Self-hosted error monitoring

Run error monitoring and bug triage in infrastructure your team controls.

Logister is an open source, self-hosted error monitoring and bug triage app for teams that want a forkable alternative to Bugsnag, Sentry, and Bugzilla-style workflows.

Best fit

Choose this path when data ownership and operational control matter.

Logister is built for teams that want application errors, logs, metrics, transactions, spans, check-ins, ownership, and notification preferences in a self-hosted app. The app runs as a Rails web process plus a Sidekiq worker, backed by PostgreSQL and Redis. S3-compatible archive storage can be added when you need export-before-prune retention, and ClickHouse can be added later for higher-volume analytics.

After the baseline runs, use the focused guides for notifications and data retention rather than trying to configure every option from one page.

Infrastructure

The supported self-hosting stack is explicit.

PartRole
Rails web processServes the UI, auth, project pages, public pages, and ingest endpoints.
PostgreSQLStores accounts, projects, API keys, events, error groups, monitors, notification preferences, and project sharing.
Redis and SidekiqRun background jobs, caching, email delivery, project notifications, digest scheduling, optional ClickHouse writes, and archive/prune tasks.
SMTP / Amazon SESSends auth mail, project alerts, digests, monitor notices, workflow notifications, and operational notices.
S3-compatible object storageOptionally stores compressed JSONL telemetry archive exports before old non-error hot telemetry is pruned from PostgreSQL.
ClickHouseOptionally stores raw event/span analytics copies and one-minute rollups after schema readiness checks pass.
Registry imagesProvide versioned release images such as ghcr.io/taimoorq/logister:v2.8 or docker.io/taimoorq/logister:v2.8, with optional Quay mirrors when configured.

Why it helps

Self-hosting does not have to mean inventing the product surface yourself.

  • Grouped errors become assignable work with status, occurrence history, and request context.
  • Dashboard explorer charts use server-backed aggregate endpoints instead of loading every event into the browser.
  • Archived projects stay accessible but disappear from active dashboards and revoke active API keys.
  • Release notes, Docker images, docs, and environment variable references are published together for repeatable operations.

Evaluate

Prove the baseline before enabling optional services.

  1. Deploy the web process, worker process, PostgreSQL, Redis, and SMTP with secrets managed outside the repository.
  2. Run migrations and confirm the health endpoint responds.
  3. Create one project, invite one teammate if team triage matters, and generate a project API key.
  4. Send one test event from the runtime your team uses most.
  5. Confirm inbox grouping, detail context, assignment, status changes, and notification delivery before adding S3 archives or ClickHouse.

Next steps

Start with the deployment docs, then connect one service.