Go Live guide

Welcome to the DX Cloud Go Live Checklist.

This guide is here to ensure a smooth and successful deployment of your project into production. Each point in this checklist addresses critical aspects that need to be reviewed and confirmed before the "Go Live" date.

From infrastructure readiness to automation, we walk you through the essential steps, providing clarity on where to assess each item and noting automation opportunities for streamlined processes. By following this checklist, you’ll be fully prepared to confidently deploy your project.

Standard project checklist

There are several items that need to checked before you go live. Completing this list does not necessarily mean going live will be entirely successful, but it helps minimize potential deployment issues.

The checklist is interactive, so you can use this page to tick each item off your list as you go if so desired.

Administration

  • You have established contacts for emergency site situations including primary technical contacts for your project and this has been communicated to the DX Cloud team.

  • The superuser credentials such as the secret and password have been updated for production including for both Public and Author instances.

Architecture

  • Your Magnolia pods and database pods for PostgreSQL are on the same node.

    For more on this topic, see Kubernetes fundamentals.

  • Your Magnolia Public instances are running in different availability zones. This helps to minimize any potential down time for your site.

    For more on this topic, see Multiregion clusters.

Database and memory

  • You use PostgreSQL 15 or later.

  • Database persistent volume for Magnolia is less than 50%. If there is an issue with persistent volume, see Database persistent volume is almost full.

  • You have at least 2 Magnolia public instances.

    For more on this topic, see Architecture: Deployment options.

  • You have a minimum 6gb JVM Max Heap available for public instances.

    Having less than 6gb JVM Max heap can cause issues like a container being OOMKilled.

  • You have at least 4gb available to public instances. (excluding Max Heap)

  • You have a minimum 8gb JVM Max Heap available for author instances.

    Having less than 8gb JVM Max heap can cause issues like a container being OOMKilled.

  • You have at least 4gb available to author instances. (excluding Max Heap)

  • Your Magnolia instances do not have a CPU limit set.

    For more on this topic, see CPU requests and limits on DX Cloud.

  • Your memory request and memory limit are the same.

    Why is this important?

    If you set a lower memory request than a memory limit, Kubernetes might run your Magnolia instance or Magnolia frontend instance on a cluster node that has enough memory to meet the memory request but not have enough memory to provide all memory of the memory limit.

    For Magnolia pods, you can set the memory request and limit used for the Magnolia author and public through the Helm chart values.

Helm

  • You do not use a custom helm chart.

    You must use the offical DX Cloud Helm chart.

  • You use a recent Helm chart version that is applicable to the version of Magnolia you use in our project.

    The latest Helm chart version is 1.15.2.

Monitoring

  • You have defined the JMX jar and JMX configuration in CATALINA_OPTS.

    For more on this topic, see JMX Monitoring.

Networking and infrastructure

  • You do not use JCR clustering.

  • You use the standalone redirects server.

    For more on this topic, see Networking: Configure redirect server.

  • If you are using the PaaS CDN, check with your DX Cloud contact to ensure we have set up the service. Be sure you are connected to Fastly.

  • Ensure the Web Application Firewall (WAF) is set up and ready. You can contact your DX Cloud representative to be sure of this.

  • Ingresses serving life traffic must have Fastly annotations (if using PaaS CDN).

  • You do not use snippets in your values.yml file.

    All ingress updates should go through the values.yml file as per the official Helm chart.

Additional headless checks

These checks are only if the project is a headless project or contains a headless element. They are in addition to the standard checks.

  • The service type for the frontend is clusterIp.

    A different service type can cause serious infrastructure issues.

  • You have at least 2 instances for your frontend.

Final checks

  • Your project contains only Magnolia, the redirect server, Solr (if applicable), frontend, and database deployments. No other helm deployments are allowed.

  • Check with your DX Cloud technical contact to ensure response time metrics are ready for launch.

Feedback

PaaS

×

Location

This widget lets you know where you are on the docs site.

You are currently perusing through the Magnolia PaaS docs.

Main doc sections

DX Core Headless PaaS Legacy Cloud Incubator modules