This topic describes the specifications of persistent memory-optimized instances of the ApsaraDB for Redis Enhanced Edition (Tair), including the instances in the standard or cluster architecture. These specifications include the memory capacity, maximum number of connections, and maximum bandwidth.
Instance purchase
Terms
- persistent memory-optimized instance
- Persistent memory-optimized instances adopt Intel Optane DC persistent memory (AEP) to support large-capacity in-memory databases that are compatible with open source Redis. A persistent memory-optimized instance persists each operation and does not use disks to implement data persistence. Compared with an ApsaraDB for Redis Community Edition instance, it reduces costs by up to 30% and delivers almost the same throughput and latency. This improves the reliability of business data. For more information, see Persistent memory-optimized instances.
- ESSD
- Enhanced SSDs (ESSDs) are disks provided by Alibaba Cloud. For more information about
the terms and performance levels of ESSDs, see ESSDs.
Note ESSDs are used only to store system operating data, such as logs, and backup data and are not used to read or write data.
Instance types
The following table lists the specifications of a single shard. The performance of a cluster instance can be calculated by using the following formula: Performance of a cluster instance = Number of shards in the instance × Performance of each shard specified by instance specifications.
Instance type | InstanceClass value (used in API operations) | Specification | Maximum number of connections | Maximum bandwidth value (Mbit/s) | QPS reference value | |
---|---|---|---|---|---|---|
Number of CPU cores | Persistent memory (GB) | |||||
Master-replica 4 GB persistent memory | tair.scm.standard.1m.4d | 3 | 4 | 10,000 | 96 | 100,000 |
Master-replica 8 GB persistent memory | tair.scm.standard.2m.8d | 3 | 8 | 10,000 | 96 | 100,000 |
Master-replica 16 GB persistent memory | tair.scm.standard.4m.16d | 3 | 16 | 10,000 | 96 | 100,000 |
Master-replica 32 GB persistent memory | tair.scm.standard.8m.32d | 3 | 32 | 10,000 | 96 | 100,000 |
Master-replica 64 GB persistent memory | tair.scm.standard.16m.64d | 3 | 64 | 10,000 | 96 | 100,000 |
Performance
The specifications described in this topic apply to persistent memory-optimized instances in the standard or cluster architecture. For more information about the standard architecture and the cluster architecture, see Standard master-replica instances and Cluster master-replica instances.
Architecture | Performance |
---|---|
Standard | The performance of standard instances is as described in the instance specifications table. |
Cluster | The performance of cluster instances is calculated by using the following formula:
Performance of a cluster instance = Number of shards in the instance × Performance
of each shard specified by instance specifications.
Note When a cluster instance that runs in proxy mode is accessed, the maximum bandwidth
is 2,048 Mbit/s, the maximum total QPS is 10,000,000, and the maximum number of connections
is 500,000. After one of the upper limits is reached, the instance performance cannot
be improved even if you add more shards to the cluster instance. To handle ultra-high
network traffic for your business, you can enable the direct connection mode. For
more information, see Enable the direct connection mode.
For example, a cluster instance has four shards of the tair.scm.standard.2m.8d type.
This type is the second instance type in Instance types. The following items show the performance of each shard:
The following items show the performance of the cluster instance:
|
Notes about CPU cores
To ensure service stability, the system reserves a CPU core to process background tasks. In a cluster instance or a read/write splitting instance, the system reserves a CPU core for each data shard or each read replica to process background tasks.