Skip to main content

Test Patterns for Chunks Endpoint

Background: For more information about the Memfault Data pipeline in general see this tutorial.

In the following tutorial we will provide several test patterns which can be used to check that the data pipeline is hooked up correctly without needing to integrate any parts of the Memfault Firmware SDK.

Posting a Chunk

In this article we will explore sending a heartbeat event for a device with serial number TESTSERIAL.

Normally, the Memfault Firmware SDK would produce events (and other kinds of data) and packetize the data into one or more "chunks". Each chunk needs to be uploaded the chunks HTTP endpoint.

Below we will list some example calls to the chunks endpoint as a curl command but they can be converted to whatever format is easiest for the your setup.

We also provide an iOS library, MemfaultCloud, to facilitate posting chunks to Memfault from an iOS app.

Project Key

A key will be needed in order to communicate with Memfault's web services. Go to, navigate to the project you want to use and select 'Settings'→'General'. The 'Project Key' listed is what you will want to use.

Note you will have to replace <YOUR_PROJECT_KEY> with the one you grabbed in the setup instructions above for the commands below to work.

Event message encoded in a single chunk

Download the single chunk test payload and save it in your working directory. This file contains the data of a single chunk, containing the complete test event.

Run this curl command to post the chunk:

# Chunk 1/1
curl -v -X POST \
-H 'Memfault-Project-Key: <YOUR_PROJECT_KEY>' \
-H 'Content-Type: application/octet-stream' \
--data-binary "@chunk_v2_single_chunk_msg.bin"

The HTTP request that curl sent, looks more or less like this:

POST /api/v0/chunks/TESTSERIAL HTTP/1.1
Memfault-Project-Key: <YOUR_PROJECT_KEY>
Content-Type: application/octet-stream
Content-Length: <length of binary chunk data in bytes>

<binary chunk data here>

Same event message encoded in two chunks

In this example, we batch-upload 2 chunks in a single HTTP request.

This achieves better throughput compared to uploading them in separate requests (Note: it's not allowed to parallelize uploads of chunks from the same device).

Download chunk test payload part 1 of 2 and chunk test payload part 2 of 2 and save them in your working directory.

Run this curl command to post both chunks in one request:

# Event message spanning across two individual chunks
# Note: Event will appear in UI once both chunks have arrived

curl -v -X POST \
-H 'Memfault-Project-Key: <YOUR_PROJECT_KEY>' \
-H 'Content-Type: multipart/mixed' \
--form 'file=@chunk_v2_chunk_msg_pt1_of_2.bin;headers="Content-Length: 46"' \
--form 'file=@chunk_v2_chunk_msg_pt2_of_2.bin;headers="Content-Length: 44"'

The HTTP request that curl sent, looks more or less like this:

POST /api/v0/chunks/TESTSERIAL HTTP/1.1
Memfault-Project-Key: <YOUR_PROJECT_KEY>
Content-Length: <total length>
Content-Type: multipart/mixed; boundary=77AC153B-FCD2-4224-A8F8-1788947663A0
Content-Length: <chunk 1 of 2 binary data length in bytes>

<chunk 1 of 2 binary data>
Content-Length: <chunk 2 of 2 binary data length in bytes>

<chunk 2 of 2 binary data>

The curl commands inserts an additional Content-Disposition header with each part. These headers are not necessary and will get ignored.

Verifying arrival of the test events

Upon success you will be able to see the event contained inside the chunk(s) by going to and selecting 'Event'. It will look like this:



When a request to the chunks HTTP endpoint is responded to with 202 Accepted, it means the posted chunks have been enqueued for processing. It is possible for failures to occur later on in the post-processing phase. To troubleshoot these failures, the Integration Hub / Chunks Debug Log can be used.

This view lists recently received chunks and any associated post-processing errors:


You can reach this view by going to and selecting 'Chunks Debug' from the menu. If this option is missing in your project, please contact us to get it enabled.

Further Reading

We recommend taking a look at the detailed documentation of the chunks endpoint. To learn more about the semantics of the API and error codes refer to the chunks endpoint API documentation