Two weeks ago at FutureStack16 in San Francisco, we announced the general availability of New Relic Infrastructure, which includes simple configuration and dynamic dashboards and alerting for 12 of the most used Amazon Web Services (AWS) products out-of-the-box as part of our Professional version. But we’re just getting started.

This week we’re proud to announce another integration, offering support for AWS Lambda, the leading serverless compute platform. For New Relic Infrastructure customers, this means setup and configuration in seconds; dynamic, filterable dashboards and alerting; event correlation; live inventory; and 13 months of data retention. You can find out more about how seamlessly New Relic Infrastructure works with AWS services, including setup, on Adam Larson’s new blog post (New Relic Infrastructure in the Real World: AWS), so I’d like to focus here on how we can help you get the most of your Lambda investment.

Lambda enables engineers to run applications and functions without having to configure, maintain, and scale the underlying infrastructure. It has dramatically changed the way a number of our customers deploy their applications. Instead of maintaining hardware or virtual machines that sit idle when not being used, Lambda functions execute only when prompted by an external event or on a regular interval, such as data being added to a DynamoDB table or on a regular basis by polling a Kinesis stream, ultimately saving time and money for operations and development teams.

The unique, out-of-the-box support that New Relic offers Lambda customers includes:

  • Configuration changes and live state inventory: Just like all the data we process from hosts and the other 12 AWS integrations, we enable our customers to gather, store, and analyze the configuration of their Lambda functions and applications and the associated aliases mappings, highlight any changes in the event feed, and keep a live state of the currently deployed system, enabling operations and engineering teams to correlate any changes to the Lambda configurations with other potential problems throughout a system.
  • Extensibility: Everything in New Relic Infrastructure is processed inside of New Relic Insights, which delivers real-time operational dashboards, and it’s easy to extend the data model by adding custom events using the New Relic Insights API.
// Write two custom metrics to Insights from the Lambda function
// Full code here: https://github.com/smithclay/calming-metrics-lambda
var memorySample = process.memoryUsage();
request.write(JSON.stringify([
    {
      eventType: 'CustomServerlessTiming',
      functionName: process.env.AWS_LAMBDA_FUNCTION_NAME,
      segmentName: 'insightsQuery',
      duration: duration
    },
    {
      eventType: 'CustomServerlessMemoryUsage',
      functionName: process.env.AWS_LAMBDA_FUNCTION_NAME,
      rss: memorySample.rss,
      heapTotal: memorySample.heapTotal,
      heapUsed: memorySample.heapUsed
    }
    ]
    ));
request.end();

This means that not only does the integration gather the simple metrics for invocations, errors, throttling, and duration, but engineers can extend the integration to gather custom instrumentation beyond aggregate metrics to track individual requests across tens, hundreds, or even thousands of functions. All of these metrics can then be sliced, diced, and analyzed inside of New Relic Insights using any of the individual AWS Lambda transaction attributes, including such things as request latency and memory usage.

  • Ease of configuration: After going through the initial AWS setup process that’s a part of the base New Relic Infrastructure installation, customers can enable the Lambda integration in seconds by editing the Lambda account or accounts that they want to monitor, check a single box, and within seconds curated dashboards are created with real-time data available for analysis.

What could this look like for you? The two images below show a dashboard for a sample Lambda environment and then an explainer of what’s in that sample dashboard. In the real world, of course, a team monitoring a Lambda app would likely combine these Lambda metrics with analytic dashboards for other non-Lambda services in New Relic Insights dashboards, but today we’ll focus on Lambda as a stand-alone entity.

AWS lambda screenshot

This dashboard displays built-in Lambda metrics with custom instrumented metrics sent to Insights for greater visibility into the performance of the function.

aws lambda screenshot

Detailed information about inventory, configuration state, built-in, and custom metrics are available in real time.

 

We hope that you’ll start a 30-day free trial of New Relic Infrastructure today or, if you already have New Relic Infrastructure Professional edition, that you’ll give this new integration a try. The full documentation for our Lambda support is available here.

If you’re attending AWS re:Invent this week, stop in to see a demo at our booth (#610) and meet the New Relic Infrastructure team. And don’t miss my colleague Lee Atchison’s session on Wednesday, November 30, at 2:00 p.m.: ARC303 – Cloud Monitoring – Understanding, Preparing, and Troubleshooting Dynamic Apps on AWS.

We’d love to hear your feedback on Twitter @NewRelic or in the New Relic Discussion Forums!

 

Aaron Johnson is VP of product management, looking after Alerts, Insights, and New Relic Infrastructure. View posts by .

Interested in writing for New Relic Blog? Send us a pitch!