Thread: SQL: Chaining versus Pipelining
Hi PgSQL general mailing lists,
In Postgres Weekly, I encountered this interesting article. The documentation on the use of AND CHAIN is here. A co-worker is interested in the difference between chaining and using the pipelining syntax.
They look pretty similar to me; it could be that one implementation is PG specific and not part of the SQL standard or something. Can anyone shed any light for us?
Thanks in advance for any assistance anyone can provide,
Joseph Hammerman
On Mon, 2024-07-08 at 15:26 -0700, Joseph Hammerman wrote: > In Postgres Weekly, I encountered this interesting article. The documentation on the use > of AND CHAIN is here. A co-worker is interested in the difference between chaining > and using the pipelining syntax. > > They look pretty similar to me; it could be that one implementation is PG specific and > not part of the SQL standard or something. Can anyone shed any light for us? COMMIT AND CHAIN and pipelining have little in common, except that they can both reduce the overhead caused by a high latency between database client and server. COMMIT AND CHAIN just saves you a round trip. Pipelining allows you to send the next database request before you get the response to the first one. Here is an article that describes pipelining: https://www.cybertec-postgresql.com/en/pipeline-mode-better-performance-on-slow-network/ Yours, Laurenz Albe