This topic describes the specifications of read/write splitting instances of ApsaraDB for Redis Enhanced Edition (Tair). These specifications include the number of read-only replicas, memory capacity, maximum number of concurrent connections to each instance, bandwidth, and queries per second (QPS) reference value.

Instance specifications

Instance type InstanceClass (used in API operations) Number of I/O threads Number of read/write nodes Number of read-only replicas Bandwidth (MB/s) Maximum number of new connections per second Maximum number of concurrent connections QPS reference value
1 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.small.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
2 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.mid.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
4 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.stand.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
8 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.large.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
16 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.2xlarge.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
32 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.4xlarge.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
64 GB performance-enhanced read/write splitting instance (1 shard, 1 read-only replica) redis.amber.logic.splitrw.8xlarge.1db.1rodb.6proxy.multithread 4 1 1 192 20,000 60,000 480,000
1 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.small.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
2 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.mid.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
4 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.stand.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
8 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.large.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
16 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.2xlarge.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
32 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.4xlarge.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
64 GB performance-enhanced read/write splitting instance (1 shard, 3 read-only replicas) redis.amber.logic.splitrw.8xlarge.1db.3rodb.12proxy.multithread 4 1 3 384 40,000 120,000 960,000
1 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.small.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
2 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.mid.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
4 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.stand.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
8 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.large.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
16 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.2xlarge.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
32 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.4xlarge.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000
64 GB performance-enhanced read/write splitting instance (1 shard, 5 read-only replicas) redis.amber.logic.splitrw.8xlarge.1db.5rodb.18proxy.multithread 4 1 5 576 50,000 180,000 1,440,000

Calculation rules for bandwidth

  • If network resources are sufficient, the bandwidth is unlimited. However, if network resources are insufficient, the limit on maximum internal bandwidth takes effect.
  • The bandwidth value in the table refers to the bandwidth of the instance. It is the sum of bandwidth of all shards or nodes on the instance.
  • The upper limit of total bandwidth for a cluster or read/write splitting instance (as shown in the table) is 2,048 MB/s. After the upper limit is reached, the bandwidth cannot be increased even if the number of shards or nodes is increased.
  • The bandwidth applies to the upstream bandwidth and 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 listed in the table is the internal bandwidth of the ApsaraDB for Redis instance. The Internet bandwidth is determined by the internal bandwidth and is limited by the bandwidth of the connection between the instance and a client. We recommend that you connect to the instance over an internal network to maximize the performance.

Calculation rules for connections

  • The following example explains the maximum number of new connections that can be added per second: The maximum number of new connections per second of an instance is 10,000. The maximum number of connections is 50,000. The actual number of connections at the Nth second after the instance starts to run is 12,000. Then, at the (N+1)th second, the number of connections is 22,000 (12,000 + 10,000).
  • The maximum number of new connections per second and the maximum number of connections are the sum of the corresponding values of all shards or nodes on the instance. After 500,000 is reached, the maximum number of connections cannot be increased even if the number of shards or nodes is increased.