Hybrid-storage instances are no longer available for purchase. For details, see End-of-sale for Tair hybrid-storage instances. To meet similar workload requirements, use persistent memory-optimized instances.
This page lists the specifications of Tair (Enterprise Edition) hybrid-storage cluster instances, including memory and disk capacity, CPU cores, shard count, maximum connections, bandwidth, and queries per second (QPS) reference values.
Instance specifications
| Specification | InstanceClass (for API use) | CPU cores | Shards | Max connections (direct connection mode) | Max connections (proxy mode) | Bandwidth (Mbps) | QPS reference value |
|---|---|---|---|---|---|---|---|
| 64 GB memory, 256 GB disk capacity (4 shards) | redis.amber.sharding.16g.4db.0rodb.12proxy.4x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 64 GB memory, 512 GB disk capacity (4 shards) | redis.amber.sharding.16g.4db.0rodb.12proxy.8x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 128 GB memory, 512 GB disk capacity (4 shards) | redis.amber.sharding.32g.4db.0rodb.12proxy.4x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 128 GB memory, 1,024 GB disk capacity (4 shards) | redis.amber.sharding.32g.4db.0rodb.12proxy.8x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 256 GB memory, 1,024 GB disk capacity (4 shards) | redis.amber.sharding.64g.4db.0rodb.12proxy.4x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 256 GB memory, 2,048 GB disk capacity (4 shards) | redis.amber.sharding.64g.4db.0rodb.12proxy.8x.ext4.default | 24 | 4 | 120,000 | 500,000 | 384 | 120,000 |
| 128 GB memory, 512 GB disk capacity (8 shards) | redis.amber.sharding.16g.8db.0rodb.24proxy.4x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 128 GB memory, 1,024 GB disk capacity (8 shards) | redis.amber.sharding.16g.8db.0rodb.24proxy.8x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 256 GB memory, 1,024 GB disk capacity (8 shards) | redis.amber.sharding.32g.8db.0rodb.24proxy.4x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 256 GB memory, 2,048 GB disk capacity (8 shards) | redis.amber.sharding.32g.8db.0rodb.24proxy.8x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 512 GB memory, 2,048 GB disk capacity (8 shards) | redis.amber.sharding.64g.8db.0rodb.24proxy.4x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 512 GB memory, 4,096 GB disk capacity (8 shards) | redis.amber.sharding.64g.8db.0rodb.24proxy.8x.ext4.default | 48 | 8 | 240,000 | 500,000 | 768 | 240,000 |
| 256 GB memory, 1,024 GB disk capacity (16 shards) | redis.amber.sharding.16g.16db.0rodb.48proxy.4x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
| 256 GB memory, 2,048 GB disk capacity (16 shards) | redis.amber.sharding.16g.16db.0rodb.48proxy.8x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
| 512 GB memory, 2,048 GB disk capacity (16 shards) | redis.amber.sharding.32g.16db.0rodb.48proxy.4x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
| 512 GB memory, 4,096 GB disk capacity (16 shards) | redis.amber.sharding.32g.16db.0rodb.48proxy.8x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
| 1,024 GB memory, 4,096 GB disk capacity (16 shards) | redis.amber.sharding.64g.16db.0rodb.48proxy.4x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
| 1,024 GB memory, 8,192 GB disk capacity (16 shards) | redis.amber.sharding.64g.16db.0rodb.48proxy.8x.ext4.default | 96 | 16 | 480,000 | 500,000 | 1,536 | 480,000 |
Table notes:
Bandwidth: The value listed is the total instance bandwidth — the combined bandwidth of all shards. Each shard supports a maximum of 768 Mbps (96 MB/s). Bandwidth is symmetric: upstream and downstream limits are both equal to the listed value. The value reflects internal bandwidth; public network bandwidth is further limited by the connection between the instance and the client. For best performance, connect over an internal network.
Max connections (proxy mode): 500,000 is an instance-wide ceiling. Adding more shards does not increase this limit. For instances created before March 1, 2020 in proxy mode, the ceiling is 200,000.
Max connections (direct connection mode): Equals the per-shard limit multiplied by the number of shards. The per-shard limit for Tair (Enterprise Edition) is 30,000.
QPS reference value: A reference figure based on standard test conditions. Actual QPS varies with workload characteristics.
Limitations
| Limitation | Details |
|---|---|
| Proxy endpoint bandwidth ceiling | When using the default proxy endpoint, the maximum instance bandwidth is capped at 16,384 Mbps (2,048 MB/s), regardless of shard count. |
| CPU core reservation | One CPU core per shard (or per read replica in a read/write splitting instance) is reserved for background tasks. This reservation is factored into the CPU core counts shown in the table. |
| Direct connection mode scope | Applicable to cluster instances only. In direct connection mode, maximum connections and bandwidth scale linearly with shard count: max connections = 30,000 x number of shards; max bandwidth = 96 MB/s x number of shards. For example, a 128-shard instance reaches 12,288 MB/s. Note that connection performance may degrade slightly because traffic bypasses proxy nodes. |
| New connections per second | Per shard: 2,000. For a 4-shard instance, the instance-level limit is 8,000 new connections per second. |
What's next
Enable the direct connection mode to scale bandwidth and connections beyond proxy mode limits.
Adjust the bandwidth of an instance to handle unexpected or scheduled traffic peaks.
Persistent memory-optimized instances — the recommended alternative to hybrid-storage cluster instances.