syncmux 0.1.1

Fast file synchronization for remote development over SSH

syncmux

Fast file synchronization for remote development over SSH. syncmux is designed for developers working on remote servers who need low-latency file syncing with persistent connections and robust logging.

Features

• 🚀 Lightning Fast: Uses SSH ControlMaster to maintain persistent connections, eliminating handshake overhead for every file.
• 📦 Smart Batching: Automatically switches to batch mode when multiple files change at once (e.g., switching git branches).
• 🔄 Auto-Reconnect: Automatically attempts to reconnect if the network drops or the SSH connection is severed.
• 🎯 Git-Aware: Respects .gitignore and logs git diffs for changes.
• 📝 Comprehensive Logging: Keeps detailed logs of what changed on both the local machine and the remote server.

Prerequisites

• Passwordless SSH: You must have SSH key-based authentication set up for the remote host (e.g., via ssh-copy-id), allowing usage without entering a password.
• rsync: Must be installed on both the local machine and the remote server.

Installation

pip install syncmux

Usage

Quick Start

Start watching a directory and syncing to a remote server:

syncmux --local-dir ~/myproject --remote-host dev-server --remote-dir /home/user/myproject

Initial Sync

If you want to ensure the remote directory is an exact mirror of your local directory before starting the watcher (useful for fresh starts), use the --initial-sync flag.

⚠️ WARNING: This will delete files on the remote directory that are not present locally.

syncmux --local-dir ~/myproject --remote-host dev-server --remote-dir /home/user/myproject --initial-sync

Command Line Reference

usage: syncmux [-h] [--version] --local-dir LOCAL_DIR --remote-host REMOTE_HOST
               --remote-dir REMOTE_DIR [--debounce DEBOUNCE]
               [--batch-threshold BATCH_THRESHOLD]
               [--keepalive-interval KEEPALIVE_INTERVAL]
               [--keepalive-count KEEPALIVE_COUNT]
               [--connection-check-interval CONNECTION_CHECK_INTERVAL]
               [--reconnect-attempts RECONNECT_ATTEMPTS]
               [--reconnect-delay RECONNECT_DELAY] [--initial-sync]

Required arguments:
  --local-dir DIR       Local directory to watch
  --remote-host HOST    Remote SSH host (as defined in ~/.ssh/config or user@host)
  --remote-dir DIR      Remote directory path

Sync behavior:
  --debounce SEC        Seconds to wait after last change before syncing (default: 2.0)
  --batch-threshold N   Number of files to trigger batch sync mode (default: 5)

Connection tuning:
  --keepalive-interval N  SSH keepalive interval in seconds (default: 15)
  --reconnect-attempts N  Number of reconnection attempts (default: 5)

How It Works

Persistent Connection

syncmux sets up a master SSH connection (ssh -M) with ControlPersist. All subsequent sync operations (rsync) reuse this socket. This drastically reduces latency, especially on high-latency networks (like VPNs), as it avoids the SSH handshake overhead for transfer.

Logging and Diffs

syncmux maintains logs on both ends so you always know what happened.

Local Log (.syncmux-log in your local dir): Records every file synced along with the local git diff at the time of sync. This allows you to audit exactly what code changes were pushed.

============================================================
Timestamp: 2024-03-20 10:30:01
Synced: src/main.py
Remote: dev-server:/app/src/main.py
----------------------------------------
Diff:
diff --git a/src/main.py b/src/main.py
index ...
...

Remote Log (.sync-log in your remote dir): Records incoming syncs. If the remote directory is a git repo, it also attempts to capture the diff of the applied changes relative to the remote HEAD.

Contributing

See CONTRIBUTING.md for details on how to contribute to syncmux.

License

MIT License - see LICENSE file