Memfault is a full device monitoring solution which can be integrated into any embedded device utilizing ARM Cortex-M or ESP32 MCUs irrespective of the RTOS in use or connectivity path in place.
How does it work?
- Using the lightweight memfault-firmware-sdk, full device coredumps, backtraces, debug registers, logs, and custom metrics can be collected.
- Data structures collected are framed into packets that can be as small as 9 bytes so they can be sent over whatever transport you use (e.g. Bluetooth, LTE, WiFi, LoRa, Zigbee, or even proprietary protocols).
- The Memfault cloud reassembles data structures from received packets and analyzes them to classify the issue, identify a possible root cause, and recover items such as variable and function names from the binary data.
- The Memfault cloud automatically de-duplicates similar issues. Notifications are used to let you know when problems are detected. From the Memfault web application, individual issues can be drilled into and overall fleet health can be examined via dashboard views.
The memfault-firmware-sdk is designed as a collection of components, so you can include only what is needed for your project.
We have step-by-step guides which cover how to add the SDK to a project, capture an error, and publish data to the Memfault cloud:
The following guides assume you already have a Memfault Account. If you don't yet have one, you can create one here.
|MCU Architecture||Getting Started Guide|
|ARM Cortex-M||ARM Cortex-M Integration Guide|
|nRF Connect SDK||nRF Connect SDK Integration Guide|
|ESP32 ESP-IDF||ESP32 ESP-IDF Integration Guide|
|ESP8266||ESP8266 RTOS Integration Guide|
|Dialog DA1469x||DA1469x Integration Guide|
|MCUXpresso RT1060||MCUXpresso SDK for i.MX RT Guide|
Once the memfault-firmware-sdk is in place, a number of modules can easily be integrated into the system:
- Coredumps - Allows one to collect the system state when the device hangs, crashes or asserts unexpectedly.
- Reboot Reason Tracking - allows one to collect why devices are rebooting in the field (whether it be due to a crash, brown out, crash or a firmware update). Summaries and aggregates and then be viewed from the Memfault web application.
- Device Metrics - A flexible way to collect system health vitals that are important to your application (whether it be battery life, connectivity statistics, performance measurements or something else!). From the Memfault web application, metrics are plotted over time for any device and fleet wide summaries can be viewed via dashboards.
- Error Tracking with Trace Events - allows one to collect lightweight events (<40 bytes) when error conditions occur. These events can be collected while the system is running and can be used to discover how often unexpected errors are occurring (flash write failures, I2C bus timeouts, etc).
The Memfault data collection system was designed from the ground up with embedded devices in mind. Deeper explorations into the designs of the subsystems can be found below: