Re: index prefetching

Peter Geoghegan <pg@bowt.ie>

From: Peter Geoghegan <pg@bowt.ie>

To: Andres Freund <andres@anarazel.de>

Cc: Tomas Vondra <tomas@vondra.me>, Thomas Munro <thomas.munro@gmail.com>, Nazir Bilal Yavuz <byavuz81@gmail.com>, Robert Haas <robertmhaas@gmail.com>, Melanie Plageman <melanieplageman@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Georgios <gkokolatos@protonmail.com>, Konstantin Knizhnik <knizhnik@garret.ru>, Dilip Kumar <dilipbalaut@gmail.com>

Date: 2025-08-15T18:05:06Z

Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

aio: io_uring: Trigger async processing for large IOs
- a9ee66881744 19 (unreleased) landed
read stream: Split decision about look ahead for AIO and combining
- 8ca147d582a5 19 (unreleased) landed
read_stream: Only increase read-ahead distance when waiting for IO
- f63ca3379025 19 (unreleased) landed
read_stream: Prevent distance from decaying too quickly
- 6e36930f9aaf 19 (unreleased) landed
Reduce ExecSeqScan* code size using pg_assume()
- b227b0bb4e03 19 (unreleased) cited
Fix rare bug in read_stream.c's split IO handling.
- b421223172a2 19 (unreleased) cited
Fix multiranges to behave more like dependent types.
- 3e8235ba4f9c 17.0 cited
Add EXPLAIN (MEMORY) to report planner memory consumption
- 5de890e3610d 17.0 cited
Optimize nbtree backward scan boundary cases.
- c9c0589fda0e 17.0 cited
Increment xactCompletionCount during subtransaction abort.
- 90c885cdab8b 14.0 cited
Add nbtree Valgrind buffer lock checks.
- 4a70f829d86c 14.0 cited
Add nbtree high key "continuescan" optimization.
- 29b64d1de7c7 12.0 cited
Reduce pinning and buffer content locking for btree scans.
- 2ed5b87f96d4 9.5.0 cited
Teach btree to handle ScalarArrayOpExpr quals natively.
- 9e8da0f75731 9.2.0 cited

On Fri, Aug 15, 2025 at 1:23 PM Andres Freund <andres@anarazel.de> wrote:
> Somewhat random note about I/O waits:
>
> Unfortunately the I/O wait time we measure often massively *over* estimate the
> actual I/O time. If I execute the above query with the patch applied, we
> actually barely ever wait for I/O to complete, it's all completed by the time
> we have to wait for the I/O. What we are measuring is the CPU cost of
> *initiating* the I/O.

I do get that.

This was really obvious when I temporarily switched the prefetch patch
over from using READ_STREAM_DEFAULT to using READ_STREAM_USE_BATCHING
(this is probably buggy, but still seems likely to be representative
of what's possible with some care). I noticed that that change reduced
the reported "shared read" time by 10x -- which had exactly zero impact on
query execution time (at least for the queries I looked at). Since, as
you say, the backend didn't have to wait for I/O to complete either
way.

--
Peter Geoghegan