Web Services (AWS) Relational Database Service (RDS) or Azure database.
Prerequisites
- You are using a supported version of PostgreSQL. For more information, see the Product life cycle page.
You have the following details:
-
db-host
: Denotes your PostgreSQL instance Domain Name System (DNS) or IP address -
db-port
: Denotes your PostgreSQL instance port number, such as5432
-
username
: Denotes the user name to connect to your PostgreSQL instance -
password
: Denotes the password to connect to your PostgreSQL instance
-
- You have installed the Red Hat Developer Hub Operator.
- Optional: You have a CA certificate, Transport Layer Security (TLS) private key, and TLS certificate so that you can secure your database connection by using the TLS protocol. For more information, refer to your PostgreSQL vendor documentation.
By default, Developer Hub uses a database for each plugin and automatically creates it if none is found. You might need the Create Database
privilege in addition to PSQL Database
privileges for configuring an external PostgreSQL instance.
Procedure
Optional: Create a certificate secret to configure your PostgreSQL instance with a TLS connection:
cat <<EOF | oc -n <your-namespace> create -f - apiVersion: v1 kind: Secret metadata: name: <crt-secret> 1 type: Opaque stringData: postgres-ca.pem: |- -----BEGIN CERTIFICATE----- <ca-certificate-key> 2 postgres-key.key: |- -----BEGIN CERTIFICATE----- <tls-private-key> 3 postgres-crt.pem: |- -----BEGIN CERTIFICATE----- <tls-certificate-key> 4 # ... EOF
Create a credential secret to connect with the PostgreSQL instance:
cat <<EOF | oc -n <your-namespace> create -f - apiVersion: v1 kind: Secret metadata: name: <cred-secret> 1 type: Opaque stringData: 2 POSTGRES_PASSWORD: <password> POSTGRES_PORT: "<db-port>" POSTGRES_USER: <username> POSTGRES_HOST: <db-host> PGSSLMODE: <ssl-mode> # for TLS connection 3 NODE_EXTRA_CA_CERTS: <abs-path-to-pem-file> # for TLS connection, e.g. /opt/app-root/src/postgres-crt.pem 4 EOF
- 1
- Provide the name of the credential secret.
- 2
- Provide credential data to connect with your PostgreSQL instance.
- 3
- Optional: Provide the value based on the required Secure Sockets Layer (SSL) mode.
- 4
- Optional: Provide the value only if you need a TLS connection for your PostgreSQL instance.
Create a
Backstage
custom resource (CR):cat <<EOF | oc -n <your-namespace> create -f - apiVersion: rhdh.redhat.com/v1alpha1 kind: Backstage metadata: name: <backstage-instance-name> spec: database: enableLocalDb: false 1 application: extraFiles: mountPath: <path> # e g /opt/app-root/src secrets: - name: <crt-secret> 2 key: postgres-crt.pem, postgres-ca.pem, postgres-key.key # key name as in <crt-secret> Secret extraEnvs: secrets: - name: <cred-secret> 3 # ...
NoteThe environment variables listed in the
Backstage
CR work with the Operator default configuration. If you have changed the Operator default configuration, you must reconfigure theBackstage
CR accordingly.-
Apply the
Backstage
CR to the namespace where you have deployed the RHDH instance.
1.2. Configuring an external PostgreSQL instance using the Helm Chart
You can configure an external PostgreSQL instance by using the Helm Chart. By default, the Helm Chart creates and manages a local instance of PostgreSQL in the same namespace where you have deployed the RHDH instance. However, you can change this default setting to configure an external PostgreSQL database server, for example, Amazon Web Services (AWS) Relational Database Service (RDS) or Azure database.
Prerequisites
- You are using a supported version of PostgreSQL. For more information, see the Product life cycle page.
You have the following details:
-
db-host
: Denotes your PostgreSQL instance Domain Name System (DNS) or IP address -
db-port
: Denotes your PostgreSQL instance port number, such as5432
-
username
: Denotes the user name to connect to your PostgreSQL instance -
password
: Denotes the password to connect to your PostgreSQL instance
-
- You have installed the RHDH application by using the Helm Chart.
- Optional: You have a CA certificate, Transport Layer Security (TLS) private key, and TLS certificate so that you can secure your database connection by using the TLS protocol. For more information, refer to your PostgreSQL vendor documentation.
By default, Developer Hub uses a database for each plugin and automatically creates it if none is found. You might need the Create Database
privilege in addition to PSQL Database
privileges for configuring an external PostgreSQL instance.
Procedure
Optional: Create a certificate secret to configure your PostgreSQL instance with a TLS connection:
cat <<EOF | oc -n <your-namespace> create -f - apiVersion: v1 kind: Secret metadata: name: <crt-secret> 1 type: Opaque stringData: postgres-ca.pem: |- -----BEGIN CERTIFICATE----- <ca-certificate-key> 2 postgres-key.key: |- -----BEGIN CERTIFICATE----- <tls-private-key> 3 postgres-crt.pem: |- -----BEGIN CERTIFICATE----- <tls-certificate-key> 4 # ... EOF
Create a credential secret to connect with the PostgreSQL instance:
cat <<EOF | oc -n <your-namespace> create -f - apiVersion: v1 kind: Secret metadata: name: <cred-secret> 1 type: Opaque stringData: 2 POSTGRES_PASSWORD: <password> POSTGRES_PORT: "<db-port>" POSTGRES_USER: <username> POSTGRES_HOST: <db-host> PGSSLMODE: <ssl-mode> # for TLS connection 3 NODE_EXTRA_CA_CERTS: <abs-path-to-pem-file> # for TLS connection, e.g. /opt/app-root/src/postgres-crt.pem 4 EOF
- 1
- Provide the name of the credential secret.
- 2
- Provide credential data to connect with your PostgreSQL instance.
- 3
- Optional: Provide the value based on the required Secure Sockets Layer (SSL) mode.
- 4
- Optional: Provide the value only if you need a TLS connection for your PostgreSQL instance.
Configure your PostgreSQL instance in the Helm configuration file named
values.yaml
:# ... upstream: postgresql: enabled: false # disable PostgreSQL instance creation 1 auth: existingSecret: <cred-secret> # inject credentials secret to Backstage 2 backstage: appConfig: backend: database: connection: # configure Backstage DB connection parameters host: ${POSTGRES_HOST} port: ${POSTGRES_PORT} user: ${POSTGRES_USER} password: ${POSTGRES_PASSWORD} ssl: rejectUnauthorized: true, ca: $file: /opt/app-root/src/postgres-ca.pem key: $file: /opt/app-root/src/postgres-key.key cert: $file: /opt/app-root/src/postgres-crt.pem extraEnvVarsSecrets: - <cred-secret> # inject credentials secret to Backstage 3 extraEnvVars: - name: BACKEND_SECRET valueFrom: secretKeyRef: key: backend-secret name: '{{ include "janus-idp.backend-secret-name" $ }}' extraVolumeMounts: - mountPath: /opt/app-root/src/dynamic-plugins-root name: dynamic-plugins-root - mountPath: /opt/app-root/src/postgres-crt.pem name: postgres-crt # inject TLS certificate to Backstage cont. 4 subPath: postgres-crt.pem - mountPath: /opt/app-root/src/postgres-ca.pem name: postgres-ca # inject CA certificate to Backstage cont. 5 subPath: postgres-ca.pem - mountPath: /opt/app-root/src/postgres-key.key name: postgres-key # inject TLS private key to Backstage cont. 6 subPath: postgres-key.key extraVolumes: - ephemeral: volumeClaimTemplate: spec: accessModes: - ReadWriteOnce resources: requests: storage: 1Gi name: dynamic-plugins-root - configMap: defaultMode: 420 name: dynamic-plugins optional: true name: dynamic-plugins - name: dynamic-plugins-npmrc secret: defaultMode: 420 optional: true secretName: dynamic-plugins-npmrc - name: postgres-crt secret: secretName: <crt-secret> 7 # ...
- 1
- Set the value of the
upstream.postgresql.enabled
parameter tofalse
to disable creating local PostgreSQL instances. - 2
- Provide the name of the credential secret.
- 3
- Provide the name of the credential secret.
- 4
- Optional: Provide the name of the TLS certificate only for a TLS connection.
- 5
- Optional: Provide the name of the CA certificate only for a TLS connection.
- 6
- Optional: Provide the name of the TLS private key only if your TLS connection requires a private key.
- 7
- Provide the name of the certificate secret if you have configured a TLS connection.
Apply the configuration changes in your Helm configuration file named
values.yaml
:helm upgrade -n <your-namespace> <your-deploy-name> openshift-helm-charts/redhat-developer-hub -f values.yaml --version 1.4.0
1.3. Migrating local databases to an external database server using the Operator
By default, Red Hat Developer Hub hosts the data for each plugin in a PostgreSQL database. When you fetch the list of databases, you might see multiple databases based on the number of plugins configured in Developer Hub. You can migrate the data from an RHDH instance hosted on a local PostgreSQL server to an external PostgreSQL service, such as AWS RDS, Azure database, or Crunchy database. To migrate the data from each RHDH instance, you can use PostgreSQL utilities, such as pg_dump
with psql
or pgAdmin
.
The following procedure uses a database copy script to do a quick migration.
Prerequisites
Procedure
Configure port forwarding for the local PostgreSQL database pod by running the following command on a terminal:
oc port-forward -n <your-namespace> <pgsql-pod-name> <forward-to-port>:<forward-from-port>
Where:
-
The
<pgsql-pod-name>
variable denotes the name of a PostgreSQL pod with the formatbackstage-psql-<deployment-name>-<_index>
. -
The
<forward-to-port>
variable denotes the port of your choice to forward PostgreSQL data to. The
<forward-from-port>
variable denotes the local PostgreSQL instance port, such as5432
.Example: Configuring port forwarding
oc port-forward -n developer-hub backstage-psql-developer-hub-0 15432:5432
-
The
Make a copy of the following
db_copy.sh
script and edit the details based on your configuration:#!/bin/bash to_host=<db-service-host> 1 to_port=5432 2 to_user=postgres 3 from_host=127.0.0.1 4 from_port=15432 5 from_user=postgres 6 allDB=("backstage_plugin_app" "backstage_plugin_auth" "backstage_plugin_catalog" "backstage_plugin_permission" "backstage_plugin_scaffolder" "backstage_plugin_search") 7 for db in ${!allDB[@]}; do db=${allDB[$db]} echo Copying database: $db PGPASSWORD=$TO_PSW psql -h $to_host -p $to_port -U $to_user -c "create database $db;" pg_dump -h $from_host -p $from_port -U $from_user -d $db | PGPASSWORD=$TO_PSW psql -h $to_host -p $to_port -U $to_user -d $db done
- 1
- The destination host name, for example,
<db-instance-name>.rds.amazonaws.com
. - 2
- The destination port, such as
5432
. - 3
- The destination server username, for example,
postgres
. - 4
- The source host name, such as
127.0.0.1
. - 5
- The source port number, such as the
<forward-to-port>
variable. - 6
- The source server username, for example,
postgres
. - 7
- The name of databases to import in double quotes separated by spaces, for example,
("backstage_plugin_app" "backstage_plugin_auth" "backstage_plugin_catalog" "backstage_plugin_permission" "backstage_plugin_scaffolder" "backstage_plugin_search")
.
Create a destination database for copying the data:
/bin/bash TO_PSW=<destination-db-password> /path/to/db_copy.sh 1
- 1
- The
<destination-db-password>
variable denotes the password to connect to the destination database.
NoteYou can stop port forwarding when the copying of the data is complete. For more information about handling large databases and using the compression tools, see the Handling Large Databases section on the PostgreSQL website.
-
Reconfigure your
Backstage
custom resource (CR). For more information, see Configuring an external PostgreSQL instance using the Operator. Check that the following code is present at the end of your
Backstage
CR after reconfiguration:# ... spec: database: enableLocalDb: false application: # ... extraFiles: secrets: - name: <crt-secret> key: postgres-crt.pem # key name as in <crt-secret> Secret extraEnvs: secrets: - name: <cred-secret> # ...
NoteReconfiguring the
Backstage
CR deletes the correspondingStatefulSet
andPod
objects, but does not delete thePersistenceVolumeClaim
object. Use the following command to delete the localPersistenceVolumeClaim
object:oc -n developer-hub delete pvc <local-psql-pvc-name>
where, the
<local-psql-pvc-name>
variable is in thedata-<psql-pod-name>
format.- Apply the configuration changes.
Verification
Verify that your RHDH instance is running with the migrated data and does not contain the local PostgreSQL database by running the following command:
oc get pods -n <your-namespace>
Check the output for the following details:
-
The
backstage-developer-hub-xxx
pod is in running state. The
backstage-psql-developer-hub-0
pod is not available.You can also verify these details using the Topology view in the OpenShift Container Platform web console.
-
The
Chapter 2. Running Red Hat Developer Hub behind a corporate proxy
In a network restricted environment, configure Red Hat Developer Hub to use your proxy to access remote network resources.
You can run the RHDH application behind a corporate proxy by setting any of the following environment variables before starting the application:
-
HTTP_PROXY
: Denotes the proxy to use for HTTP requests. -
HTTPS_PROXY
: Denotes the proxy to use for HTTPS requests.
Additionally, you can set the NO_PROXY
environment variable to exclude certain domains from proxying. The variable value is a comma-separated list of hostnames that do not require a proxy to get reached, even if one is specified.
2.1. Configuring proxy information in Operator deployment
For Operator-based deployment, the approach you use for proxy configuration is based on your role:
- As a cluster administrator with access to the Operator namespace, you can configure the proxy variables in the Operator’s default ConfigMap file. This configuration applies the proxy settings to all the users of the Operator.
- As a developer, you can configure the proxy variables in a custom resource (CR) file. This configuration applies the proxy settings to the RHDH application created from that CR.
Prerequisites
- You have installed the Red Hat Developer Hub application.
Procedure
Perform one of the following steps based on your role:
As an administrator, set the proxy information in the Operator’s default ConfigMap file:
-
Search for a ConfigMap file named
backstage-default-config
in the default namespacerhdh-operator
and open it. -
Find the
deployment.yaml
key. Set the value of the
HTTP_PROXY
,HTTPS_PROXY
, andNO_PROXY
environment variables in theDeployment
spec as shown in the following example:Example: Setting proxy variables in a ConfigMap file
# Other fields omitted deployment.yaml: |- apiVersion: apps/v1 kind: Deployment spec: template: spec: # Other fields omitted initContainers: - name: install-dynamic-plugins # command omitted env: - name: NPM_CONFIG_USERCONFIG value: /opt/app-root/src/.npmrc.dynamic-plugins - name: HTTP_PROXY value: 'http://10.10.10.105:3128' - name: HTTPS_PROXY value: 'http://10.10.10.106:3128' - name: NO_PROXY value: 'localhost,example.org' # Other fields omitted containers: - name: backstage-backend # Other fields omitted env: - name: APP_CONFIG_backend_listen_port value: "7007" - name: HTTP_PROXY value: 'http://10.10.10.105:3128' - name: HTTPS_PROXY value: 'http://10.10.10.106:3128' - name: NO_PROXY value: 'localhost,example.org'
-
Search for a ConfigMap file named
As a developer, set the proxy information in your custom resource (CR) file as shown in the following example:
Example: Setting proxy variables in a CR file
spec: # Other fields omitted application: extraEnvs: envs: - name: HTTP_PROXY value: 'http://10.10.10.105:3128' - name: HTTPS_PROXY value: 'http://10.10.10.106:3128' - name: NO_PROXY value: 'localhost,example.org'
- Save the configuration changes.
2.2. Configuring proxy information in Helm deployment
For Helm-based deployment, either a developer or a cluster administrator with permissions to create resources in the cluster can configure the proxy variables in a values.yaml
Helm configuration file.
Prerequisites
- You have installed the Red Hat Developer Hub application.
Procedure
Set the proxy information in your Helm configuration file:
upstream: backstage: extraEnvVars: - name: HTTP_PROXY value: '<http_proxy_url>' - name: HTTPS_PROXY value: '<https_proxy_url>' - name: NO_PROXY value: '<no_proxy_settings>'
Where,
<http_proxy_url>
- Denotes a variable that you must replace with the HTTP proxy URL.
<https_proxy_url>
- Denotes a variable that you must replace with the HTTPS proxy URL.
<no_proxy_settings>
Denotes a variable that you must replace with comma-separated URLs, which you want to exclude from proxying, for example,
foo.com,baz.com
.Example: Setting proxy variables using Helm Chart
upstream: backstage: extraEnvVars: - name: HTTP_PROXY value: 'http://10.10.10.105:3128' - name: HTTPS_PROXY value: 'http://10.10.10.106:3128' - name: NO_PROXY value: 'localhost,example.org'
- Save the configuration changes.
Chapter 3. Configuring Red Hat Developer Hub deployment by using the operator
The Red Hat Developer Hub operator exposes a rhdh.redhat.com/v1alpha2
API Version of its Custom Resource Definition (CRD). This CRD exposes a generic spec.deployment.patch
field, which gives you full control over the Developer Hub Deployment resource. This field can be a fragment of the standard apps.Deployment
Kubernetes object.
Procedure
- Create a Developer Hub Custom Resource Definition with the following fields:
Example
apiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template:
labels
Add labels to the Developer Hub pod.
Example adding the label
my=true
apiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template: metadata: labels: my: true
volumes
Add an additional volume named
my-volume
and mount it under/my/path
in the Developer Hub application container.Example additional volume
apiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template: spec: containers: - name: backstage-backend volumeMounts: - mountPath: /my/path name: my-volume volumes: - ephemeral: volumeClaimTemplate: spec: storageClassName: "special" name: my-volume
Replace the default
dynamic-plugins-root
volume with a persistent volume claim (PVC) nameddynamic-plugins-root
. Note the$patch: replace
directive, otherwise a new volume will be added.Example
dynamic-plugins-root
volume replacementapiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template: spec: volumes: - $patch: replace name: dynamic-plugins-root persistentVolumeClaim: claimName: dynamic-plugins-root
cpu
requestSet the CPU request for the Developer Hub application container to 250m.
Example CPU request
apiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template: spec: containers: - name: backstage-backend resources: requests: cpu: 250m
my-sidecar
containerAdd a new
my-sidecar
sidecar container into the Developer Hub Pod.Example side car container
apiVersion: rhdh.redhat.com/v1alpha2 kind: Backstage metadata: name: developer-hub spec: deployment: patch: spec: template: spec: containers: - name: my-sidecar image: quay.io/my-org/my-sidecar:latest
Additional resources
- To learn more about merging, see Strategic Merge Patch.
Chapter 4. Configuring an RHDH instance with a TLS connection in Kubernetes
You can configure an RHDH instance with a Transport Layer Security (TLS) connection in a Kubernetes cluster, such as an Azure Red Hat OpenShift (ARO) cluster, any cluster from a supported cloud provider, or your own cluster with proper configuration. However, You must use a public Certificate Authority (CA)-signed certificate to configure your Kubernetes cluster.
Prerequisites
- You have set up an Azure Red Hat OpenShift (ARO) cluster with a public CA-signed certificate. For more information about obtaining CA certificates, refer to your vendor documentation.
You have created a namespace and setup a service account with proper read permissions on resources.
Example: Kubernetes manifest for role-based access control
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: backstage-read-only rules: - apiGroups: - '*' resources: - pods - configmaps - services - deployments - replicasets - horizontalpodautoscalers - ingresses - statefulsets - limitranges - resourcequotas - daemonsets verbs: - get - list - watch #...
- You have obtained the secret and the service CA certificate associated with your service account.
You have created some resources and added annotations to them so they can be discovered by the Kubernetes plugin. You can apply these Kubernetes annotations:
-
backstage.io/kubernetes-id
to label components -
backstage.io/kubernetes-namespace
to label namespaces
-
Procedure
Enable the Kubernetes plugins in the
dynamic-plugins-rhdh.yaml
file:kind: ConfigMap apiVersion: v1 metadata: name: dynamic-plugins-rhdh data: dynamic-plugins.yaml: | includes: - dynamic-plugins.default.yaml plugins: - package: ./dynamic-plugins/dist/backstage-plugin-kubernetes-backend-dynamic disabled: false 1 - package: ./dynamic-plugins/dist/backstage-plugin-kubernetes disabled: false 2 # ...
NoteThe
backstage-plugin-kubernetes
plugin is currently in Technology Preview. As an alternative, you can use the./dynamic-plugins/dist/backstage-plugin-topology-dynamic
plugin, which is Generally Available (GA).Set the kubernetes cluster details and configure the catalog sync options in the
app-config-rhdh.yaml
file:kind: ConfigMap apiVersion: v1 metadata: name: app-config-rhdh data: "app-config-rhdh.yaml": | # ... catalog: rules: - allow: [Component, System, API, Resource, Location] providers: kubernetes: openshift: cluster: openshift processor: namespaceOverride: default defaultOwner: guests schedule: frequency: seconds: 30 timeout: seconds: 5 kubernetes: serviceLocatorMethod: type: 'multiTenant' clusterLocatorMethods: - type: 'config' clusters: - url: <target-cluster-api-server-url> 1 name: openshift authProvider: 'serviceAccount' skipTLSVerify: false 2 skipMetricsLookup: true dashboardUrl: <target-cluster-console-url> 3 dashboardApp: openshift serviceAccountToken: ${K8S_SERVICE_ACCOUNT_TOKEN} 4 caData: ${K8S_CONFIG_CA_DATA} 5 # ...
- 1
- The base URL to the Kubernetes control plane. You can run the
kubectl cluster-info
command to get the base URL. - 2
- Set the value of this parameter to
false
to enable the verification of the TLS certificate. - 3
- Optional: The link to the Kubernetes dashboard managing the ARO cluster.
- 4
- Optional: Pass the service account token using a
K8S_SERVICE_ACCOUNT_TOKEN
environment variable that you can define in yoursecrets-rhdh
secret. - 5
- Pass the CA data using a
K8S_CONFIG_CA_DATA
environment variable that you can define in yoursecrets-rhdh
secret.
- Save the configuration changes.
Verification
Run the RHDH application to import your catalog:
kubectl -n rhdh-operator get pods -w
- Verify that the pod log shows no errors for your configuration.
- Go to Catalog and check the component page in the Developer Hub instance to verify the cluster connection and the presence of your created resources.
If you encounter connection errors, such as certificate issues or permissions, check the message box in the component page or view the logs of the pod.