Telemetry and Persistence

How AutoDoctor builds telemetry, writes SQLite rows, stores metadata, and keeps API/dashboard reads consistent with current run context.
Categories:
Tags:

Who This Is For

  • Developers troubleshooting missing dashboard data.
  • Operators validating write/read consistency.

Telemetry Collection Flow

  1. Invoke-AutoDoctorTelemetryCollection builds telemetry object.
  2. JSON file is written as UTF-8 without BOM.
  3. Write-AutoDoctorTelemetry inserts module and system snapshot rows into SQLite.

Key telemetry object sections:

  • RunID, GeneratedAt, Hostname, AutoDoctorVersion
  • ExecutionStats
  • System
  • Modules

DB Write Flow

Diagnostics

Write-AutoDoctorDiagnostics inserts per-module status and summary into diagnostics.

Remediation

Write-AutoDoctorRemediation inserts remediation module status into remediation.

Telemetry

Write-AutoDoctorTelemetry writes:

  • module-level status and keys into telemetry_modules
  • derived system snapshot into system_info

Alerts

Write-AutoDoctorAlerts maps root-cause detected issues into alerts with severity.

Metadata File for Dashboard

latest_run.json is written at:

  • server/latest_run.json

Fields:

  • run_id
  • host_name
  • generated_time

The API endpoint /api/dashboard/meta reads this file and provides fallbacks if missing/corrupt.

Concurrency Notes

API DB connections set:

  • PRAGMA journal_mode=WAL
  • PRAGMA busy_timeout=5000

This supports concurrent writer/reader behavior between agent and API.

Validation Queries 

SELECT COUNT(*) FROM diagnostics;
SELECT COUNT(*) FROM alerts;
SELECT COUNT(*) FROM telemetry_modules;
SELECT * FROM system_info ORDER BY timestamp DESC LIMIT 5;

Next Steps