This topic describes the specifications of standard hybrid-storage instances of ApsaraDB for Redis Enhanced Edition (Tair), such as the memory capacity, maximum number of concurrent connections to each instance, maximum internal bandwidth, and queries per second (QPS) reference value.

Instance specifications

Note The maximum internal bandwidth is applicable to both upstream bandwidth and downstream bandwidth. If network resources are sufficient, the bandwidth is unlimited for ApsaraDB for Redis instances. However, if network resources are insufficient, the limit on maximum internal bandwidth takes effect for the instances.
Instance type InstanceClass (used in API operations) Number of new connections per second Maximum number of concurrent connections Maximum internal bandwidth (MB/s) QPS reference value Description
Master-replica instance with 16 GB memory and 32 GB disk storage redis.amber.master.16g.2x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 16 GB memory and 64 GB disk storage redis.amber.master.16g.4x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 16 GB memory and 128 GB disk storage redis.amber.master.16g.8x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 32 GB memory and 64 GB disk storage redis.amber.master.32g.2x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 32 GB memory and 128 GB disk storage redis.amber.master.32g.4x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 32 GB memory and 256 GB disk storage redis.amber.master.32g.8x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 64 GB memory and 128 GB disk storage redis.amber.master.64g.2x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 64 GB memory and 256 GB disk storage redis.amber.master.64g.4x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance
Master-replica instance with 64 GB memory and 512 GB disk storage redis.amber.master.64g.8x.ext4.default 10,000 50,000 48 40,000 Master-replica hybrid-storage instance

Bandwidth calculation rules

  • If network resources are sufficient, Apsaradb for Redis instances do not have bandwidth limits. However, if network resources are insufficient, the limits in the above table take effect.
  • The bandwidth value in the table represents the bandwidth of the instance, that is, the total bandwidth of all shards or nodes in the instance.
  • The maximum bandwidth of an instance (including all shards or nodes) is 2,048 MB. Bandwidth does not increase even if the number of shards or nodes increases.
  • The bandwidth applies to the upstream bandwidth and the downstream bandwidth respectively. If the bandwidth of an instance is 10 MB/s, the upstream and downstream bandwidth of the instance are both 10 MB/s.
  • The bandwidth in the table is the internal bandwidth of the Redis instance. The Internet bandwidth is determined by the internal bandwidth and is limited by the bandwidth of the connection between the Redis instance and the client. We recommend that you connect to the instance over an internal network to maximize its performance.

Connection number calculation rules

  • The following example explains the maximum number of new connections per second: If the maximum number of new connections per second of an instance is 10000, the actual number of concurrent connections at the Nth second after the instance starts to run is 12000, at the (N+1)th second, the actual number of concurrent connections is 22000, then the number of new connections within that second is 22000-12000=10000, which is not greater than the maximum number of new connections per second.
  • The maximum number of new connections per second and the maximum number of concorrent connections are the sum of the corresponding values of all shards or nodes in the instance. After 500000 is reached, the maximum number of concorrent connections is not increased even if the number of shards or nodes is increased.