Do you mean you want recommendations for open‑source error tracking tools (self‑hosted), or help choosing/setting one up? Here’s a concise overview of the best open‑source options, what they’re good for, and tips for choosing and deploying one.
Top open‑source error‑tracking / crash-reporting projects
- Sentry (open core)
- Mature, feature‑rich, wide language/SDK support (JavaScript, Python, Java, Go, mobile, etc.).
- Issue grouping, releases, breadcrumbs, performance tracing (some enterprise features behind paid plan).
- Can be self‑hosted with Docker Compose/Kubernetes. Large community and docs.
- Good if you want production‑quality feature set and many integrations.
- GlitchTip
- Open‑source Sentry alternative built to be lighter and easier to self‑host.
- Compatible with Sentry SDKs and provides similar UI and features (error grouping, releases).
- Simpler to run than full Sentry for smaller teams.
- Errbit
- Ruby on Rails app inspired by Airbrake. Integrates with Airbrake protocol.
- Best if you prefer a simple Rails app and are handling Ruby apps.
- Less actively developed/featureful than Sentry.
Logging and aggregated observability (complementary tools)
- ELK / OpenSearch (Elasticsearch + Logstash + Kibana or OpenSearch)
- Powerful full‑text log search and dashboards. Good for error logs and forensic analysis.
- Resource‑heavy at scale.
- Grafana Loki + Promtail + Grafana
- Log aggregation designed to be cheaper than ELK, integrates well with Grafana dashboards.
- Pair with traces and metrics for full observability.
- Graylog
- Centralized log management with alerting and a UI, easier to operate than ELK for some teams.
- Fluentd / Fluent Bit / Vector
- Log collectors/forwarders to route app logs to Sentry/ELK/Loki/etc.
Choosing the right approach
- If you want a full error‑tracking experience (stack traces, grouping, releases, user context): use Sentry or GlitchTip.
- If you only need aggregated logs and search: use ELK/OpenSearch or Loki + Grafana.
- If you prefer light weight + Sentry API compatibility: GlitchTip.
- If you need something simple for Ruby apps: Errbit.
- Consider hosted Sentry if you don’t want to operate infrastructure.
Deployment & scaling notes
- Self‑hosting Sentry requires PostgreSQL, Redis, and Kafka/Redis for event pipeline in production; uses lots of storage for events.
- GlitchTip is simpler; still needs PostgreSQL and object storage (S3) for attachments.
- For logs, plan index/retention strategy (ELK/OpenSearch) or use Loki with retention rules to control storage costs.
- Use object storage (S3/MinIO) for attachment retention and backups.
Integrations & SDK support
- Sentry-compatible SDKs cover most frameworks and languages; GlitchTip is Sentry API compatible so you can reuse Sentry SDKs.
- Logging libraries (logback, winston, bunyan, python logging) can be configured to send to these services.
- For mobile/native crashes, Sentry has robust SDKs; alternatives may be limited.
Security & privacy
- Self‑hosting gives control over sensitive error data. Ensure TLS, access controls, and data retention policies.
- Redact sensitive fields (PII) before sending errors; many SDKs support before_send hooks.
Costs
- Self‑hosting costs: VM/container resources, storage, backups, maintenance time.
- Hosted Sentry / commercial providers may be cheaper operationally for small teams.
Want help?
- I can recommend the best option based on your stack, traffic, and whether you prefer self‑hosted or managed.
- I can also provide step‑by‑step setup for Sentry or GlitchTip (Docker compose or Kubernetes), or help integrate an SDK into your app. Which do you want?