your ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/weekyour ad here · $149/week
Launched September 8, 2026
Pooml - Poor Man's Logger

Pooml - Poor Man's Logger

SELECT * FROM logs;

Visit website

Free2,806 impressions#5 of its week2 comments

Comments

>log in to comment
  • MixFox· 21d ago

    Nice Logo to remember.

  • Arpad Kun· 23d ago

    Can Pooml help diagnose performance issues with vector similarity queries? For example, query latency, index usage, or throughput, or would those metrics need to be instrumented and sent to Pooml separately?

Pooml (stands for "Poor Man's Logger") is a self-hosted, single-binary log and metrics server on top of SQLite, built for solo developers and small teams who want observability for their side projects and VPS deployments without running an ELK stack, a Prometheus + Grafana pair, or paying per-GB SaaS prices. One binary, two SQLite files, a web UI. Ship logs from FluentBit or plain HTTP, push metrics via OTLP or let pooml scrape Prometheus endpoints, query everything with real SQL, draw charts, build dashboards, and get alerts on your phone. The assumption is that you will self-host pooml on a VPS or a small cloud instance (Coolify, plain Docker, a systemd unit - anything works).

Pooml is a self-hosted single-binary log and metrics server using SQLite with a SQL query UI.

for
Solo developers and small teams needing lightweight observability for side projects or VPS deployments.
pricing
open source
license
AGPL-3.0
n0rdy/pooml 32 0Goupdated 6 days ago
works withFluentBitOTLPPrometheusDockersystemd

Key features

6 features of Pooml - Poor Man's Logger
  • SQL query language — Query logs and metrics directly with standard SELECT, WHERE, GROUP BY etc., no custom DSL.
  • FluentBit-compatible ingestion — Accepts logs via HTTP with automatic JSON, CLF or plain text detection and live tail.
  • Metrics collection — Push OTLP/HTTP metrics or scrape Prometheus endpoints, supporting counters and gauges.
  • Dashboards & charts — Build typed dashboards with stream, chart, and number panels that retain their SQL queries.
  • Alerting — Define alerts as SQL queries and deliver via Pushover or Once Campfire with cooldowns.
  • Built-in ops — Hourly retention cleanup, S3 backups via SQLite backup API, and self-monitoring metrics.

Use cases

  • Debug a recent incident by running a SQL query on recent error logs
  • Monitor checkout latency over time with hourly aggregated metrics
  • Set up an alert that notifies on high error rates via Pushover

Pooml - Poor Man's Logger FAQ

How do I ingest logs into Pooml?+

Send HTTP POST requests to the ingestion API (default port 8080) with an API key, or configure FluentBit to forward logs.

Can I query logs with AI assistants?+

Yes, the @pooml/mcp package lets Claude or other MCP clients send read-only SQL queries to your instance.

What storage does Pooml use?+

All logs and metrics are stored in three SQLite files on disk, which can be backed up or opened with any SQLite tool.

Is Pooml suitable for high-availability setups?+

No, it runs as a single-node binary with one login secret and does not provide clustering or HA.

How are alerts delivered?+

Alert rules are SQL queries; when they fire, notifications are sent via Pushover or Once Campfire.

What are the resource requirements?+

The binary idles at a few dozen megabytes of RAM and can run on the cheapest VPS instance.

Summarized by DevHunt from pooml.com · Sep 28, 2026. Details may change; check the official site.