Point the Sentry SDKs you already use at one binary on your own server. Issues, traces, logs and alerts, with your data never leaving it.
- JavaScript
- Python
- Go
- Rust
- Ruby
- PHP
- Java
- .NET
Your frontends, APIs, workers and mobile apps report to one Rustrak. Thousands of events go in; a short list of issues comes out.
Your frontends, APIs, workers and mobile apps report to one Rustrak. Thousands of events go in; a short list of issues comes out.
01 Issues
Thousands of events become a short list of issues, with how often each one happens, who it reaches and whether it is new.
Read the docsGrouped by fingerprint
Events with the same cause land on one issue instead of a thousand rows.
New, regressed, escalating
What changed since you last looked is flagged on the row.
Trends at a glance
A 24-hour sparkline on every issue, without opening anything.
Assign and resolve
Hand an issue to a teammate, or resolve it in the next release.
02 Debug
Stack traces open on your own code, with the source around the failing line, the breadcrumbs before it and the release it shipped in.
Your frames first. Application code is expanded; library frames step back.
Source maps. Upload them and minified JavaScript reads as the code you wrote.
Breadcrumbs and context. The clicks, requests and logs that led up to the error.
Release health. Crash-free sessions per release, so a bad deploy shows up fast.
03 Performance
Transactions and spans from the same SDKs: latency over time for every endpoint, and the waterfall of one slow request down to the query that held it up.
Ranked by time spent. Every endpoint by its share of total time, with p95 and failure rate beside it.
Span waterfall. Every database call, request and render inside one trace.
Any window. The last day, two weeks or all time, per environment.
Linked to errors. An error inside a trace opens on the request it broke.
04 Alerts
Alert rules decide what fires, check it before firing and route it to the people who can fix it. Nobody acknowledges it, it escalates.
Read the docsConditions
New issues, regressions, or a level and a rate you choose.
Channels
Slack, email and webhooks, one rule to as many as you need.
Digests
A daily summary for what does not need anyone right now.
Escalation
Unacknowledged alerts move on to the next person on their own.
05 Migrate
Rustrak speaks the Sentry protocol, so the official SDKs you already ship talk to it as they are. Point the DSN at your server; point it back to roll back.
Official SDKs. JavaScript, Python, Go, Rust, Ruby, PHP, Java and more.
No code changes. The DSN is the only line that moves.
Reversible. Switching back is the same one-line change.
Your data stays. Events never leave the server you run.
Compare
Sentry is a platform, and self-hosting it means running that platform. Rustrak is the part most teams use every day, built to run on the machine you already have.
Memory
Services to run
| Rustrak | Self-hosted Sentry | Sentry cloud | |
|---|---|---|---|
| Where your data lives | Your server | Your server | Sentry's cloud |
| Storage | SQLite or PostgreSQL | PostgreSQL, ClickHouse, Kafka, Redis | Managed |
| Price | Free, AGPL-3.0 | Free, plus the hardware | From $26/mo, by volume |
| Works with the Sentry SDKs | Yes | Yes | Yes |
06 Deploy
One container, SQLite by default or PostgreSQL when you want it. The first boot creates your admin user, and the dashboard is served from the same port.
- <60 MB
- 5k/s
- 1
Open source under AGPL-3.0. No per-event pricing, no sampling you did not ask for.