Prerequisites

Before proceeding, ensure you have the following prerequisites in place:

1. Create two kubernetes node groups (for application & analysis workloads) and label the nodes in the node groups (eg: workload: deepsource-application & workload: deepsource-analysis)
   1. The minimum specifications for the nodes are as follows:
      1. Application Node group
         1. CPU: 16 Cores
         2. Memory:32 GiB
         3. Disk:128 GiB
      2. Analysis Node group
         1. CPU:16 Cores
         2. Memory:32 GiB
         3. Disk:256 GiB
2. Create two namespaces
   1. deepsource-application - kubectl create ns deepsource-application
   2. deepsource-analysis - kubectl create ns deepsource-analysis
3. Create an external database using the following guide: https://docs.deepsource.com/docs/enterprise-server-configuration#bring-your-own-database. You can skip this step for pilot installation.
4. Obtain a license by reaching out to [email protected]
5. Make sure to use at least helm v3.8.0, prior versions require HELM_EXPERIMENTAL_OCI=1environment variable to be set.

Install

1. Login to the helm registry using the obtained credentials
   Shell

   helm registry login registry.deepsource.com --username <username> -—password <password>
   

2. Obtain the default values.yaml file from the chart and write to a file
   Shell

   helm show values oci://registry.deepsource.com/helm-charts/deepsource > values.yaml
   

3. Make the necessary changes to the newly created values.yaml file by following the configuration guide.
4. Install DeepSource Enterprise Server
   Shell

   helm install deepsource --set-file global.license.token=<path-to-license-file> -n deepsource-application --timeout 10m0s --debug oci://registry.deepsource.com/helm-charts/deepsource -f values.yaml
   

Upgrade

1. Login to the registry using the obtained credential
   Shell

   helm registry login registry.deepsource.com --username <username> —password <password>
   

2. Upgrade DeepSource Enterprise Server using your existing values.yaml

   Go through the changelog.md in the helm package to make sure there aren't any breaking changes.

   Shell

   helm upgrade deepsource --set-file global.license.token=./license.txt -n deepsource-application --timeout 10m0s --debug oci://registry.deepsource.com/helm-charts/deepsource --version <version-number> -f values.yaml
   

Configuration

• All the values with changeme as the default value must be changed and are considered a security risk if left unchanged.
• Wherever applicable, If you have a value that contains newlines (for example, a private key), use a multi-line literal style yaml string (starts with |).
• The Persistent Volumes (PVs) don't get cleaned up automatically when the chart is uninstalled. You need to manually delete the PVs if you want to clean up the data.

Global configuration creation parameters

Cluster Settings configuration parameters

Analysis node group configuration parameters

Application node group configuration parameters

Version Control System (VCS) configuration parameters

Github configuration parameters

Github Enterprise configuration parameters

Gitlab configuration parameters

Bitbucket configuration parameters

Bitbucket Data Center configuration parameters

AzureDevOps configuration parameters

GSR configuration parameters

Backend Services/Dependencies configuration parameters

RabbitMQ configuration parameters

Redis configuration parameters

PostgreSQL configuration parameters

• When using embedded PostgreSQL, the values once set are not changed. If you want to change the values, you need to delete the PostgreSQL PVCs & PVs and then change the values.
• To configure your PostgreSQL database, follow the guide.

Minio/S3 Compatible Storage configuration parameters

Auth configuration parameters

SAML configuration parameters

SCIM configuration parameters

Networking configuration parameters

Ingress configuration parameters

Controller configuration parameters

TLS configuration parameters

CustomCA configuration parameters

Integrations configuration parameters

Slack integration configuration parameters

Jira integration configuration parameters

Deployment Settings configuration parameters

External Registry configuration parameters

Currently we support Sonatype Nexus as an external registry.

Reporting configuration parameters

License configuration parameters

Troubleshooting

Generate support-bundle

1. Make sure yq is installed on your system. To Install. yq follow the installation guide: https://mikefarah.gitbook.io/yq/v/v3.x/
2. We use kubectl support-bundle plugin from troubleshoot.sh. To install support-bundle follow this guide:
   1. Install krew for your system from https://krew.sigs.k8s.io/docs/user-guide/setup/install/
   2. Once krew is installed and working run the following command
      Shell

      kubectl krew install support-bundle
      

3. Generate the support-bundle by running the following command:
   cURL

   curl https://assets.enterprise.deepsource.com/generate-support-bundle | sh -s <path-to-values-file>
   

Admin permission issues during initial setup

If you are having permission issues despite your email being in the global.auth.admins block in values file, run the following command after making sure all the DeepSource application pods are up and running.

APP_NAMESPACE="<deepsource-application-namespace>"; kubectl exec $(kubectl get pods -l app.kubernetes.io/name=asgard-main -n $APP_NAMESPACE -o jsonpath="{.items[0].metadata.name}") -n $APP_NAMESPACE -- python manage.py update_superusers

Manually running the migration

During installation or upgrade, the migrate job that runs post-install might fail due to various reasons after the backoff-limit is reached. To manually trigger the migration process run the following command

APP_NAMESPACE="<deepsource-application-namespace>"; kubectl exec $(kubectl get pods -l app.kubernetes.io/name=asgard-celery-serveanlytics -n $APP_NAMESPACE -o jsonpath="{.items[0].metadata.name}") -n $APP_NAMESPACE -- make bucket compress migrate