All the nodes of PolarDB for PostgreSQL are dedicated nodes. The CPU, memory, storage, and I/O resources allocated to dedicated nodes are not shared by other nodes. Therefore, the performance of dedicated nodes is more stable and reliable.
PolarDB for PostgreSQL provides nodes of the following specifications.
| Node type | CPU and memory | Maximum storage capacity | Maximum number of connections 1 | Internal bandwidth | Maximum IOPS | I/O bandwidth |
|---|---|---|---|---|---|---|
| polar.pg.x4.medium | 2 cores 8 GB | 5 TB | 800 | 1 Gbps | 16,000 | 1 Gbps |
| polar.pg.x4.large | 4 cores 16 GB | 10 TB | 1,600 | 10 Gbps | 64,000 | 4 Gbps |
| polar.pg.x4.xlarge | 8 cores 32 GB | 10 TB | 3,200 | 10 Gbps | 128,000 | 8 Gbps |
| polar.pg.x8.xlarge | 8 cores 64 GB | 30 TB | 3,200 | 10 Gbps | 160,000 | 10 Gbps |
| polar.pg.x8.2xlarge | 16 cores 128 GB | 50 TB | 12,800 | 10 Gbps | 256,000 | 16 Gbps |
| polar.pg.x8.4xlarge | 32 cores 256 GB | 50 TB | 25,600 | 10 Gbps | 384,000 | 24 Gbps |
| polar.pg.x8.8xlarge | 64 cores 512 GB | 100 TB | 36,000 | 10 Gbps | 409,600 | 24 Gbps |
| polar.pg.x8.12xlarge | 88 cores 710 GB | 100 TB | 36,000 | 25 Gbps | 512,000 | 32 Gbps |
- Maximum number of connections 1: the limit on the number of connections, which is indicated by the value of the max_connections parameter. The actual number of connections supported by a compute node may vary based on your business. For more information, see Maximum number of connections.
- In the minor version 1.1.7 that was released in December 2020, the maximum numbers of connections supported by some compute node specifications were changed. The latest specifications are listed in the preceding table. The changes apply to all clusters that are created after the minor version 1.1.7 was released. For existing clusters, you can use one of the following methods to modify the maximum number of connections:
- Change the specifications of clusters.
- Go to the Quota Center. Click Apply in the Actions column corresponding to polardb_pg_max_connections.
| Node type | CPU and memory | Maximum storage capacity | Internal bandwidth | Maximum IOPS | I/O bandwidth |
|---|---|---|---|---|---|
| polar.pg.x4.medium | 2 cores 8 GB | 5 TB | 1 Gbps | 16,000 | 1 Gbps |
| polar.pg.x4.large | 4 cores 16 GB | 10 TB | 10 Gbps | 64,000 | 4 Gbps |
| polar.pg.x4.xlarge | 8 cores 32 GB | 10 TB | 10 Gbps | 128,000 | 8 Gbps |
| polar.pg.x8.xlarge | 8 cores 64 GB | 30 TB | 10 Gbps | 160,000 | 10 Gbps |
| polar.pg.x8.2xlarge | 16 cores 128 GB | 50 TB | 10 Gbps | 256,000 | 16 Gbps |
| polar.pg.x8.4xlarge | 32 cores 256 GB | 50 TB | 10 Gbps | 384,000 | 24 Gbps |
| polar.pg.x8.8xlarge | 64 cores 512 GB | 100 TB | 10 Gbps | 409,600 | 24 Gbps |
| polar.pg.x8.12xlarge | 88 cores 710 GB | 100 TB | 25 Gbps | 512,000 | 32 Gbps |
Maximum number of connections
- If the number of concurrent connections in your PolarDB for PostgreSQL cluster exceeds the maximum number, new connections cannot be established, or existing connections will time out.
The amount of memory consumed by each connection varies in different business systems. As a result, the actual number of connections supported by your PolarDB cluster may be different from the maximum number in the specifications.
- Query the maximum number of connections under the current system configurations.
show max_connections; - Query the current number of connections.
select count(1) from pg_stat_activity; - We recommend that you monitor the status of your cluster and make sure that the number of connections does not exceed the recommended value. You can calculate the recommended value based on following formula:
LEAST({DBInstanceClassMemory/11MB}, 5000) - If your application requires a larger number of connections than the maximum number, you must use a PolarDB cluster with larger memory.