Skip to content
03

Data and messaging

State & Queue

Data and messaging keep SecAudit durable — PostgreSQL for platform state, Redis Sentinel for queues, locks, and schedule coordination.

PostgreSQL stores platform state. Redis Sentinel provides highly available queues, distributed locks, and scheduling

Where it sits in the stack

  1. 01
    Access layer
  2. 02
    Control plane
  3. 03
    Data and messaging Focus
  4. 04
    Execution layer
  5. 05
    Managed targets
PostgreSQLRedis SentinelQueuesLocksBeat

How requests move

  1. 01

    Control plane writes durable state

  2. 02

    Tasks enter Redis-backed queues

  3. 03

    Locks serialize contested work

  4. 04

    Workers claim jobs and stream progress

What this layer does

  • PostgreSQL as the source of truth for platform state

  • Redis Sentinel for highly available task queues

  • Distributed locks that help reduce duplicate execution

  • Schedule coordination for recurring jobs and reports

PostgreSQL holds inventory, profiles, job definitions, run history, waivers, and audit events so operators can rebuild context after any restart.

Redis Sentinel backs Celery queues and distributed locks, which helps reduce duplicate work when multiple workers compete for the same task.

Beat-driven schedules land here too: recurring compliance jobs and report delivery stay coordinated even when individual workers recycle.

Other layers

Back to Technology
01

Access layer

Operators work in a single PatternFly interface with Keycloak OIDC or local authentication and object-level RBAC

 02

Control plane

FastAPI provides REST and WebSocket APIs. secaudit_core manages profiles, AuditFlow, jobs, and reports

 04

Execution layer

Celery workers execute checks, remediation jobs, discovery scans, and scheduled tasks while streaming logs in real time

 05

Managed targets

Agentless SSH, WinRM, and network protocols connect to Linux, Windows, and network devices