MQTT Retained Messages: Beginner's Guide with Example
Table of Contents
What is Retained Messages?
If you know MQTT even for just a little bit, you may already know that for each MQTT message, there is a topic name, and there is the payload. If you dig a little deeper, you’ll find that there are also message properties and flags. One of the flags is called Retain
, which is what this post is about.
Upon receiving a message with the Retain
flag set, the MQTT broker must store the message for the topic to which the message was published, and it must store only the latest message. So the subscribers which are interested in this topic can go offline, and reconnect at any time to receive the latest message instead of having to wait for the next message from the publisher after the subscription.
As illustrated below, when a client subscribes to a topic, if there is a retained message for this topic, the message is sent to the client immediately.
When to use MQTT Retained Messages?
While allowing publishers to decouple subscribers, the publish-subscribe pattern also has the disadvantage that subscribers cannot actively fetch messages from publishers. When a subscriber receives a message depends on when the publisher publishes it, which is inconvenient in some scenarios.
New subscribers can get the latest data immediately without waiting for unpredictable times with retained messages. Below are some examples:
- Smart home devices only send state data when the state changes, but the control APP needs to know the device's current state whenever the user opens the APP.
- The interval between sensors reporting data can be very long, but subscribers may need to get the latest data immediately after subscribing.
- Properties such as sensor version and serial number that do not change frequently can be published as a retained message for later subscribers to get the information.
How to use MQTT Retained Messages?
For MQTT client SDKs, there are typically APIs or parameters to set the Retain
flag. For example the paho MQTT Java client library, and the the Erlang MQTT client emqtt.
For MQTT client tools, either with a command line or graphic interface, you should be able to find where to set the Retain
flag.
In this post, we are not going to dig into the programming SDKs. We will try to demonstrate MQTT retained messages using the open-source cross-platform MQTT 5.0 desktop client - MQTTX.
If you start the MQTTX application for the first time, you will see the main window below. Click the New Connection
button to create an MQTT connection.
We only need to fill in a connection Name
and leave the other parameters as default. The Host
will default to the public MQTT Broker provided by EMQ. Finally, click the Connect
button in the upper right corner to create an MQTT connection.
After the successful connection, publish a message to the topic sensor/t1
in the message input box.
Next, we check the Retain flag and publish two retained messages to the topic sensor/t2
.
Then click the New Subscription
button to create a subscription.
We subscribe to the wildcard topic sensor/+,
which will match the topics sensor/t1
and sensor/t2
.
Check out the blog Understanding MQTT Topics & Wildcards by Case for more details.
Finally, we will see that the subscription successfully receives the second retained message, neither the normal message for sensor/t1
nor the first retained message for sensor/t2
. This shows that the MQTT Broker will only store the latest retained message for each topic.
Q & A about MQTT Retained Messages
How do I know an MQTT message is a retained message?
When a message is originated from the Retain
storage in the broker, the Retain
flag is set, so the subscriber knows that this is not a new message after its subscription.
That is, if a retained message is published after the subscription, the subscriber will receive it as a regular message (without the Retain
flag). After a retained message is delivered, if the subscriber wishes to receive the retained message again, it needs to resubscribe.
In the example below, we subscribe to the topic sensor/t2
and then publish a retained message to the topic, the subscriber receives the message immediately, but without the ‘retain’ flag. Then we delete the subscription and re-subscribe to sensor/t2
to receive the message again with the ‘retain’ flag set.
How long are retained messages stored? How to delete it?
The broker will only store the latest retained message for each topic, and the validity of the retained message is related to the broker's settings. If the broker is set to store retained messages in memory, they are lost when the MQTT Broker is restarted; if they are stored on disk, they remain after the broker is restarted.
Retained messages are not part of session states, meaning retained messages are not deleted when the publishing session terminates. There are several ways to delete retained messages.
- When a client publishes a retained message with an empty payload to a topic, the broker deletes the retained message under that topic.
- Delete on the MQTT Broker, e.g., the EMQX MQTT Broker provides the ability to delete retained messages from management API or from the Dashboard.
- MQTT 5.0 protocol added Message Expiry Interval property, which can be used to set the expiration time of the message when publishing. The message will be automatically deleted after the expiration time, regardless of whether it is a retained message.
MQTT Retained Messages in EMQX
EMQX is the world's most scalable MQTT broker that supports advanced features such as MQTT 5.0, MQTT-SN, and MQTT over QUIC. It supports masterless clustering for high availability and horizontal scalability.
EMQX supports viewing and setting retained messages in the built-in Dashboard. You may use the following command to install EMQX open-source version for trial.
docker run -d --name emqx -p 1883:1883 -p 8083:8083 -p 8084:8084 -p 8883:8883 -p 18083:18083 emqx/emqx:latest
After successful installation, use your browser to visit http://127.0.0.1:18083/
to experience the new EMQX Dashboard.
The default Username is admin, and the Password is public.
After successful login, you can click the Configuration->MQTT
menu to view the list of retained messages. You can also view the Payload of retained messages or delete a retained message.
Click on the Settings
menu under Retainer
, and you will see that EMQX supports setting the Storage (memory or disk), the Max Retained Messages, the Expire, and other parameters in the Dashboard.
Summary
This article has provided a comprehensive overview of MQTT Retained Messages, demonstrating their practical applications and benefits. By leveraging Retained Messages, users can significantly enhance their MQTT communication, enabling immediate data retrieval upon subscription.
While Retained Messages are a powerful feature, it's important to remember that MQTT offers a wealth of other valuable capabilities. To deepen your understanding and explore more advanced MQTT applications, we encourage you to check out EMQ's MQTT Guide series. These resources will provide you with the knowledge and tools needed to develop sophisticated MQTT applications and services.
Related Resources
- Introduction to MQTT Publish-Subscribe Pattern
- How to Set Parameters When Establishing an MQTT Connection?
- MQTT Topics and Wildcards: A Beginner's Guide
- MQTT Persistent Session and Clean Session Explained
- MQTT QoS 0, 1, 2 Explained: A Quickstart Guide
- MQTT Will Message (Last Will & Testament) Explained and Example
- What is the MQTT Keep Alive parameter for?