finds.dev← search

// the find

timescale/promscale

★ 1,315 · Go · Apache-2.0 · updated Mar 2024

[DEPRECATED] Promscale is a unified metric and trace observability backend for Prometheus, Jaeger and OpenTelemetry built on PostgreSQL and TimescaleDB.

Promscale was a Prometheus remote storage and Jaeger trace backend built on top of PostgreSQL and TimescaleDB. The pitch was one familiar database instead of Elasticsearch or Cassandra, with SQL available for ad-hoc trace analysis. It is officially discontinued as of late 2023.

- SQL-queryable traces is a genuinely underrated idea — being able to JOIN span data with business tables in Postgres is more useful than most Jaeger UIs allow, and the query examples in the docs show this off well

- Two-component architecture (connector + database) is much leaner than the Elasticsearch or Cassandra alternatives that Jaeger users typically had to operate

- prom-migrator, the included Prometheus-to-Prometheus data migration tool, is a standalone utility that may still be useful even though the main project is dead

- 100% PromQL compliance was independently verified by promlabs, not just claimed in the README

- Deprecated and unmaintained — Timescale shut this down in favor of their cloud product; the official docs are gone and no security patches are coming, so this is a dead end for any new adoption

- PostgreSQL is not a natural fit for high-cardinality metrics at scale; TimescaleDB helps, but teams pushing millions of series per second were always going to hit walls that purpose-built systems like Thanos or Mimir don't have

- Required installing a separate PostgreSQL extension (promscale_extension) in addition to TimescaleDB, meaning you owned three moving parts before writing a single byte of telemetry data

- The Jaeger storage plugin model it relied on has itself been superseded by OpenTelemetry Collector pipelines, so the integration story would need rethinking even if the project were alive

View on GitHub → Homepage ↗

// want more like this?

We dig through GitHub every week and send a few repos picked for what you actually care about — each with an honest take like this one.

Get finds in your inbox → Search again →