Error Tracking
Definition
Tools and practices for automatically detecting, recording, and alerting developers about errors and crashes in applications.
Use Cases
- Google: Detecting and prioritizing production exceptions across services — Teams use Google Cloud Error Reporting to automatically group similar exceptions, view stack traces, and set alerting policies so on-call engineers are notified when new error types appear or spike. (Faster identification of regressions after releases and reduced mean time to resolution (MTTR) by focusing engineers on the highest-impact error groups.)
- Microsoft: Monitoring web app reliability and diagnosing crashes in production — Engineering teams use Azure Monitor Application Insights to collect exceptions, correlate them with requests/dependencies, and investigate failures using distributed tracing and end-to-end transaction views. (Improved incident response by correlating errors to specific deployments, endpoints, and upstream dependencies, enabling quicker root-cause analysis.)
- GitHub: Alerting on application errors and debugging issues in production — Teams use Sentry to capture unhandled exceptions, group them by fingerprint, attach stack traces and breadcrumbs, and route alerts to the right owners via integrations (e.g., issue trackers and chat/incident tools). (Reduced time spent reproducing bugs and improved release confidence by catching high-severity errors quickly after deployment.)
Provider Equivalents
- AWS: Amazon CloudWatch Application Signals
- Azure: Azure Monitor Application Insights
- GCP: Google Cloud Error Reporting
- OCI: OCI Application Performance Monitoring (APM)
Frequently Asked Questions
- What's the difference between error tracking and application performance monitoring (APM)?
- Error tracking focuses on capturing crashes and exceptions (what broke, where, and how often). APM is broader: it includes performance metrics (latency, throughput), traces, dependency calls, and often error tracking too. In practice, many APM tools include error tracking, but dedicated error trackers go deeper on grouping, alerting, and developer-friendly debugging details.
- When should I use error tracking?
- Use it for any production application where you need to know quickly when users hit crashes or unhandled exceptions. It’s especially useful after frequent deployments, when multiple services interact (microservices), or when you support mobile/desktop clients where crashes are hard to reproduce. If you already have logs/metrics, add error tracking when you want automatic grouping, stack traces, and actionable alerts instead of manually searching logs.
- How much does error tracking cost?
- Cost usually depends on event volume (number of errors/exceptions captured), retention period, number of users/seats, and advanced features (alerting, integrations, performance tracing, session replay). Cloud-provider options may charge based on ingested telemetry and alerting usage, while SaaS tools like Sentry or Bugsnag typically price by monthly events and team size. To control cost, sample low-value errors, filter noisy exceptions, and set retention appropriate to your debugging needs.
Category: software
Difficulty: intermediate
Related Terms
See Also