Skip to content

About

Notification server for Honeywell Galaxy Flex alarm systems

Topics

Resources

Stars

4 stars

Watchers

1 watching

Forks

Repository files navigation

Honeywell Galaxy SIA Notification Server

SIA-Server is a lightweight, self-hosted Python service that receives SIA protocol messages from Honeywell Galaxy alarm systems and forwards them as prioritized push notifications to your phone or computer.

It was created as a replacement for the discontinued free Honeywell push notification service, allowing users to regain full control over their alarm alerts without ongoing subscription costs.

Notifications are delivered via ntfy.sh — a free, open-source push notification service. Just install the ntfy app on your phone (Android/iOS), add your chosen topic to the app, and you're done. No account, no registration, no subscription required. Additional built-in providers include Telegram, Pushover and webhook for custom integrations. A plugin system also allows adding your own providers — see providers/README.md for details.

The project was originally developed and tested on a Honeywell Galaxy Flex 20. It has since also been verified with a Honeywell Galaxy Dimension 48 running firmware 6.93. Other Honeywell Galaxy models may be compatible, but have not been verified.

If your Galaxy notifications suddenly stopped working, or you are looking for a simple self-hosted alternative without installing a full home automation ecosystem, this project is for you.

IMPORTANT SECURITY NOTICE By default, the communication between the alarm panel and this server is unencrypted. This server is designed to be run on a trusted local network. An optional encryption module is available for supported platforms — see Security & Privacy Guidelines for details and recommended hardening steps before installation.

If you are having trouble self-hosting and would like to test against a live server before setting up your own, feel free to reach out via PM on DIYNOT or Reddit (username: ZebMcKayhan on both). I can set up a test account on my personal server to help you verify your panel is communicating correctly. Note that this is a personal setup with no guarantees of uptime or longevity.

Features

  • Self-Hosted: Runs on any local Windows or Linux machine (like a Raspberry Pi), or in a Docker container.
  • Real-time Notifications: Instantly forwards alarm events to your devices.
  • Prioritized Alerts: Uses notification priorities to distinguish between urgent alarms and routine events.
  • Advanced Notification Routing: Route notifications for different accounts to different channels or topics, each with its own optional authentication (Bearer Token or User/Pass).
  • Robust Protocol Handling: Correctly parses the multi-message protocol used by Galaxy Flex panels.
  • Broad SIA Level Support: The flexible parser can correctly handle event data from SIA Levels 0, 1, 2, and 3.
  • Optional IP Check Service: Handles the proprietary Honeywell "IP Check" heartbeat protocol as an integrated service, with connection watchdog monitoring that sends notifications when a panel stops sending heartbeats.
  • Connection Security Policies: Per-account ENABLED policy (Yes/No/Secure) and a configurable REJECT_POLICY (respond/drop) to control how invalid connections are handled.
  • Protocol State Machine: Enforces correct SIA message ordering. Any connection that does not start with a valid ACCOUNT_ID is immediately rejected or silently dropped.
  • Character Encoding Fixes: Decodes the proprietary character set used by Galaxy panels (e.g., Å, Ä, Ö).
  • Highly Configurable: Most user settings are in a simple sia-server.conf file, with advanced protocol constants located in the galaxy/ directory.
  • Extensible Notification Providers: Built-in support for ntfy.sh, Pushover and Telegram with a plugin architecture for custom providers. Drop a new provider file into the providers/ directory and it is auto-discovered at startup. See providers/README.md.

Prerequisites

  • A Honeywell Galaxy Flex alarm system with an Ethernet module (e.g., A083-00-10 or E080-4).
  • A Linux or Windows machine on the same network as the alarm system (a Raspberry Pi running Raspberry Pi OS is perfect).
  • Python 3.10 or newer.
  • The python3-requests package and the optional python3-uvloop package (for Linux).

File Structure

The project is structured to separate the server logic, protocol parsing, and configuration.

.
├── sia-server.py           # The main server application
├── sia-server.conf         # Main user configuration file.
├── configuration.py        # Loads and validates all configuration.
├── notification.py         # Handles formatting and sending of notifications.
├── ip_check.py             # IP Check Service module — handles the proprietary heartbeat protocol.
├── watchdog.py             # Generic heartbeat watchdog monitor (used by both IP Check and event heartbeats).
├── sia_server_tester.py    # Test client for sending SIA packets to the server.
├── Dockerfile              # Docker container definition for portable deployment.
├── docker-compose.yml      # Docker Compose configuration for easy container management.
├── README.md               # This file.
├── PanelSetup.md           # Panel Configuration help.
├── requirements.txt        # Required python packages.
├── galaxy/
│   ├── __init__.py
│   ├── README.md           # Technical description of the protocol.
│   ├── constants.py        # Constants used in the SIA protocol.
│   ├── parser.py           # Handles parsing of the Galaxy SIA protocol.
│   ├── protocol.py         # Low-level SIA block framing and checksum helpers.
│   └── encryption/         # Optional encryption module (platform-specific).
│       ├── __init__.py
│       └── *.so            # Compiled encryption module for supported platforms.
├── providers/
│   ├── __init__.py
│   ├── README.md           # How to create custom notification providers.
│   ├── base.py             # Abstract base class for notification providers.
│   ├── ntfy.py             # Built-in ntfy.sh notification provider.
│   ├── pushover.py         # Built-in Pushover notification provider.
│   ├── telegram.py         # Built-in Telegram notification provider.
│   └── webhook.py          # Built-in Webhook event sender.
└── asuswrt-merlin/
    ├── README.md           # Install instructions for Asuswrt-Merlin.
    ├── S99siaserver        # Entware service (init.d) file.
    └── check-sia.sh        # Watchdog script for Entware service.

Installation & Setup

This guide will walk you through the five main steps to get your server running.

Step 1: Download the Server Code

The recommended way is to download the latest stable release.

  1. Go to the Releases page on GitHub.
  2. Under the latest release, download the Source code (zip) file.
  3. Unzip the file to your chosen directory (e.g., /home/pi/Scripts/sia-server on Linux or C:\siaserver on Windows).
For Developers: Cloning with Git

If you want the latest development code, you can clone the repository directly:

git clone https://github.com/ZebMcKayhan/SIA-Server.git sia-server
cd sia-server

Step 2: Install Python and Dependencies

This server requires Python 3. The installation steps are different for Linux and Windows.

For Linux (e.g., Raspberry Pi, Debian, Ubuntu)

  1. Install Python (if needed): Most modern Linux systems come with Python 3 pre-installed. You can check with python3 --version. If you need to install it
    sudo apt update
    sudo apt install python3
  2. Install Dependencies: Use apt to install the required packages. uvloop is an optional performance enhancement.
    sudo apt update
    sudo apt install python3-requests python3-uvloop
  3. Install Encryption Dependencies (Optional): Only required if your platform has a compatible encryption module in galaxy/encryption/.
    sudo apt install python3-pycryptodome

For Windows

  1. Install Python: Download and install the latest Python 3 from the official Python website. Important: During installation, make sure to check the box that says "Add Python to PATH".
  2. Install Dependencies: Open a PowerShell or Command Prompt. It is strongly recommended to use python -m pip to ensure you are installing packages for the correct Python interpreter.
    python -m pip install requests pyopenssl

    Note: The extra package pyopenssl are optional but recommended to avoid potential HTTPS/SSL errors when sending notifications from Windows.

Step 3: Get the Notification App and Topic

Before configuring the server, get the notification app on your phone or computer. The instructions below are for ntfy.sh — the recommended default provider. For other providers, see the configuration file examples or providers/README.md.

  1. Follow the instructions at the ntfy.sh documentation to get the app.
  2. Inside the app, add a new topic. Choose a long, random, unguessable name for your topic to keep it private (e.g., alarm-skUHvisapP2J382MDI2).
  3. You will use the full URL of this topic (e.g., https://ntfy.sh/alarm-skUHvisapP2J382MDI2) in the configuration file.

Step 4: Configure Your Alarm Panel

Log into your Galaxy Flex panel's installer menu and configure the Ethernet module. The numbers in parentheses are the menu codes for a Galaxy Flex 20.

  • ARC IP Address: The IP of the machine running sia-server.py (e.g., 192.168.128.10). (Menu 56.1.1.1.4.1)
  • ARC Port: The port for the [SIA-Server] and optionally the [IP-Check] server. (Menu 56.1.1.1.4.1)
  • Protocol: SIA. Levels 0-3 are supported; Level 3 is recommended for the most detail. (Menu 56.1.1.1.4.2)
  • Account Number: Your 4 or 6-digit alarm account number. SIA Level 3 requires 6 digits. (Menu 56.1.2.1.1)
  • Encryption: Set to Off unless you have a compatible encryption module installed galaxy/encryption/. (Menu 56.3.3.5)
  • IP-Check: (Optional) To use the heartbeat feature, enable it by setting a time interval (e.g., 00:30 for 30 minutes). 00:00 means disabled. (Menu 56.3.3.7.1)
  • Eng. Test: Use this to send a test notification without generating a fault. (Menu 56.7.1)

Step 5: Configure the Server

Edit the sia-server.conf file to match your setup. The file is pre-populated with examples to guide you.

# On Linux
nano /path/to/your/sia-server/sia-server.conf

On Windows, simply edit the file with a text editor like Notepad.

Configuration

For most users, only three things need to be changed in sia-server.conf:

  1. Your account number and notification provider — Add a section with your panel's account number and configure your preferred notification provider (default: ntfy.sh).
  2. Port numbers — Match the ports you configured on the alarm panel.
  3. Logging — Choose Screen for testing, or File with a path for permanent use.

Everything else works out of the box with sensible defaults.

Configuration Explained

The primary configuration is done in sia-server.conf. This file is designed to be user-friendly and not sensitive to Python syntax. Advanced, technical constants are located in galaxy/constants.py.

  • Site Sections ([012345]): Each site is defined by a section where the header is the panel's unique Account Number.

    • SITE_NAME: A friendly name for the site (e.g., "Main House"). If omitted, the account number is used.

    • ENABLED: Controls the connection policy for this account. Accepts Yes, No, or Secure.

      • Yes — Accept all connections to this account (default).
      • No — Reject all connections from this account.
      • Secure — Only accept encrypted connections. Plaintext connections will be rejected. You will need a compatible encryption module in galaxy/encryption/ to enable encryption.
    • PROVIDER: The notification provider to use. Set to ntfy for ntfy.sh notifications, or None to disable notifications for this account. Custom providers are supported — see providers/README.md.

    • NTFY_TOPIC, NTFY_TITLE: Configure ntfy.sh notification delivery for this site.

    • NTFY_AUTH: Set to None, Token, or Userpass for private topics and provide the corresponding NTFY_TOKEN or NTFY_USER/NTFY_PASS keys.

  • [Default] Section: A special section for events from account numbers not specifically listed.

  • [SIA-Server] Section: Configure the ports and addresses for the main server.

    • ENABLED: Controls whether the SIA Event Server starts. Set to No if you only want to run the IP Check Service. Default: Yes.

    • REJECT_POLICY: Controls how invalid or unauthorised connections are handled. Accepts respond or drop.

      • respond — Send a SIA REJECT frame to the client (default).
      • drop — Silently close the connection without sending anything.
    • Event Heartbeats: Some panels can send heartbeat information through the normal SIA event path instead of using the separate IP Check service.

    • EVENT_HEARTBEAT_WATCHDOG: Enables watchdog monitoring from event heartbeats. Default: No.

    • EVENT_HEARTBEAT_EVENTTYPE: Event type that identifies an event heartbeat. Default: Old.

    • EVENT_HEARTBEAT_EVENTCODE: Event code that identifies an event heartbeat. Default: RX.

    • EVENT_HEARTBEAT_TEXT: Text prefix that identifies an event heartbeat. Default: [HEARTBT.].

      When enabled, a matching event heartbeat is parsed for its heartbeat interval. The interval is normally read from the HH:MM value in the event's action_text. For example:

      [HEARTBT.] 00:05 00001
      

      represents a 5-minute heartbeat interval.

      Some automatic test messages include a VAmmmm modifier, where mmmm is the heartbeat interval in minutes.

      The interval from the action_text is used first. If no valid interval is found there, the Value from a VAmmmm modifier is used as a fallback and converted from minutes to seconds.

      Successfully recognized event heartbeats with interval are consumed by the watchdog and are not sent as normal notifications.

  • [IP-Check] Section: Configure the optional IP Check Service.

    Note: The IP Check server validates all incoming heartbeat packets before responding. > It verifies the packet length and header. Invalid packets are dropped.

    • ENABLED: Controls whether the IP Check Service starts. Default: No.
    • LISTEN_ADDR, LISTEN_PORT: Configure the address and TCP port used by the IP Check Service. The default port is 10001.
    • Watchdog configuration is described in the [WATCHDOG] Section below. Watchdog keys may also be placed in this section for backwards compatibility with older configuration files.
  • [WATCHDOG] Section: Configure the generic heartbeat watchdog used by both the IP Check Service and event heartbeats.

    Note: This is the preferred location for watchdog configuration. > For backwards compatibility, watchdog keys may also be placed in the [IP-Check] section. > Existing configurations using the legacy [IP-Check] location continue to work.

    • WATCHDOG_THRESHOLD: Controls how long the server waits after a missed heartbeat before declaring the connection lost. The timeout is calculated as: WATCHDOG_THRESHOLD × heartbeat interval. For example, with a heartbeat interval of 10 minutes:

      • 2.1 results in a timeout after 21 minutes.
      • 1.1 results in a timeout after 11 minutes.

      This allows occasional missed heartbeats without immediately generating an alert. Values between 1.1 and 10.0 are supported. Set the value to 0 (or any value <= 1.0) to disable the watchdog entirely.

    • Notification Priorities: Watchdog events have independent notification priorities:

      • MONITORING_STARTED_PRIO: Notification priority when monitoring starts for an account. Default: 2.
      • CONNECTION_RESTORED_PRIO: Notification priority when a previously disconnected heartbeat connection is restored. Default: 2.
      • INTERVAL_CHANGED_PRIO: Notification priority when the panel changes its heartbeat interval. Default: 3.
      • WATCHDOG_TIMEOUT_PRIO: Notification priority when the watchdog detects a lost connection. Default: 4.

      Notification priorities range from 1 (lowest) to 5 (highest/urgent).

    • Custom Watchdog Notifications: The four watchdog events can have independently configurable notification messages:

      • MONITORING_STARTED: Generated when the first valid heartbeat is received for an account.
      • CONNECTION_RESTORED: Generated when a previously timed-out connection receives a heartbeat again.
      • INTERVAL_CHANGED: Generated when a connected panel changes its heartbeat interval.
      • WATCHDOG_TIMEOUT: Generated when the watchdog determines that a heartbeat has been missed for longer than the configured threshold.

      MONITORING_STARTED and INTERVAL_CHANGED have no notification by default. CONNECTION_RESTORED and WATCHDOG_TIMEOUT retain their original notification messages for backwards compatibility.

      Notification formats use %field placeholders. Fields that represent timestamps or durations can additionally use {format} to control their representation, for example:

      %new_panel_time{%YYYY-%MM-%DD %hh:%mm:%ss}
      %elapsed{%DDd %hh:%mm:%ss}
      %new_interval{%hh:%mm:%ss}
      

      The formatting rules are:

      • Timestamp fields use %YYYY, %YY, %MM, %DD, %hh, %mm, and %ss.
      • Duration fields use %DD, %hh, %mm, and %ss.
      • For durations, units are calculated from the largest unit requested to the smallest, regardless of the order in which the placeholders appear in the format. Omitted higher-order units are folded into the next requested unit. For example, 2 hours 15 minutes formatted as %mm:%ss becomes 135:00.
      • For duration %DD, leading zeroes are removed, but 0 is retained when the value is zero. The other duration fields always use minimum two digits. This allows formats such as 2d 05h 02m 25s or 2d 05:02:25.
      • If no format is specified, each field uses its default representation.

      Available placeholders:

      • General fields:

        • %account: Account number.
        • %site_name: Configured site name.
        • %current_state: State before the current heartbeat event.
        • %new_state: State after the current heartbeat event.
      • Last heartbeat fields:

        • %last_panel_time: Panel time from the last received heartbeat.
        • %last_interval: Heartbeat interval from the last received heartbeat, in seconds.
        • %last_threshold: Watchdog timeout calculated from the last heartbeat interval.
        • %last_seen: Server time when the last heartbeat was received.
        • %elapsed: Time elapsed since the last heartbeat.
      • New heartbeat fields:

        • %new_panel_time: Panel time from the heartbeat currently being processed.
        • %new_interval: Heartbeat interval from the heartbeat currently being processed, in seconds.
        • %new_threshold: Watchdog timeout calculated from the new heartbeat interval.
      • Server date/time fields:

        • %YYYY: Four-digit server year.
        • %YY: Two-digit server year.
        • %MM: Server month (01–12).
        • %DD: Server day (01–31).
        • %hh: Server hour (00–23).
        • %mm: Server minute (00–59).
        • %ss: Server second (00–59).

      The last_* and new_* naming describes the heartbeat data involved in the event:

      • last_* refers to the most recently known heartbeat before the current event.
      • new_* refers to the heartbeat currently being processed.

      A normal heartbeat event can contain both last_* and new_* values. A WATCHDOG_TIMEOUT has no new_* values because no new heartbeat was received. Conversely, MONITORING_STARTED has no last_* values because it is the first heartbeat received.

      Fields that are not applicable to a particular event are left empty. Optional sections using [ ... ] are also supported, following the same syntax as SIA notification formats.

      For example:

      MONITORING_STARTED = Monitoring started at %new_panel_time{%YYYY-%MM-%DD %hh:%mm:%ss}, interval: %new_interval{%hh:%mm:%ss}
      CONNECTION_RESTORED = Monitoring restored at %new_panel_time{%YYYY-%MM-%DD %hh:%mm:%ss}, after %elapsed{%DDd %hh:%mm:%ss}
      INTERVAL_CHANGED = Heartbeat interval changed from %last_interval{%hh:%mm:%ss} to %new_interval{%hh:%mm:%ss}
      WATCHDOG_TIMEOUT = Heartbeat lost. Last heartbeat: %last_panel_time{%hh:%mm:%ss}. Last seen: %last_seen{%YYYY-%MM-%DD %hh:%mm:%ss}
      

      \n can be used to insert a line break into a notification.

  • [Logging] Section: Control the log level and output destination.

    • LOG_LEVEL: Set the verbosity of logs (DEBUG, INFO, WARNING, ERROR). INFO is recommended for normal use.

    • LOG_TO: Choose Screen, File, or Syslog.

      • Screen is best for manual testing and standard systemd services.
      • File is best for creating a dedicated log file (e.g., on Windows or for cron jobs).
    • File-Specific Settings: LOG_FILE, LOG_MAX_MB, and LOG_BACKUP_COUNT are only used when LOG_TO = File.

Advanced Logging: Using LOG_TO = Syslog

Setting LOG_TO = Syslog integrates the server's logging with the native operating system logger. This is an advanced option recommended for embedded systems like routers.

  • Log Format: When using Syslog, the server uses a simpler log format (SIA-Server: LEVEL - Message) because the syslog service adds its own timestamps and hostname.

  • On Linux/Unix Systems:

    • The server will attempt to write to the standard /dev/log socket.
    • For non-standard systems, you can specify a different path with the optional SYSLOG\_SOCKET key.
    • You can also change the syslog "facility" (which controls how the system syslogd categorizes the messages) using the optional SYSLOG_FACILITY key. Common values are daemon or local0 through local7. This can be useful for tailoring log filtering rules on your specific system.
  • On Windows Systems:

    • This will log messages to the Windows Event Log under the "Application" section with the source name "SIA-Server".
    • Dependencies: This feature requires the pywin32 package. You must install it from an Administrator prompt: python -m pip install pywin32.
    • Permissions: The very first time you run the server with this option, it must be run as an Administrator to register the "SIA-Server" source in the Windows Registry. After that, it can be run as a normal user.
    • If either the dependency is missing or the registration fails due to permissions, the server will print a clear warning and automatically fall back to logging to the screen.
  • [Notification] Section: Configures notification formatting, event priorities, and the resilient retry queue.

    • Custom Notification Formats:

      • Notification formats use %field to insert information from the GalaxyEvent or optionally server date & time.

      • Available placeholders:

        • GalaxyEvent fields:

          • %account: Account number.
          • %site_name: Site name.
          • %time: Event time (from panel, if present).
          • %action_text: Human-readable text from the SIA ASCII block.
          • %event_type: Event type (New/Old).
          • %event_code: SIA event code (e.g. OP, BA, RX).
          • %event_description: Description of the event code.
          • %user_id: User ID.
          • %zone: Zone address.
          • %group: Group number.
          • %peripheral: Peripheral/module number.
          • %value: Event value.
        • Server Date & Time fields (server local time when event is processed):

          • %YYYY: 4-digit server year (e.g. 2026).
          • %YY: 2-digit server year (e.g. 26).
          • %MM: 2-digit server month (01-12).
          • %DD: 2-digit server day (01-31).
          • %hh: 2-digit server hour (00-23).
          • %mm: 2-digit server minute (00-59)
          • %ss: 2-digit server second (00-59)
        • Sections enclosed in [ ] are optional. The entire section is omitted if any field used inside the section is unavailable.

        • Use \n to insert a newline in the notification (can also be used inside optional sections, e.g. [\nGroup: %group]).

        • Examples

          NOTIFICATION_FORMAT_ASCII = %hh:%mm %action_text[\t(Group: %group)]
          NOTIFICATION_FORMAT_DATA = %YYYY-%MM-%DD %hh:%mm:%ss %event_type event: %event_code (%event_description)[\nUser: %user_id][\nZone: %zone][\nGroup: %group][\nPeripheral: %peripheral][\nValue: %value]
          
    • NOTIFICATION_FORMAT_ASCII: Format used for SIA Level 3 events where human-readable action_text is available from the panel (default: %time %action_text [(Group: %group)]).

    • NOTIFICATION_FORMAT_DATA: Format used for SIA Level 0-2 events where human-readable action_text is not available (default: %time Event: %event_code [(%event_description)][ User: %user_id][ Zone: %zone][ Group: %group][ Peripheral: %peripheral][ Value: %value] ).

    • MAX_QUE_SIZE: Number between 1 and 1000 Maximum number of events to queue up if failed to send.

    • MAX_RETRIES: 0 = infinite, Maximum number of times to retry to send.

    • MAX_RETRY_TIME: Number between 1-1000, Max minutes between retries. Retries will always be after 1, 2, 4, 8, 16 minutes but this puts a limit on how much it can be increased.

    • PRIORITY_1 through PRIORITY_5: Assign SIA Event Codes to different priority levels (1 = lowest, 5 = urgent).

    • DEFAULT_PRIORITY: The priority to use for any unlisted or unknown event code.

Usage

For Linux

Manual Start (for testing)

Note: It's convenient to set LOG_TO = Screen in sia-server.conf to see live events in your terminal.

cd /path/to/your/sia-server
python3 sia-server.py

Press Ctrl+C to stop.

Custom Configuration File Path

By default, the server looks for sia-server.conf in the current directory. You can specify a different path using the --config argument:

python3 sia-server.py --config /path/to/your/sia-server.conf

As a Service (Recommended)

Note: Set LOG_TO = File in sia-server.conf to keep a persistent log.

  1. Create the Service File: sudo nano /etc/systemd/system/sia-server.service
  2. Paste this content, changing the paths in WorkingDirectory and ExecStart.
    [Unit]
    Description=Galaxy SIA Alarm Server
    After=network.target
    [Service]
    Type=simple
    User=pi
    WorkingDirectory=/home/pi/Scripts/sia-server
    ExecStart=/usr/bin/python3 /home/pi/Scripts/sia-server/sia-server.py
    Restart=on-failure
    RestartSec=5s
    [Install]
    WantedBy=multi-user.target

    Note: You may need to add firewall rules (e.g., via ExecStartPre= & ExecStopPost=). If your firewall commands require root, you may need to remove or comment out the User=pi directive.

  3. Enable and Start:
    sudo systemctl daemon-reload
    sudo systemctl enable sia-server.service
    sudo systemctl start sia-server.service
  4. Manage:
    • Check status & recent logs: sudo systemctl status sia-server.service
    • Stop the service: sudo systemctl stop sia-server.service
    • Start the service: sudo systemctl start sia-server.service
    • Restart the service: sudo systemctl restart sia-server.service
    • View live logs: journalctl -u sia-server.service -f (if not logging to a file) or tail -f /path/to/your/log/file.log (if logging to a file).
    • Reload configuration (accounts and log level): kill -HUP $(pgrep -f sia-server.py) or systemctl kill -s HUP sia-server.service

For Windows

Manual Start (for testing)

Note: Set LOG_TO = Screen in sia-server.conf to see live events.

cd C:\path\to\your\sia-server
python sia-server.py

Note: If your network setting is set to public, instead of private, you may need to add a rule windows firewall to accept inbound connections on the port (i.e. 10000) you are using. Search online how to do this for your windows version.

Press Ctrl+C to stop.

Custom Configuration File Path

By default, the server looks for sia-server.conf in the current directory. You can specify a different path using the --config argument:

python sia-server.py --config C:\path\to\your\sia-server.conf

As a Service (Recommended)

Note: Set LOG_TO = File in sia-server.conf to keep a persistent log.

  1. Download NSSM.
  2. Open a Command Prompt as an Administrator.
  3. Run the installer: C:\path\to\nssm.exe install SIA-Server
  4. In the GUI that pops up:
    • Path: Browse to your Python executable (e.g., C:\Python312\python.exe).
    • Startup directory: Browse to your script folder.
    • Arguments: sia-server.py
  5. Click Install service. You can now manage it from the Windows Services app (services.msc).

Using Docker

Prerequisites

Setup and Start

  1. Place your configured sia-server.conf in the same directory as docker-compose.yml.
  2. Check that the ports in docker-compose.yml match the ports configured in sia-server.conf:
ports:
  - "10000:10000"   # Must match [SIA-Server] LISTEN_PORT
  - "10001:10001"   # Must match [IP-Check] LISTEN_PORT
  1. Start the container:
docker compose up -d

The Compose file uses the published SIA-Server image from Docker Hub. Docker will pull the image automatically if it is not already available locally.

  1. The server is now running. To verify:
docker compose ps

View Logs

docker compose logs -f

Build Locally

If you have cloned the SIA-Server repository and want to build the image from the local source instead of using the published Docker Hub image:

docker compose build
docker compose up -d

Stop the Server

docker compose down

Reload Configuration

To reload accounts and log level without restarting the container:

docker compose kill --signal=HUP sia-server

File Logging (Optional)

By default the server logs to screen, which is captured by Docker and visible via docker compose logs. If you prefer a persistent log file, set the following in sia-server.conf:

LOG_TO = File
LOG_FILE = /logs/sia-server.log

The log file will appear as ./logs/sia-server.log on the host machine. The ./logs directory will be created automatically by Docker if it does not exist.

Timezone & Health Check

By default the container runs in UTC. Set your local timezone in docker-compose.yml:

environment:
  - TZ=Europe/London    # Change to your local timezone

An optional application-level health check can be enabled to verify that the SIA Event Server and IP-Check module are alive and responsive.

To use the Docker health check, set the ports you want to test in the environment variables:

environment:
  - HEALTHCHECK_PORT1=10000 # Port to test
  - HEALTHCHECK_PORT2=10001 # Optional, set to 0 if not used

Then uncomment the healthcheck section in docker-compose.yml:

healthcheck:
  test: ["CMD-SHELL", "python3 -u sia_server_tester.py --ping $${HEALTHCHECK_PORT1:-0} $${HEALTHCHECK_PORT2:-0} --quiet"]
  interval: 60s
  timeout: 5s
  retries: 3
  start_period: 10s

Docker will then report the container as healthy or unhealthy based on the health check result. Docker does not automatically restart or otherwise act on an unhealthy container unless additional configuration is used.

Health Check

SIA-Server provides an application-level health check for both the SIA Event Server and the IP-Check module. It can be used by monitoring applications to verify that the services are alive and responsive.

The health check uses a simple TCP PING/PONG exchange. Send the ASCII text PING to the listening port and the server will respond with the ASCII text PONG.

PING must be an exact match. Additional characters are not accepted, so for example PING\n will not trigger a response.

Access Control

Health check requests are subject to the configured REJECT_POLICY.

When REJECT_POLICY = drop, PING is only answered when the connection originates from localhost (127.0.0.1). With other REJECT_POLICY settings, PING is answered regardless of the source address.

This allows the health check to be used safely by local monitoring and container health checks without creating an unrestricted response mechanism when rejected connections are configured to be silently dropped.

Monitoring Applications

The health check can be used with monitoring applications such as Uptime Kuma, Zabbix, Nagios, or Icinga to monitor the SIA Event Server and IP-Check module and trigger alerts or other actions when they become unavailable.

The sia_server_tester.py utility includes support for performing these checks:

python3 sia_server_tester.py --ping port1 [port2 ...]

The tester sends PING to each specified port and verifies the response. It exits with status code 0 when all specified ports respond successfully, and status code 1 if any check fails.

Port numbers set to 0 are skipped, allowing individual checks to be disabled.

For Docker deployments, this functionality is available as an optional Docker health check. See Using Docker.

Security & privacy guidelines

1. Local Network Communication (Panel to Server)

The communication between your alarm panel and this server is unencrypted by default. Run it on a trusted local network (LAN).

Warning: Do not expose the server's listening ports directly to the public internet. If you must, use a VPN (e.g., WireGuard) or enable the optional encryption module (see below). Note that the encryption module is only available for selected platforms.

If your server is exposed to the internet, enabling encryption on the panel combined with the following configuration makes the server much more resilient against unwanted access:

  • Set ENABLED = Secure for all accounts — plaintext connections are rejected.
  • Set ENABLED = No for the [Default] account — unknown accounts are rejected.
  • Set REJECT_POLICY = drop — rejected connections are silently dropped without response.
  • Use random, unguessable account numbers on the panel.

Note: Encrypted sessions require the optional encryption module, which is available for selected platforms only, see ./galaxy/encryption/ folder.

2. Notification Privacy (Server to Notification Providers)

  • Transport Security: All built-in providers use HTTPS and are secure.
  • Topic Privacy (ntfy.sh): ntfy.sh topics are public by default. To secure them:
    • Use a long, unguessable topic name.
    • Consider a generic Site Name that cannot be linked to your address.
    • Alternatively: Subscribe to ntfy.sh PRO to set up private channels with authentication. This server fully supports authentication via the NTFY_AUTH settings.
    • Alternatively: Host ntfy yourself for private channels free of charge (requires a machine with a public IP).

Disclaimer: You are ultimately responsible for securing your own setup.

Alternative Versions

Looking for a fully containerized setup with Apprise notification routing and MQTT integration? @claesmathias maintains a fork of this project with Docker Compose, Apprise and Mosquitto pre-configured.

Acknowledgments

  • This project was developed through a collaborative effort with Anthropic's AI assistant, Claude.
  • The initial socket server structure was inspired by the nimnull/sia-server project.
  • Some protocol information was found in dklemm/FlexSIA2MQTT project.
  • sia_server_tester.py was contributed by @claesmathias.

License

This project is licensed under the MIT License.

About

Notification server for Honeywell Galaxy Flex alarm systems

Topics

Resources

Stars

4 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages