Re: [PATCH] Fall back to serializing when parallel apply runs out of DSM segments

From: Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>
To: Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>
Cc: Baji Shaik <baji(dot)pgdev(at)gmail(dot)com>, Zhijie Hou <houzhijie22(at)gmail(dot)com>, PostgreSQL-development <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: [PATCH] Fall back to serializing when parallel apply runs out of DSM segments
Date: 2026-11-18 17:42:08
Message-ID: CALj2ACXBHXpGXy+-8=RNjQsvemsYAspbzhOKAouovGFmhFVk9g@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

On Tue, Nov 17, 2026 at 2:53 PM Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> wrote:
>
> I've reviewed both patches and both look good to me.

Thanks for taking a look.

> Regarding backpatching these patches, I'm inclined to backpatch both
> of them. Regarding the parallel apply worker issue, it seems a bug to
> me.
>
> Regarding the 0002 patch, it simply violates what DSA_ALLOC_NO_OOM
> promises. IIUC make_new_segment() is implemented to exit gracefully
> even in erroneous cases so that its caller can handle the situation. A
> caller using DSA_ALLOC_NO_OOM gets an error if DSM slots are used up,
> instead of InvalidDsaPointer.

Agreed to both of the above. My understanding was that although these
are not behaving as expected, and so are bugs, I haven't seen such
issues in practice or reported from the field, and even if they were,
the errors are recoverable or reported to the callers. That said, I'm
fine to classify them as back-patchable here. Thanks.

I marked it as RfC.

--
Bharath Rupireddy
Amazon Web Services: https://aws.amazon.com

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2026-11-18 17:46:58 Re: [PATCH] Normalize numeric/bigint scientific notation after rounding
Previous Message Greg Burd 2026-11-18 17:31:49 Re: nbtree: opt-in leaf key prefix compression