Grafana Alloy App/Add-on: Ship HAOS Logs to Loki (Promtail Replacement)

Share your Projects! Custom Integrations
1

Grafana Alloy Add-on: Ship HAOS Logs to Loki (Promtail Replacement)

If you’ve been using the official Promtail add-on to ship your Home Assistant logs to Loki, you may have noticed it silently stopped working sometime after upgrading to HAOS 11+. I built a drop-in replacement using Grafana Alloy and published it as a community add-on.

Repository: GitHub - ecohash-co/ha-addon-alloy: Home Assistant add-on: Ship HAOS logs to Loki using Grafana Alloy

Why Promtail is Broken on Modern HAOS

The official Promtail add-on (v2.2.0) ships Promtail 2.6.1, which was released in 2022. Since then:

  • systemd 252+ introduced a compact journal format that Promtail 2.6.1 can’t read
  • HAOS 11+ ships with this newer systemd version
  • The add-on hits "received error during sdjournal follow: Timeout expired" in a loop
  • Promtail itself was deprecated by Grafana — EOL is March 2, 2026
  • The official HA add-on repository hasn’t been updated to address this

The result: your HAOS instance generates logs, but nothing makes it to Loki. No errors in the HA UI — it just silently fails.

What is Grafana Alloy?

Grafana Alloy is the official successor to Promtail, Grafana Agent, and Grafana Agent Flow. It’s a single binary that can collect logs, metrics, and traces using a component-based pipeline architecture (think modular building blocks connected in a DAG).

For our purposes, the key advantage is simple: it can read modern systemd journals.

Installation

1. Add the repository

  • Settings > Add-ons > Add-on Store
  • Click the overflow menu (three dots, top-right) > Repositories
  • Paste: https://github.com/ecohash-co/ha-addon-alloy
  • Click Add, then Close

2. Install and configure

  • Find Grafana Alloy in the store and click Install
  • In the Configuration tab, set your Loki URL:
loki_url: "http://<your-loki-ip>:3100/loki/api/v1/push"
log_level: info
  • Start the add-on
  • Enable Start on boot and Watchdog (recommended)

3. Verify

Check the add-on log — you should see:

Grafana Alloy for Home Assistant
Loki URL: http://192.168.1.45:3100/loki/api/v1/push
Journal path: /var/log/journal

And in Grafana, query {job="systemd-journal"} — your HAOS logs should start appearing.

What Gets Shipped

The add-on reads the full HAOS systemd journal and forwards everything to Loki with useful labels:

Label Example Description
job systemd-journal Static identifier
unit docker.service Systemd unit
hostname homeassistant Machine hostname
syslog_identifier addon_core_whisper Process/add-on name
container_name addon_a0d7b954_spotify Docker container
level info, error Log priority
transport journal, stdout How the log entered journald

This means you can filter in Grafana like:

  • {job="systemd-journal", level="error"} — all errors across HAOS
  • {job="systemd-journal", syslog_identifier="homeassistant"} — just HA Core logs
  • {job="systemd-journal", unit="docker.service"} — all add-on container logs

There’s also a debug UI at http://<haos-ip>:12345 where you can inspect the pipeline, see component health, and troubleshoot.

Don’t Have Loki Yet? Here’s Why You Should

If you’re running a homelab and don’t have centralized logging yet, you’re flying blind when things break. Here’s a minimal setup to get started.

Why Loki?

  • It’s free and lightweight. Loki is designed to be cost-effective — it indexes labels, not the full log content, so storage requirements are modest.
  • You already have the dashboards. If you run Grafana (and many HA users do for energy dashboards, sensor history, etc.), Loki plugs right in as a data source. Same query interface, same alerting.
  • “What happened at 3 AM?” — When Home Assistant restarts unexpectedly, an automation misfires, or a Z-Wave device drops off, the journal has the answer. But HAOS only keeps a limited journal buffer. Loki gives you weeks or months of searchable history.
  • Multi-node visibility. If you run other Docker hosts, NAS boxes, or network gear alongside HAOS, Loki can ingest logs from all of them in one place. One Grafana dashboard, every device.

Quick Loki Setup (Docker)

On any machine with Docker (a NAS, a Raspberry Pi, a VM — doesn’t need to be your HA host):

# docker-compose.yml
services:
  loki:
    image: grafana/loki:latest
    container_name: loki
    ports:
      - "3100:3100"
    volumes:
      - loki-data:/loki
      - ./loki-config.yml:/etc/loki/local-config.yaml:ro
    command: -config.file=/etc/loki/local-config.yaml
    restart: unless-stopped

volumes:
  loki-data:

# loki-config.yml
auth_enabled: false

server:
  http_listen_port: 3100

common:
  path_prefix: /loki
  storage:
    filesystem:
      chunks_directory: /loki/chunks
      rules_directory: /loki/rules
  replication_factor: 1
  ring:
    kvstore:
      store: inmemory

schema_config:
  configs:
    - from: 2020-10-24
      store: tsdb
      object_store: filesystem
      schema: v13
      index:
        prefix: index_
        period: 24h

limits_config:
  reject_old_samples: true
  reject_old_samples_max_age: 168h

analytics:
  reporting_enabled: false

docker compose up -d

Then add Loki as a data source in Grafana: Connections > Data Sources > Add > Loki > URL: http://<loki-ip>:3100.

Grafana Starter Query

Once logs are flowing, try this in Grafana’s Explore view:

{job="systemd-journal"} |= "error" | line_format "{{.syslog_identifier}}: {{.__line__}}"

This shows all error-containing lines with the originating service name — great for a quick health check.

Technical Details

The add-on uses s6-overlay to manage the Alloy process. On startup, a oneshot init script reads your add-on options, detects the journal path, and generates an Alloy config. The longrun service then starts Alloy with that config.

For advanced users, the additional_config option lets you append arbitrary Alloy config blocks — for example, to scrape the home-assistant.log file in addition to the journal.

Feedback Welcome

This is a v1.0 born out of personal frustration with the broken Promtail add-on. If you run into issues or have feature requests, please open an issue on the GitHub repo.

Happy logging!

2

Thanks! I’ve been debugging the promtail addon for a few days now, after I noticed it stop working. This addon works great.