| From: | solai v <solai(dot)cdac(at)gmail(dot)com> |
|---|---|
| To: | Aleksander Alekseev <aleksander(at)tigerdata(dot)com> |
| Cc: | Jacob Brazeal <jacob(dot)brazeal(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Andrey Borodin <x4mmm(at)yandex-team(dot)ru> |
| Subject: | Re: SERIALIZABLE can commit write skew after SSI summarization |
| Date: | 2026-11-02 11:56:11 |
| Message-ID: | CAF0whufQsPmxUi8oUr-aKSMLJ41d2XmmMkb8S+c1R2stv7dfZw@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
I have tested the v2 patches for "SERIALIZABLE can commit write skew after
SSI summarization."
The following are my testing results
- Reproduced the issue before applying the patches, where both
serializable transactions committed and the final count became 2.
- After applying the patches, PostgreSQL raised a serialization failure
when the first transaction attempted to update the second row.
- Verified that predicate-lock summarization occurred (locks_summarized
= true) and the conflict was detected even after summarization.
- Verified the final table contents: only one row had v = 1 (final count
= 1).
regards,
solai
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Sami Imseih | 2026-11-02 12:11:29 | Re: Report buffer access strategies in EXPLAIN |
| Previous Message | Pavel Stehule | 2026-11-02 11:52:34 | Re: PL/pgSQL: apply the too_many_rows extra check to EXECUTE ... INTO |