The public preview of PolarDB for PostgreSQL compatibility with PostgreSQL 14 has ended. The official version of PolarDB for PostgreSQL compatibility with PostgreSQL 14 has been released on October 17, 2022. To apply for a trial, click here.

End date of public preview

September 30, 2022

Official release date

October 17, 2022

Note
You can apply for a free trial of PolarDB for PostgreSQL compatibility with PostgreSQL 14 in the first month. After your application is approved, you can use PolarDB for PostgreSQL compatibility with PostgreSQL 14 free of charge for one month. To apply for a free trial, click here.

Billing methods

PolarDB for PostgreSQL compatibility with PostgreSQL 14 supports the subscription and pay-as-you-go billing methods. For more information about the pricing of PolarDB for PostgreSQL compatibility with PostgreSQL 14, see Billable items.

Features

PolarDB for PostgreSQL is compatible with PostgreSQL 14. In addition to using distributed loads, PostgreSQL 14 further improves the performance of high concurrency, high write loads, parallel queries, and logical replication.

The new and optimized features of PostgreSQL 14 include:
  • Optimized high concurrency of a large number of connections
  • Enhanced indexing
  • Enhanced parallel computing
  • Improved performance of partitioned tables
  • Optimized IN queries of large tables
  • LZ4 compression algorithm
  • Enhanced security

Optimized high concurrency of a large number of connections

In case of a large number of connections, PostgreSQL 14 significantly improves the throughput of transactions. In test scenarios, the throughput can be doubled.

Enhanced indexing

PostgreSQL 14 reduces the resource consumption of B-tree indexes, including table bloats caused by frequent index updates. The generalized search tree (GiST) method can be used to pre-sort indexes when they are created. The space-partitioned generalized search tree (SP-GiST) method can significantly improve the speed to create indexes and reduce their size.

Enhanced parallel computing

PostgreSQL 14 enhances parallel query features. RETURN QUERY in PL/pgSQL functions supports parallel computing of query results. REFRESH MATERIALIZED VIEW supports parallel queries. Parallel sequential scan can be performed on chunks, which improves the scan throughput of large-scale data and approaches the throughput limit for block devices.

Improved performance of partitioned tables

PostgreSQL 14 improves partition pruning and reduces the number of subplans and cached plans of subpartitions. You can execute the ALTER TABLE DETACH|ATTACH PARTITION CONCURRENTLY statement to avoid lock conflicts when you add or remove partitions.

Optimized IN queries of large tables

PostgreSQL 14 supports linear search TO hash table probe to improve the processing performance of IN statements. The performance significantly improves when multiple conditions are specified or a large amount of data is filtered out.

LZ4 compression algorithm

PostgreSQL 14 supports the LZ4 algorithm to compress data columns. LZ4 improves compression performance and supports the original PGLZ compression method.

Enhanced security

PostgreSQL 14 can grant the read-only and write-only permissions by using two predefined roles of pg_read_all_data and pg_write_all_data. PostgreSQL 14 can also grant read and write permissions on tables, views, and sequences. By default, the SCRAM-SHA-256 algorithm is used for passwords.

References

Create a cluster