Re: BUG: ReadStream look-ahead exhausts local buffers when effective_io_concurrency>=64

From: Feike Steenbergen <feikesteenbergen(at)gmail(dot)com>
To: Eduard Stepanov <crtxcz(at)gmail(dot)com>
Cc: Induja Sreekanthan <indujas(at)google(dot)com>, pgsql-hackers(at)postgresql(dot)org, Simhachala Sasikanth Gottapu <simhachala(at)google(dot)com>, Vishal Bagga <vishalbagga(at)google(dot)com>, Madhukar <madhukarprasad(at)google(dot)com>, Shihao Zhong <shihaozhong(at)google(dot)com>, Yi Ding <yidin(at)google(dot)com>, Hardik Singh Negi <hardiksnegi(at)google(dot)com>
Subject: Re: BUG: ReadStream look-ahead exhausts local buffers when effective_io_concurrency>=64
Date: 2026-05-06 12:24:38
Message-ID: CAK_s-G2qY+D2X=cBoNBP3pxj+zmsqSjtZvmmXr8nE586f-0Fbw@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi, all,

> On Tue, 10 Mar 2026 at 23:22, Induja Sreekanthan <indujas(at)google(dot)com>
wrote:
> >
> > Hi,
> >
> > I found an issue where Postgres (with effective_io_concurrency of 64 or
higher) runs out of local buffers during a sequential scan on a temporary
table with TOAST data.
> >

Just to confirm, I have a production db (20TB, high concurrency, lot's of
updates/merges) that
is a heavy user of temp tables running into this as well with
io_method='io_uring'.

This started after having upgraded PostgreSQL 18 (from 17).
For now, the problem disappears again when switching to io_method='worker'

Thank you for providing the patches,

kind regards,

Feike Steenbergen

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Feike Steenbergen 2026-05-06 12:47:46 Re: BUG: ReadStream look-ahead exhausts local buffers when effective_io_concurrency>=64
Previous Message Tatsuo Ishii 2026-05-06 12:19:07 Re: Proposal: tighten validation for legacy EUC encodings or document that accepted byte sequences may be unconvertible to UTF8