Connector Sentry

Not currently available – coming soon

OSentry is a REST API connector from Omnis Studio to Sentry, linking Omnis applications with professional error monitoring. Errors, warnings, and relevant runtime events are transmitted directly from Omnis to Sentry, where they are evaluated in a structured way. This makes problems visible before they have to be painstakingly reconstructed through support. The adapter collects important information from Omnis, prepares it for Sentry, and helps developers understand, prioritise, and fix errors more quickly.

Sentry is a platform for error monitoring and application monitoring. It captures errors from applications, groups similar incidents, shows affected environments, and provides context about the cause. This turns a simple error message into a comprehensible report with technical and functional insights.

Features

• • • Direct connection of Omnis to Sentry via DSN
  • • Capturing of exceptions, messages, logs, and SQL errors
  • • Support for severity levels such as fatal, error, warning, info, and debug
  • • Automatically generated Omnis context information
  • • Omnis stack trace with classes, methods, lines, and context
  • • Passing of tags for categorising events
  • • Passing of extras for additional diagnostic information
  • • Optional user context with ID, name, and email
  • • Proxy support for protected networks
  • • Enabling and disabling Sentry transmission
  • • Local saving of reports in case of transmission errors
  • • JSON preparation tailored to the Sentry interface
  • • Support for revision and version information for better tracking of errors across different releases

OSentry consists of a central low-level adapter and a high-level layer. The OSentry class handles the actual communication with Sentry: initialisation, data preparation, JSON generation, HTTP transmission, and error handling. Initialisation is done via the Sentry DSN, from which the required information such as public key, host, and project ID are extracted, allowing the application to send events directly to the desired Sentry project.

The high-level class OSentryHL simplifies typical use cases. It provides methods to capture general exceptions, info messages, log entries, SQL errors, deprecated function calls, or licence violations with minimal effort. This allows OSentry to be quickly integrated into existing Omnis applications without having to manually construct each event.

A major advantage of the adapter is the automatically appended application context. OSentry adds information such as the Omnis version, platform, operating system version, error code, open windows, open reports, and the current print file. This information helps to understand errors not in isolation, but in the actual state of the application.

Particularly valuable is the Omnis-specific stack trace. OSentry reads the current Omnis stack and removes internal adapter calls, so that the relevant classes, methods, and lines are visible in Sentry. Additionally, context lines, parameters, and variable states can be included. This makes it possible to understand exactly where an error occurred and what data was involved at that point.

Via tags and extras, Sentry events can be flexibly enriched. Tags are suitable for filtering and grouping, for example by database, module, client, or environment. Extras provide additional details, such as SQL text or native database error messages. A user context can also be set if desired, so that affected cases can be better classified.

For production use, OSentry offers practical operational features. Transmission can be enabled or disabled, proxy information can be configured, and if transmission fails, the generated report is saved locally. This ensures that important diagnostic information is not lost, even if the network or Sentry is temporarily unavailable.

With OSentry, Omnis developers gain a compact tool to make applications more transparent, maintainable, and support-friendly. Errors become visible more quickly, causes are easier to trace, and support cases can be better prioritised.