Migrate from Sentry to MOJAQ

Why switch

Sentry is usually wired deep into production, across frontends, services, workers and release pipelines. A full instrumentation rewrite is not a realistic plan. MOJAQ takes the practical path: keep the Sentry SDK, replace the DSN, and move new error data into an EU-hosted environment.

The compliance reason is just as direct. Error events often contain request paths, user IDs, IP addresses, headers and fragments of state, which can be personal data under GDPR even after scrubbing. MOJAQ keeps that data in Helsinki, includes a DPA, and is built to avoid US transfers. And because it is free during the open beta with no fixed end date, you can run it in parallel before switching production traffic.

Step by step

1. Get a MOJAQ DSN for your project from the dashboard. You can keep the existing SENTRY_DSN variable name, the SDK only cares about the value.
   Swap the DSN value

   # Before
   SENTRY_DSN=https://<key>@o123.ingest.sentry.io/456
   
   # After
   SENTRY_DSN=https://<key>@errors.mojaq.com/<project>

2. Point the SDK at MOJAQ. Leave the rest of your config (integrations, traces, release, environment, before-send) exactly as it is.
   Sentry.init

   Sentry.init({
     dsn: process.env.SENTRY_DSN,        // now your MOJAQ DSN
     environment: process.env.NODE_ENV,
     release: process.env.COMMIT_SHA,
   });

3. Send a test event from each runtime so you can confirm ingestion end to end.
   Capture a test exception

   Sentry.captureException(new Error("MOJAQ migration test"));
   await Sentry.flush(2000);

4. Verify it arrives in MOJAQ with the right environment and release, readable stack frames and breadcrumbs present. Keep the same release and dist values so uploaded source maps still match.

What carries over

Carries over: your Sentry SDKs, DSN ingestion, issue grouping, releases, environments, breadcrumbs, handled and unhandled exceptions, and source-map matching when release identifiers stay consistent. What does not copy automatically: historical issues, alert rules and saved views, recreate the ones you still want, then let new production errors build a clean EU-hosted history.

Questions