Heroku OpEx
Heroku’s operational experience lets you focus on what’s most important – maintaining application health and providing an optimal experience for your end users. Sit back and let the platform monitor the key indicators of app health like responsiveness and request failures, alert you proactively so you can find (and fix) issues before your users do, and effortlessly scale to meet increases in demand. We run one of the world’s largest PaaS platforms hosting millions of apps, so we understand the value of operations and know your time is valuable. That’s why we provide integrated tools to increase application visibility, diagnose issues, receive proactive notifications and remediate issues to help developers and operations staff maintain apps in a healthy state. The platform manages all downstack components freeing you to focus on application operations, not infrastructure.
Application Metrics
Application metrics help you identify, investigate, and diagnose issues. Metrics provides key attributes of app health including response time, throughput, errors, dyno load, and memory, unified on a single time axis so you can see correlations. Accessible via the Heroku Dashboard, Application Metrics are available on all paid dynos.
Threshold Alerting
Threshold Alerting lets you specify limits on web dyno 95th percentile response time and the percentage of failed requests, above which an alert will be triggered. Email, PagerDuty, and dashboard notifications are supported. Threshold Alerting is available to apps running on Professional or Private dynos.
Autoscaling
Autoscaling horizontally scales your app’s web dynos to meet your specified 95th percentile response time threshold, based on your existing throughput. With autoscaling there’s no need to anticipate your traffic spikes. Autoscaling is included for free on Performance and Private dynos.
Consolidated Logs
Tail and view the logs for your application within the Dashboard or via the command line interface (CLI) to gain more in-depth details about events and errors. Drain your logs using one of the many logging add-on providers in the Elements marketplace for long-term storage, search, filtering and troubleshooting.
App webhooks
App webhooks allow you to subscribe to events and receive notifications when changes are made to your Heroku app. Webhooks make it easy to integrate changes like dyno formation changes or app releases into your operational workflow.
Heroku Exec
Heroku Exec allows you to connect to a dyno at runtime via SSH to make debugging easier. With Exec you can copy files off of a dyno, attach a remote debugger via local port forwarding, and take advantage of common Java debugging tools.
Runtime Metrics
Language Runtime Metrics display key language-specific performance metrics, like heap and non-heap memory and garbage collection activity, to gain insights into the root cause of application performance problems.
Discover Heroku Opex
Visibility Into App Health
With up to 7-day metrics retention, Application Metrics helps you view trends and quickly identify anomalies in application health. Language Runtime Metrics and Exec provide further insights to help pinpoint the root cause of performance bottlenecks.
Proactive Notifications
We watch your app, so you don’t have to. Heroku sends notifications if the response time or error rate for your web dynos exceeds your criteria. Webhook notifications make it easy to integrate changes to your apps into a custom workflow.
Effortless Scaling
No need to keep an eye on your traffic spikes anymore, we automatically adjust the number of web dynos to meet your specified latency threshold.
Performance Tuning Guidance
With features like language-specific Application Guidance, we help you identify and correct application problems.
No Additional Cost
Metrics are available for Hobby dynos and above at no additional cost. Advanced features are unlocked as you move to higher level dynos.
Easy Debugging
With Language Runtime Metrics you can easily diagnose when your app is leaking memory. Then use Heroku Exec to find the root cause by connecting directly to a dyno at runtime.
“Heroku is our ‘ops team.’ Compared to hiring DevOps and owning all the infrastructure responsibility, Heroku just makes it so much easier.”
Podcast A streamlined approach to application operations, and how Heroku built autoscaling
Listen to our podcast with Software Engineering Daily from February 28th, 2017.
In this podcast Andrew Gwozdziewycz, Heroku operational experience engineer, talks about the importance of application health, Heroku’s metrics platform architecture and how it laid the foundation for autoscaling.
Ready to Get Started?
Stay focused on building great data-driven applications and let Heroku tackle the rest.