Anomaly Detection

Set smart alerts on misbehaving metrics.

Anomaly detection is the best way to avoid surprises. Get alerted in real-time when metrics go over their limits. Up to date > too late.

Docs
Metric alerts with anomaly detection screenshot of AppSignal APM

Catch issues early

Be alerted when performance metrics deviate from the norm. Monitor all of your critical metrics and keep your app stable.

Filter out the noise

AppSignal filters out temporary metric spikes, alerting you only to real issues. Avoid alert fatigue and maintain focus.

Detect anomalies anywhere

Combine anomaly detection with custom metrics to monitor any part of your app. Receive tailored alerts.

Getting Started

Enable one of many available triggers.

AppSignal collects loads of performance metrics out of the box. Enable one of the default alerts on metrics that are important to you. Now you can run Ruby, Elixir, Node.js and Python apps with confidence.

Error metrics

  • Error rate
  • Error rate per namespace
  • Error count
  • Error count per namespace

Performance metrics

  • Throughput
  • Throughput per namespace
  • Action/controller performance
  • Queue time
  • Queue time per namespace

Custom metrics

  • Host metrics
  • CPU usage
  • Disk I/O and usage
  • Load average
  • Memory usage
  • Swap usage
  • Network traffic

Custom metrics

With our custom metrics feature, you can monitor any metric you like. When you combine this with anomaly detection, you’ll get alerts for the metrics that are most important for your application’s health.

Alert Configuration

Avoid noisy alerts with warmups and cooldowns.

Stop drowning in noisy notifications, hello inbox zero. AppSignal’s configuration options make sure you receive the right alerts at the right time.

  1. Warmups before alerts

    Configure how soon you get an alert after a metric goes out of its limits. Avoid alerts for short spikes in metrics.

  2. Ongoing alert

  3. Cooldown after incidents

    Prevent multiple alerts in a row by setting a cooldown. We’ll create one longer incident instead of multiple short ones.

  4. Alert ended

What our customers think

1500+ development teams rely on AppSignal to ship code with confidence. We are sharing a few of their stories with you.

Previously, we were using Honeybadger, Skylight, and Datadog, but now we have all those apps in one: AppSignal, and that is awesome!

Oleg, Developer at Benuevo

Hey y’all, we’re back. Sentry was just chaos.

Matt, Developer at Supertape

The biggest benefit of using AppSignal is the native Elixir support. … AppSignal just knows how an Elixir app is structured and builds error tracking and performance monitoring around it.

Justin Tormey, CTO at Ampla Technologies

Previously, we were using Honeybadger, Skylight, and Datadog, but now we have all those apps in one: AppSignal, and that is awesome!

Oleg, Developer at Benuevo

Hey y’all, we’re back. Sentry was just chaos.

Matt, Developer at Supertape

The biggest benefit of using AppSignal is the native Elixir support. … AppSignal just knows how an Elixir app is structured and builds error tracking and performance monitoring around it.

Justin Tormey, CTO at Ampla Technologies

We looked at Sentry, Datadog, and Honeybadger. AppSignal gave us everything in one tool for a fraction of the price.

Thilo Rusche, Developer at SportsKey

In New Relic, if you want more features, you have to upgrade… AppSignal’s pricing was easier to understand, and you get all the features for one price.

Devan Sabaratnam, Co-founder & CTO at HR Partner

Second to none support. It’s their developers themselves who are quickly and clearly answering any questions. It’s hard to overstate how nice this is.

Verified G2 Reviewer

We looked at Sentry, Datadog, and Honeybadger. AppSignal gave us everything in one tool for a fraction of the price.

Thilo Rusche, Developer at SportsKey

In New Relic, if you want more features, you have to upgrade… AppSignal’s pricing was easier to understand, and you get all the features for one price.

Devan Sabaratnam, Co-founder & CTO at HR Partner

Second to none support. It’s their developers themselves who are quickly and clearly answering any questions. It’s hard to overstate how nice this is.

Verified G2 Reviewer

Link Dashboards

Link to related dashboards.

Link your alerts to specific dashboards to inspect related metrics and debug issues faster.

Linking an AppSignal alert to a related dashboard

Add Notes

Add debugging notes.

Add a few lines of debugging information for when an alert goes off. Your future self and your colleagues will thank you later.

Debugging notes attached to an AppSignal alert

Integrations

Get metric alerts in your tool of choice.

Stop drowning in noisy notifications, hello inbox zero. AppSignal’s configuration options make sure you receive the right alerts at the right time.

AppSignal metric alerts delivered to notification integrations

Getting started

Bye bye surprises. Hello metric alerts.

Set metrics on anything

Configure unlimited metric alerts on any application metric you can track.

Included in all plans

We keep our plans simple and affordable. That’s why triggers are available in all plan sizes, for free.

For Ruby, Elixir, Node.js & Python

Metric alerts work out of the box for all supported languages and frameworks.

Dev-to-dev support

Need help setting up anomaly detection? Our developer support team is happy to help you.

$appsignal install

Make your next crash make sense.

Free for 30 days. No credit card. Two-minute install.