Watch every container,
database, and endpoint — in one place.

Three things, one dashboard.

No configuration sprawl. No agent fleet. Drop in one binary, register the targets you care about, watch them.

Read-only by design.

We don't host your database. We don't ask you to copy data anywhere. For database monitoring, you give us a user with the smallest footprint possible — and we pull metrics, never your data.

1. 01

   One read-only user

   You create a user with SELECT-only access to system catalogs and built-in stored procedures — pg_stat_*, INFORMATION_SCHEMA, performance_schema. Nothing else.

2. 02

   No access to your tables

   The user can't read application schemas, can't see row data, can't write anything. We literally cannot view what's in your database.

3. 03

   Metrics only — over your network

   We pull connection counts, slow-query stats, replication lag, cache hit ratios. The connection runs from your environment to ours; nothing inbound is required.

Why we're building this.

We were fed up juggling a different tool for every layer — one for container logs, another for database health, a third for uptime checks, a fourth for alerts that never quite worked. So we built Sentinel to put all of it in one place. Watch a Docker container and inspect its logs and crash pattern in the same view. Probe Postgres, MySQL, or Redis for slow queries and connection drops without spinning up a second dashboard. One timeline of what failed, when, and why. Pricing is pay-as-you-go — start free, pay only for what you actually monitor. Some of the building blocks stay open source so you can see exactly what runs on your machines.