Cold Data Preloading for Read Replicas

Scenarios

When a cluster TaurusDB instance is running, the primary node monitors the least recently used (LRU) linked list and synchronizes active data pages (pages read from storage or moved within a cache pool) to read replicas. The read replicas preload the pages to the cache pool to improve the cache hit ratio and reduce the performance jitter after a read replica is promoted to primary.

Figure 1 Diagram

Precautions

• To use this function, the kernel version of your TaurusDB instance must be 2.0.78.260602 or later. For details about how to check the kernel version, see How Can I Check the Version of a TaurusDB Instance?
• Data pages of both preloading and read services are cached in the cache pool. When preloading and read services are enabled at the same time for read replicas, the page hit ratio of read services may decrease.

Enabling Cold Data Preloading for Read Replicas

1. Log in to the TaurusDB console.
2. Click in the upper left corner and select a region.
3. On the Instances page, click the instance name.
4. In the navigation pane, choose Parameters.
5. Search for the innodb_rds_buf_preload_enable parameter and change its value to ON. The change is applied immediately. You do not need to reboot the instance.

   Table 1 Parameter description

   Parameter	Description
   innodb_rds_buf_preload_enable	Controls whether to enable cold data preloading for read replicas. The default value is OFF. ON: Cold data preloading for read replicas is enabled. OFF: Cold data preloading for read replicas is disabled.