Refactor your applications

At each step of your cloud adoption process, New Relic plays a critical role in helping you establish baselines, prioritize migration steps, validate improvements, and manage cloud costs. New Relic can also help you refactor, or re-architect, your software and services when you migrate them to the cloud.

One key aspect of refactoring your software is splitting out your components into different services available from your cloud provider. For example, you might move your database component away from a traditional MySQL database to a cloud-hosted solution, like Amazon Relational Database (RDS) services such as DynamoDB or Aurora. If you move to the Azure cloud, you might adopt Azure SQL or Cosmos DB. When you move to solutions like those, you increase your database redundancy and resiliency. If you evaluate your overall ROI properly, you should start to see the cost/benefit ratio taking effect once you complete your refactoring.

1. Deploy monitoring

New Relic is a true multi-tenant SaaS solution, meaning New Relic designed the agents to work at every layer of your application stack. To get started, deploy the agents:

Cloud-based integrations

As you refactor your application and integrate new cloud services, you can also use New Relic to monitor and report data about your cloud services, offering you a comprehensive view of your entire architecture in one place. Cloud-based integrations available through New Relic include Amazon Web Services (AWS) and Microsoft Azure.

This tutorial describes a database migration and focuses on APM and Infrastructure.

2. Set deployment markers

With New Relic, you can track deployments by setting deployment markers. Deployment tracking lets you quickly see how deployments affect your application’s performance and gauge its performance before and after you refactor.

You can view deployments on the Deployments page in APM:

Deployments page > (selected app) > Events > Deployments: Here is an example of the Deployments page and details about a selected deployment.

3. Establish baselines

When you install the APM agent at the application layer, New Relic immediately starts collecting and displaying metrics. Use these metrics to set your pre-migration application baseline.

APM agents collect performance metrics about outbound calls to databases. In the Monitoring section of the APM UI, click Databases to see a view of your database metrics. The following shows a database running on MySQL that will be refactored to use DynamoDB after migrating to AWS:

Databases page in APM > (select an app) > Monitoring > Databases: Use this page to view and sort detailed information about database performance.

4. Refactor your applications

To take advantage of the cloud, research what services are available from your cloud provider, and then re-architect and uncouple your applications.

5. Create dashboards to track refactoring your applications

Apply your KPIs to your baseline measurements to validate cloud improvements, and, ideally, start seeing the impact of moving to the cloud. New Relic’s reporting for database interactions goes much deeper than high-level performance metrics about calls to the database layer in APM. You get access to metrics about slow queries so you can filter by database vendor and cache type. You can also see host and instance details for databases visually within Service Maps and within the context of transaction traces.

Because all of your performance metrics are available in New Relic Insights, it is easy to build dashboards to show the performance of your database calls before, during, and after your refactoring in a single unified view.

Here is an example of an Insights dashboard for an application that uses DynamoDB and was migrated to AWS:

Insights dashboard - DynamoDB Build a dashboard that shows the performance of your database calls.

6. Create comparison dashboards

To compare both the pre- and post-refactor baselines, create an Insights dashboard that displays the KPIs you baselined.

The following dashboard shows the MySQL request query latency as compared with all of the operations for DynamoDB. You can see that, for this example, the request query latency for DynamoDB is too small to even register on the graph. Performance is clearly optimized with this cloud-based database.

Insights dashboard - request query latency Use Insights dashboards to visualize pre- and post-refactor baselines.

With APM agents and the integrations New Relic provides, you can measure the effectiveness of refactoring your applications using a broad spectrum of metrics.

For more help

Explore the New Relic Platform.

Recommendations for learning more: