Scenario

In the following scenarios, you can consider deleting messages from Kafka:

l Error messages: If there are errors or exceptions in the messages in Kafka, you need to delete them. This can be due to a malformatted message, incomplete message content, or other data quality issues. Deleting error messages keeps your data consistent and accurate.

l Retention policy change: You can set the retention time or size of messages in a topic. When you need to change the retention policy for a message, you may need to delete the old messages to apply the new policy. For example, if you want to shorten the retention period of messages, you need to delete old messages with a retention period longer than the new one.

l Test data cleanup: In a test environment, it is often necessary to clean up old test data to ensure that the environment has free space with good performance. When the test data is no longer needed, the related messages can be deleted to free up resources.

l Recovery from faults: If Kafka fails or data is lost, you may need to delete some messages to restore data. In this case, you need to choose to delete the corresponding messages based on the scope and impact of the fault to restore the data.

It should be noted that before deleting messages, you need to make sure that the messages are no longer needed or have been backed up. Deleting a message is irreversible and once a message is deleted, it cannot be restored. Therefore, it is recommended to make a backup or confirm that the message is no longer needed before performing the deletion.

Procedure

(1) Log in to the management console.

(2) Go to the Kafka console.

(3) In the operation column of the instance list page, click Manage in the target instance row.

(4) Click Topic Management to enter the topic management page, and then click More, and the pop-up window as shown below will appear.

(5) In the row of the topic, click Delete Message and select OK.

Distributed Message Service (Kafka)