Message Queue for Apache Kafka supports two storage engines: cloud storage and local storage. This topic compares the two storage engines to help you select one based on your business requirements.
|Item||Cloud storage Recommended||Local storage|
|Underlying algorithm||Alibaba Cloud disk-related algorithm||Apache Kafka-native ISR algorithm|
|Requirements on instance versions||None||Professional edition of open-source v2.2.0|
|Replica mechanism||Distributed replica set with three members
||Distributed replica set with three members
If you purchase a disk of 300 GB in size, the actual storage space you can use for your business is 100 GB and the remaining 200 GB of storage space is used to store backups.
|Message type||Common message
A small amount of disorder may occur due to cluster restart or downtime.
|Partitionally ordered message
A new leader is automatically selected from the ISR if the cluster goes down.
- For more information about the open source versions of different instance specifications, see Supported editions.
- You can select a storage engine when you create a topic for an instance of the Professional Edition. The storage engine can be local storage or cloud storage. You cannot select a storage engine when you create a topic for an instance of the Standard Edition. For more information about how to select a storage engine for a topic, see Step 1: Create a topic.
- Cloud storage has the benefits of the underlying storage of Alibaba Cloud. Compared with local storage, cloud storage provides better performance in auto-scaling, reliability, availability, and cost-effectiveness. Therefore, we recommended that you use cloud storage in most cases.
- If you have special requirements, such as Compact, idempotence, transaction, and partitionally ordered messages, we recommend that you use local storage. These scenarios are rare.
- Local storage refers to the use of the native ISR algorithm, not local disks.