Some open source goodies for PostgreSQL acceleration

Swarm64 develops high-performance extensions that accelerate free, open source PostgreSQL databases. Along the way, we have developed a few open source projects that help PostgreSQL users accelerate, scale, and measure PostgreSQL performance.

We invite you to try out the following offerings, and we always welcome questions from users and people interested in contributing. Contact us via Github or via our feedback page.

Elastic PostgreSQL scaling with parallel query processing

free open source patch to the PostgreSQL foreign data wrapper interface that enhances PostgreSQL with elastic scaling and faster query processing under concurrent user load.

  • Parallel query performance across shards. If you divide your data across separate PostgreSQL shards (via foreign data tables), the patch allows queries to execute against their respective shards in parallel for better performance and greater concurrency.
  • Dynamic cluster sizing. With a bit of load balancer cleverness, shards can be added or removed from a cluster, dynamically.

pg_dump foreign table data (FDW) backup & restore

Foreign tables extend PostgreSQL in many useful ways. Unfortunately, prior to PostgreSQL 13, pg_dump does not fully support foreign tables; pg_dump correctly dumps the DDL commands needed to define the foreign table again, but no rows will ever be dumped. If you restore that backup, the foreign tables will have been defined again, but there won’t be any data.

In order to make database administration easier for our customers, as of Swarm64 DA 3.1.0, and PostgreSQL 13, we developed a patched version of pg_dump that allows you to dump the content of a foreign table using the --include-foreign-data option.

TPC-DS benchmarking kit for PostgreSQL

If you want to test PostgreSQL query performance against other databases or on one hardware platform versus another, this kit might help. The offering includes everything you need to run industry-standard TPC-H and TPC-DS benchmarks on Postgres:

  • Schema generator – scripts to create a TPC-H or TPC-DS schema in your PostgreSQL database.
  • Data generator – generate and load TPC-H or TPC-DS data sets at any scale factor you choose.
  • Benchmark execution – a Python program runs the TPC-H or TPC-DS queries on your database. Config parameters allow you to control which queries you run, or set query timeout limits, and parallel workers, etc.

Request a Swarm64 demo