DNS logs show requested domains but miss connections that have no matching DNS trail. Understanding network activity requires both views, with a clear distinction between observed traffic and missing coverage.
— Case study 22
Network Monitor
DNS queries and passive connection logs, joined into a readable network picture

project overview
A home-network telemetry system combining AdGuard Home DNS logs, passive Zeek connection records, PostgreSQL and MaxMind geolocation. Grafana presents queries, destinations, flows and alerts. The screenshots use generated demo records only.
shipped outcome
The source includes DNS ingestion, passive flow mapping, anomaly checks and four provisioned dashboards. For this case study, an isolated Grafana and PostgreSQL instance renders synthetic DNS, flow and geolocation records. No home-network sensor or real device data was used.
- Role
- Telemetry pipeline and dashboard development
- Timeline
- 2026 — local telemetry system; synthetic demo shown
- Year
- 2026
- Team
- Solo
A NestJS collector polls AdGuard logs and tails Zeek connection and TLS logs. It maps devices, deduplicates events and enriches public destinations through a cached MaxMind lookup. PostgreSQL keeps DNS and flow records separately; provisioned Grafana dashboards query both.
The source includes DNS ingestion, passive flow mapping, anomaly checks and four provisioned dashboards. For this case study, an isolated Grafana and PostgreSQL instance renders synthetic DNS, flow and geolocation records. No home-network sensor or real device data was used.
— Interface evidence
The product, not a placeholder
These are the verified interface captures from the project build. Scroll inside taller frames to inspect each complete page.



— Delivery process
From operating constraint to shipped system
The work is sequenced around risk. Domain rules and failure states come before interface polish; automation arrives before handoff.
Collect DNS observations
Poll AdGuard queries and associate them with registered device identities.
Add the passive flow view
Read Zeek connection and TLS logs to cover traffic that DNS alone cannot explain.
Enrich destinations once
Cache IP geolocation and ASN lookups before joining them into dashboard queries.
Separate demonstration from observation
Capture the real dashboards against generated records, with a visible synthetic-data banner.
— System architecture
Clear boundaries, explicit responsibilities
Each layer has one job and a narrow contract. That keeps external services replaceable and product behavior testable.
two observation paths
DNS
AdGuard Home
Flows
Zeek · TLS / UDP
queries ≠ all connections
join / enrich / read
- device identity
- PostgreSQL event history
- cached location + ASN
- Grafana views
captures: synthetic data only
AdGuard Home
Query logs provide domains, client addresses and blocked-query status.
Passive Zeek sensor
Connection and TLS metadata add protocols, bytes and available server names.
NestJS + PostgreSQL + GeoLite2
Collectors map devices, deduplicate records and cache destination enrichment.
Provisioned Grafana dashboards
Overview, device, destination-map and flow views use committed SQL queries.
— What shipped
Product capabilities
DNS and connection views
Separate records expose both domain queries and observed internet-egress flows.
Destination geography
Cached location and ASN fields support a map and destination detail tables.
Traceable alert rules
Source-defined checks flag frequent domains and connections without a recent DNS match.
— Engineering pressure
Challenges resolved
DNS is incomplete evidence
A connection can exist without a visible DNS query; the flow view preserves that distinction.
Device attribution depends on deployment
Docker Desktop relays hide original client addresses, so the source uses native AdGuard for device attribution.
Demonstrate without exposing a household
Documentation IP ranges, demo device names and generated events keep the portfolio independent of private traffic.
verified project destinations
next transmission
Need a system this deliberate?
Bring the operating problem. We will shape it into a clear product brief and a buildable next step.
Start a project