Known Limitations
Search and analytics
- Saved searches do not support declared parameters, version history, or
per-search export/import.
dumpdataandloaddataoperate on broader Django project data. - Cross-database joins are implemented through pandas. Both inputs are materialized in application memory, so large joins can be slow and consume substantial CPU and memory.
- DataFrame conversions have the same memory constraint. JSON values must be stored with numeric types for numeric aggregation.
- Numeric mode can be absent when all values are unique, and text summaries show the three most common values.
- Dashboard sharing, versioned export/import, and packaged example dashboards or search packs are not provided.
Deployment and tenancy
- SIEMatic does not include a worker-node inventory model or database-backed coordination for duplicate crawler instances. Operators must partition crawler work deliberately.
- Metrics are not bundled. Python logging is available, but operators must integrate their own metrics and monitoring stack.
- Object-level permissions through Django Guardian and multi-tenant isolation are not implemented. Model permissions and owner-scoped views must not be treated as a tenant security boundary.
Alert subscriptions are configuration-based
Alert recipients and thresholds are configured through crawler settings.
SIEMatic does not provide a database-backed AlertSubscription model. It also
does not provide a general database-backed configuration table. Thus, users
cannot manage personal rule or severity subscriptions through the application.