Your web application just went from 200ms response time to 14 seconds. Users are bouncing. Revenue is bleeding. And you’re staring at server logs that tell you absolutely nothing useful.
This is exactly the kind of problem New Relic was built to solve. It’s a full-stack monitoring tool that tracks application performance, infrastructure health, browser behavior, and pretty much everything else your stack throws at it. Instead of guessing which microservice decided to misbehave at 2 AM, you get real-time data that points you straight to the bottleneck.
New Relic works as a software-as-a-service observability platform. You don’t host it yourself. You install lightweight agents in your apps or servers, and they ship telemetry data back to New Relic’s dashboard. It’s most famous for its application performance monitoring capabilities, but it does a lot more than that.
This guide covers how New Relic works under the hood, what its APM features actually do, and how to install and use it. No fluff, just the stuff you need to know.
What Is New Relic
New Relic is a web-based application monitoring tool used for full-stack observability. It lets you monitor applications, infrastructure, web browsers, and external services from a single platform. Think of it as a central nervous system for your entire stack.
The tool tracks your web application’s performance details in real time. Developers use that data to figure out why things are slow, what’s broken, and where the problem actually lives. Without something like this, you’re basically troubleshooting by intuition, and intuition scales about as well as a bash script held together with duct tape.
Among its services, New Relic is best known for application performance monitoring. As a SaaS product, everything runs through the browser. No on-premise servers to maintain, no infrastructure to babysit.
It supports multiple programming languages including PHP, Java, Python, and Ruby. You can integrate it with cloud providers like Amazon Web Services and Microsoft Azure, which makes it fit into most modern deployment pipelines without much friction.
How New Relic Works
To understand how New Relic works, you need to know about the four data types it uses for monitoring. They’re collectively called MELT:
Metrics are numeric measurements of your app’s performance. CPU usage, memory consumption, load time. The basics.
Events describe activities running within your application. Each event consists of multiple lines of log data and captures what happened at a specific moment.
Logs are detailed records of those events. Every step in an application gets recorded and timestamped. When something goes wrong at 3:47 AM, the logs tell you exactly what led up to it.
Traces show sequences of events. They give you a chronological view of how a request moved through your app, from the initial HTTP call to the database query to the response. This is where distributed tracing gets really valuable in microservice architectures.
New Relic collects all of this data using an agent. It’s a small piece of code you install within your web app or virtual private server. The agent instruments your application at the code level, gathering performance data without you needing to manually add logging everywhere.
Each monitoring task requires its own dedicated agent. There are separate agents for application monitoring, infrastructure monitoring, and browser monitoring. The agent collects data, sends it to New Relic’s servers, and the dashboard displays it. You drill into the data, find the bottleneck, and fix it.
New Relic APM Use Cases
New Relic APM is where most teams spend their time. It covers everything from code-level analysis to high-level performance overviews. Here’s what you can actually do with it.
Application Performance Monitoring
New Relic APM offers unified monitoring that tracks your web apps and microservices with code-level precision. For a quick health check, start with the APM Summary page. It shows the metrics that matter most:
Transaction response time measures how long a transaction takes from the requester’s perspective. It doesn’t reflect total server-side processing time, just what the user experiences.
Apdex score is the ratio of satisfactory to unsatisfactory response times. You set a threshold, and New Relic calculates whether your app meets it. An Apdex score below 0.5 means your users are having a bad time. Below 0.85 and you should probably start investigating.
Throughput tracks user activity measured in requests per minute. A sudden drop in throughput with no code changes? That’s usually a sign something upstream broke.
Error rate shows the percentage of transactions returning errors. New Relic flags unhandled exceptions automatically, so you don’t have to grep through logs to find them.
Host resources covers CPU and memory consumption on the host server. If your app’s slow and CPU is pegged at 100%, that’s a different fix than if CPU is at 5% but database queries are taking 8 seconds each.
Beyond the summary page, New Relic APM has dedicated views for distributed tracing, transaction breakdowns, Kubernetes deployments, database operations, and external service monitoring. Each one lets you drill deeper into specific parts of your stack.
SLA Reporting and Alert Configuration
New Relic APM includes service level agreement reporting that tracks uptime, downtime, and performance trends over time. You can pull reports daily, weekly, or monthly, and export them as CSV files. Handy when your manager wants proof that the app was actually up 99.97% of the time.
The alerting feature is where things get genuinely useful for operations. You define custom thresholds, and New Relic notifies you when metrics breach them. You’re not checking dashboards all day. The system comes to you.
You can customize alerts across several dimensions. Pick which data sources to monitor, define what constitutes a violation, set incident preferences for how notifications get grouped, and choose notification channels like email, Slack, OpsGenie, or PagerDuty.
The alerting engine also uses Applied Intelligence for root cause analysis. When an incident fires, it doesn’t just say “something’s wrong.” It tries to tell you what triggered it and what other systems might be affected.
How to Build Custom Dashboards
New Relic’s dashboard feature consolidates performance data into color-coded graphs on a single screen. It’s the kind of thing you throw up on a monitor in the office so everyone can see at a glance whether the world is on fire.
You can build personalized dashboards. Customize the layout, resize charts, change what data gets displayed. If you’re running services in multiple regions, create separate dashboards for each one. Share them with other teams so everyone’s looking at the same numbers.
The visual approach simplifies monitoring. Instead of scanning tables of numbers, you spot anomalies in graphs. A spike in response time shows up as an obvious line going the wrong direction. That’s faster than reading through rows of data trying to find the outlier.
Transaction Traces and Log Management
Transaction traces and log management are where you do the real detective work. Traces record your entire application’s transaction and database calls in detail. You can drill down to see exactly which service or query is dragging everything else down.
Logs give you the granular context that traces alone can’t provide. New Relic lets you see log messages tied directly to specific transaction traces. Search through them using the Logs UI with relevant keywords.
Here’s why this combination matters: a trace might show you that a particular database call took 6 seconds. The associated log tells you it was a table scan because someone forgot to add an index on a column that gets queried 10,000 times an hour. Trace finds the slow thing. Log tells you why it’s slow. Fix follows.
Key Benefits of New Relic
New Relic has specific advantages over many other monitoring tools:
Guided setup. The installation process isn’t a multi-day project. New Relic’s guided installation detects your environment automatically and configures the agent for you. Their documentation is solid, which isn’t something you can say about every monitoring tool out there.
Single-pane monitoring. Everything lives on one dashboard. You don’t need to jump between three different tools to correlate application errors with infrastructure metrics. It’s all in one place.
Custom error logging. When customer-facing errors happen, New Relic captures them with enough context to actually debug the issue. Different attributes, timestamps, stack traces. Not just a generic 500 error message.
Transaction tracking. You can see which application components take the longest to load. That’s how you find the one API call that’s adding 4 seconds to every page load because nobody noticed it was making three redundant database queries.
Clean interface. The UI is well-designed and customizable. Even people who aren’t deep in the ops world can navigate it without a training session. That matters when you need product managers or QA engineers to pull their own performance data.
How to Install New Relic APM Agent
Installing New Relic requires a language-specific APM agent. The process varies depending on your programming language and operating system. Here’s how to install the PHP APM agent on Ubuntu as a reference.
Before you start, make sure your system meets the minimum requirements and that you have root access.
Step 1. Register a New Relic account at newrelic.com/signup. Log in, go to your profile icon, then Account Settings, and copy your license key.
Step 2. Open your terminal and configure the New Relic package repository:
echo ‘deb http://apt.newrelic.com/debian/ newrelic non-free’ | sudo tee /etc/apt/sources.list.d/newrelic.list
Step 3. Add the GPG key to verify package authenticity:
wget -O- https://download.newrelic.com/548C16BF.gpg | sudo apt-key add –
Step 4. Update your local package list:
sudo apt-get update
Step 5. Install the PHP agent:
sudo apt-get install newrelic-php5
Step 6. If the installer throws a “Please run newrelic-install as root to complete installation” error, run it manually:
sudo newrelic-install install
Step 7. Enter your application name and the license key you copied earlier.
Step 8. Restart your web server. For Apache 2:
sudo service apache2 restart
After installation, generate some traffic on your app to verify the agent is working. Hit a few pages, trigger some API calls, and then check the New Relic dashboard.
If you’re not deploying at scale, the guided installation method is easier. It auto-detects your system and walks you through the process step by step.
How to Use New Relic for Troubleshooting
Once the agent is running, you can start using New Relic immediately. Log in, open the APM tab, and select your application. The dashboard shows your app’s performance data right away, no extra configuration needed.
Here’s a practical troubleshooting workflow:
Check the Apdex score graph. If it’s below your threshold, your application is underperforming. That’s your starting signal.
Look at the Web transactions time graph. Each color represents a different component: database queries, external service calls, code execution. Find the spike.
Click the spike. You’ll land on the Transactions page with more detail about what’s happening during those slow requests.
Sort by Apdex most dissatisfying. This surfaces the transactions hurting your score the most. Click the worst offender.
Drill into transaction traces. Scroll to the Transaction Traces section and click one. Open the Trace Details tab and look for segments highlighted in red or yellow.
Evaluate the segment. If a database call that should take 50ms is taking 3 seconds, you’ve found your problem. From here, it’s a code or database optimization task.
That entire workflow takes maybe 5 minutes once you’re familiar with the interface. Compare that to SSH-ing into servers and tailing log files for an hour. The time savings compound fast when you’re dealing with incidents regularly.
New Relic FAQ
Is New Relic a DevOps Tool
Yes. DevOps engineers rely on New Relic to monitor application performance after deploying changes. If a new release tanks response times or spikes error rates, the APM dashboard shows it within minutes. That feedback loop between deployment and monitoring is exactly what makes it a core DevOps tool.
How Does New Relic Collect Data
New Relic uses a lightweight agent installed within your application or on your host server. The agent instruments your code, collects performance metrics, events, logs, and traces, then sends that data to New Relic’s platform every 60 seconds. You don’t have to modify your application code to get started. The agent handles instrumentation automatically.
What Is the Difference Between Splunk and New Relic
Both are monitoring tools, but they solve different primary problems. New Relic is cloud-only and focuses on application performance monitoring and infrastructure observability. Splunk offers both cloud and on-premise deployment and leans heavily into log management and security analytics. If your main concern is APM, New Relic is the more focused choice. If you need SIEM capabilities or heavy log analysis with on-prem requirements, Splunk is the better fit.
Is New Relic Free to Use
New Relic offers a free tier that includes 100 GB of data ingestion per month and one full-platform user. It’s enough to monitor a small application or evaluate the platform before committing to a paid plan. Paid pricing scales based on data ingestion volume and number of users.
What Programming Languages Does New Relic Support
New Relic provides APM agents for PHP, Java, Python, Ruby, .NET, Node.js, and Go. Each agent has its own installation process and system requirements, but they all feed data into the same unified dashboard. Check New Relic’s compatibility docs for specific version support on your language and framework.
