This topic describes the specifications of hybrid-storage cluster instances of ApsaraDB for Redis Enhanced Edition (Tair). The specifications include the memory and disk capacities, maximum number of concurrent connections to each instance, maximum bandwidth value, and queries per second (QPS) reference value.
|Instance type||InstanceClass that is used in API operations||Number of CPU cores||Number of shards||Maximum number of new connections per second||Maximum number of connections ( (in direct connect mode) )||Maximum number of connections ( (in proxy mode) )||Maximum bandwidth value (MB/s)||QPS reference value|
|64 GB memory and 256 GB disk storage (4 shards)||redis.amber.sharding.16g.4db.0rodb.12proxy.4x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|64 GB memory and 512 GB disk storage (4 shards)||redis.amber.sharding.16g.4db.0rodb.12proxy.8x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|128 GB memory and 512 GB disk storage (4 shards)||redis.amber.sharding.32g.4db.0rodb.12proxy.4x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|128 GB memory and 1,024 GB disk storage (4 shards)||redis.amber.sharding.32g.4db.0rodb.12proxy.8x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|256 GB memory and 1,024 GB disk storage (4 shards)||redis.amber.sharding.64g.4db.0rodb.12proxy.4x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|256 GB memory and 2,048 GB disk storage (4 shards)||redis.amber.sharding.64g.4db.0rodb.12proxy.8x.ext4.default||24||4||40,000||120,000||500,000||384||120,000|
|128 GB memory and 512 GB disk storage (8 shards)||redis.amber.sharding.16g.8db.0rodb.24proxy.4x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|128 GB memory and 1,024 GB disk storage (8 shards)||redis.amber.sharding.16g.8db.0rodb.24proxy.8x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|256 GB memory and 1,024 GB disk storage (8 shards)||redis.amber.sharding.32g.8db.0rodb.24proxy.4x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|256 GB memory and 2,048 GB disk storage (8 shards)||redis.amber.sharding.32g.8db.0rodb.24proxy.8x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|512 GB memory and 2,048 GB disk storage (8 shards)||redis.amber.sharding.64g.8db.0rodb.24proxy.4x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|512 GB memory and 4,096 GB disk storage (8 shards)||redis.amber.sharding.64g.8db.0rodb.24proxy.8x.ext4.default||48||8||50,000||240,000||500,000||768||240,000|
|256 GB memory and 1,024 GB disk storage (16 shards)||redis.amber.sharding.16g.16db.0rodb.48proxy.4x.ext4.default||96||16||50,000||480,000||500,000||1,536||480,000|
|256 GB memory and 2,048 GB disk storage (16 shards)||redis.amber.sharding.16g.16db.0rodb.48proxy.8x.ext4.default||96||16||50,000||480,000||500,000||1,536||480,000|
|512 GB memory and 2,048 GB disk storage (16 shards)||redis.amber.sharding.32g.16db.0rodb.48proxy.4x.ext4.default||96||16||50,000||480,000||500,000||1,536||480,000|
|512 GB memory and 4,096 GB disk storage (16 shards)||redis.amber.sharding.32g.16db.0rodb.48proxy.8x.ext4.default||96||16||50,000||480,000||500,000||1,536||480,000|
|1,024 GB memory and 4,096 GB disk storage (16 shards)||redis.amber.sharding.64g.16db.0rodb.48proxy.4x.ext4.default||96||16||50,000||480,000||500,000||1,536||480,000|
|1,024 GB memory and 8,192 GB disk storage (16 shards)||redis.amber.sharding.64g.16db.0rodb.48proxy.8x.ext4.default||96||16||50,000||480,000||500,000||1536||480,000|
Calculation rules for bandwidth values
- The bandwidth in the preceding tables refers to the bandwidth of the instance. The
bandwidth value is the sum of the bandwidth of all shards or nodes on the instance.
Each shard has the following bandwidth limits:
- Community Edition: If the memory of each shard is less than or equal to 1 GB, the maximum bandwidth of each shard is 48 MB/s. If the memory of each shard is greater than 1 GB, the maximum bandwidth of each shard is 96 MB/s.
- Enhanced Edition (Tair): The maximum bandwidth of each shard is 96 MB/s.
- When the default proxy endpoint is used by a cluster instance, the upper limit of
bandwidth for the cluster instance is 2,048 MB/s. After the upper limit is reached,
the bandwidth cannot be increased even if you add more shards.
Note To handle excessive network traffic, you can enable the direct connect mode. For more information, see Enable the direct connection mode. The direct connect mode is applicable only to cluster instances. In direct connect mode, the upper limit of the bandwidth is equal to the maximum bandwidth of a single shard multiplied by the number of shards. For example, for a cluster instance that contains 128 shards, in which each shard is allocated with a memory of more than 1 GB, the bandwidth limit is 12,288 MB/s.
- The bandwidth value applies to the upstream bandwidth and the downstream bandwidth. For example, 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 preceding 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 use an internal network to maximize the performance of the instance.
Number of connections
|Maximum number of concurrent connections||In proxy mode, the maximum number of concurrent connections of an instance is 500,000. If the connections reach the upper limit, subsequent connections cannot be established even if you add more shards or nodes.|
|New connections per second||The following example shows the maximum number of new connections that can be created
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, which is calculated by using the formula 12,000
Note If the instance is connected in direct connect mode, connection performance degrades because the proxy node is bypassed. The maximum number of new connections per second for a single data shard is 2,000. For example, if an instance has four data shards, the maximum number of new connections per second is 8,000.