ArgoCD Example

Introduction

This guide explains how to use the provided example to deploy a complete OceanBase environment using ArgoCD. The example follows the "App of Apps" pattern, where a single parent ArgoCD application (oceanbase-stack) manages the deployment of all the other child applications, ensuring they are created in the correct order using a combination of Sync Phases and Waves.

Prerequisites

Before you begin, ensure you have the following:

A running Kubernetes cluster.
ArgoCD installed on your cluster. You can find the installation instructions in the official ArgoCD documentation.
kubectl installed and configured to connect to your cluster.

Installation

The installation process involves applying a single manifest to your ArgoCD cluster, which will then deploy all the necessary components.

Step 1: Configure Custom Health Checks

For ArgoCD to accurately report the status of all resources, you must add custom health check scripts to the argocd-cm ConfigMap.

Create a file named argocd-health-patch.yaml with the following content:

data:
  resource.customizations.health.argoproj.io_Application: |
    hs = {}
    hs.status = "Progressing"
    hs.message = ""
    if obj.status ~= nil then
      if obj.status.health ~= nil then
        hs.status = obj.status.health.status
        if obj.status.health.message ~= nil then
          hs.message = obj.status.health.message
        end
      end
    end
    return hs
  resource.customizations.health.oceanbase.oceanbase.com_OBCluster: |
    hs = {}
    if obj.status == nil or obj.status.status == nil then
      hs.status = "Progressing"
      hs.message = "Waiting for status"
      return hs
    end

    local current_status = obj.status.status
    if current_status == "running" then
      hs.status = "Healthy"
      hs.message = "Cluster is running"
    elseif current_status == "failed" then
      hs.status = "Degraded"
      hs.message = "Cluster has failed"
    else
      hs.status = "Progressing"
      hs.message = "Cluster is " .. current_status
    end
    return hs
  resource.customizations.health.oceanbase.oceanbase.com_OBTenant: |
    hs = {}
    if obj.status == nil or obj.status.status == nil then
      hs.status = "Progressing"
      hs.message = "Waiting for status"
      return hs
    end

    local current_status = obj.status.status
    if current_status == "running" then
      hs.status = "Healthy"
      hs.message = "Tenant is running"
    elseif current_status == "failed" then
      hs.status = "Degraded"
      hs.message = "Tenant has failed"
    else
      hs.status = "Progressing"
      hs.message = "Tenant is " .. current_status
    end
    return hs

Apply the patch to your argocd-cm ConfigMap:

kubectl patch configmap argocd-cm -n argocd --patch-file argocd-health-patch.yaml

Step 2: Prepare the Repository

The example files are configured to be deployed from the upstream OceanBase operator repository. If you are deploying from your own fork of the repository, you must first update the repoURL in the application manifests.

The following files need to be updated to point to your Git repository:

example/argocd/oceanbase-stack.yaml
example/argocd/apps/secrets.yaml
example/argocd/apps/obcluster.yaml
example/argocd/apps/obtenant.yaml
example/argocd/apps/ob-configserver.yaml
example/argocd/apps/obproxy.yaml

In each file, change the repoURL to your forked repository's URL.

Step 3: Deploy the Application Stack

The oceanbase-stack application is the parent application that manages all other components. Deploy it to your ArgoCD instance using the following command:

kubectl apply -f example/argocd/oceanbase-stack.yaml -n argocd

Step 4: Monitor the Deployment

Once the oceanbase-stack application is created, you can monitor the deployment process in the ArgoCD UI. The deployment follows a specific order to ensure dependencies are met:

Pre-Sync Phase: The following applications are deployed first to set up the foundational components. ArgoCD will wait for these to complete and become Healthy before moving on.
- ob-operator: Installs the OceanBase operator and its CRDs.
- secrets: Creates the necessary password secrets.
- ob-configserver: Deploys the configuration server.
Wave 1: After the Pre-Sync phase is successful, this wave deploys the main cluster components.
- obcluster: Deploys the OceanBase cluster.
- obproxy: Deploys the proxy.
Wave 2: After Wave 1 is healthy, the final wave deploys the tenant.
- obtenant: Deploys the tenant.

Verification

After the sync process is complete, all applications in the ArgoCD UI should have a Synced and Healthy status.

You can also verify the deployment using kubectl:

Check the pods in the oceanbase and oceanbase-system namespaces to ensure all components are running.
Check the status of the OBCluster and OBTenant custom resources.

Accessing the Database

Once all the applications are Synced and Healthy, the oceanbase-stack provides a fully functional OceanBase tenant that is compatible with MySQL. You can connect to this database instance from other applications within your Kubernetes cluster.

Connection Details:

Host: obproxy.oceanbase.svc.cluster.local
Port: 2883
Username: root@obtenant
Password: The password is automatically generated and stored in a Kubernetes secret.

To retrieve the password, run the following command:

kubectl get secret tenant-root-password -n oceanbase -o jsonpath='{.data.password}' | base64 -d

You can use these details to configure any MySQL-compatible client or application to connect to your OceanBase tenant.

Troubleshooting

Sync Failures due to Network Issues: The ArgoCD server needs to be able to connect to the Git repository and the Helm repository. If you see Unknown sync statuses or errors like context deadline exceeded, it indicates a network problem between your ArgoCD instance and the internet. Ensure that firewalls, proxies, and network policies are configured to allow this traffic.

Introduction​

Prerequisites​

Installation​

Verification​

Accessing the Database​

Troubleshooting​