logor

Centralized error tracking and logs in one static binary, with its brain in xcon-db. Self-hosted Sentry is famously heavy — Kafka, ClickHouse, Redis, a dozen services; logor is one process next to one database.

The ingest side speaks the Sentry envelope protocol, so every language that has a Sentry SDK already has a logor client: point the SDK's DSN at your logor node and errors start arriving. Logs ride a plain batched JSON endpoint, and the house Go client (logor-clients/go) covers both without any SDK ceremony.

What you get out of the box:

• Projects with DSN keys — one per service, revocable from the console.
• Issues — errors grouped by fingerprint (SDK-sent, else exception type and top in-app frames, else message), with first/last seen, counts, and open / resolved / ignored states.
• Logs — recent view and a live tail over SSE.
• Alerts — mail on new issues, from the same node.
• Retention by age — old partitions are dropped whole; nothing slows down as history grows.

Start with the quick start, or read how ingest works.