mpysync

Rsync-like tool for MicroPython.

• Works over serial or WiFi.
• Can be run (as an async task) in the background of your normal app (for remote updates).
• Generally 10-100x faster than using ampy.

Install

$ pip install mpysync

Performance

Take for example, a decently sized project (50 files, 836K) on an ESP32...

Copying all files with ampy:

$ time ampy --port /dev/ttyUSB0 put build/ui ui
real	8m33.000s

Vs. rsync-ing the same files (no changes):

$ time python -m mpysync build/ --host 10.0.0.179
real	0m1.511s

$ time python -m mpysync build/ --port /dev/ttyUSB0
real	0m6.858s

Worst case (no files in common), copying all files w/ mpysync is still an order of magnitude faster:

$ time python -m mpysync build/ --port /dev/ttyUSB0
real	0m59.650s

CLI Usage

Over Serial

python -m mpysync build/ --port /dev/ttyUSB0

Over WiFi

python -m mpysync build/ --host 10.0.0.179

All Options

python -m mpysync [--directory <str>] [--host <str>] [--port <value>] [--baud <int>] [--dry_run <bool>] [--clear_cache <bool>] [--verify <bool>]

OTA Usage

In your main.py or (or anywhere really):

import mpysync

Then (somewhere later in your code), start the main event loop:

import uasyncio as asyncio
asyncio.get_event_loop().run_forever()

By default it listens on port 31261.

Memory

When running in the background (for OTA updates), mpysync consumes 16k RAM if you're already using uttp, 20k otherwise.