The main specification metrics for AnalyticDB for MySQL node groups are CPU and memory. Different types of node groups support different disk space sizes. You can purchase disk space based on actual usage.
AnalyticDB for MySQL stores data in three replicas. Each node group consists of three online nodes (replicas). This improves reliability and query concurrency compared with the primary and secondary nodes or dual replicas. The three online nodes in each node group have individual specifications. For example, if you purchase a C8 node group, the three online nodes in the node group are all 8-core 64 GB nodes. In terms of pricing, you can pay for only two nodes when you purchase a node group.
AnalyticDB for MySQL offers two types of node groups: compute-intensive node groups and storage-intensive node groups.
- Compute-intensive node groups (C series): All the data of these node groups is stored in standard SSDs.
Compute-intensive node groups are ideal for services that require high performance and query concurrency.
- Storage-intensive node groups (S series): These node groups use an architecture that separates computing resources from storage
resources. All the data of these node groups is stored in HDDs.
Storage-intensive node groups are ideal for scenarios where query concurrency and performance are not highly required and a query response of more than 10 seconds is allowed.
|vCPUs||Memory (GB)||Disk space (GB)|
|Cluster Edition (compute-intensive)||C8||8||64||100 to 1,000||Allows you to scale out nodes and scale up storage space.||Provides 100 times the performance of MySQL databases and offers linear scalability.|
|C24||24||96||100 to 1,000|
|C4||4||32||100 to 200||Allows you to scale out nodes.|
|Cluster Edition (storage-intensive)||S8||8||64||1,000 to 12,000||Allows you to scale out nodes and scale up storage space.||Provides slightly lower performance compared with compute-intensive node groups and offers linear scalability.|
- The disk space displayed on the buy page is the space size of each node group. The total disk space of a cluster equals the space size multiplied by the number of node groups.
- The total number of scale-in, scale-out, upgrade, and downgrade operations on each database cannot exceed 12 per day.