Re: Improving tracking/processing of buildfarm test failures

Hello hackers,

Please take a look at the May report on buildfarm failures:
# SELECT br, COUNT(*) FROM failures WHERE dt >= '2026-05-01' AND
  dt < '2026-06-01' GROUP BY br;
REL_14_STABLE: 9
REL_15_STABLE: 6
REL_16_STABLE: 26
REL_17_STABLE: 26
REL_18_STABLE: 22
master: 164
-- Total: 253
(Counting test failures only, excluding indent-check, Configure, Build
errors, also excluding failures from fruitcrow..)

# SELECT COUNT(*) FROM (SELECT DISTINCT issue_link FROM failures WHERE
  dt >= '2026-05-01' AND dt < '2026-06-01');
25

# SELECT issue_link, COUNT(*) FROM failures WHERE dt >= '2026-06-01' AND
  dt < '2026-06-01' GROUP BY issue_link ORDER BY 2 DESC LIMIT 6;

https://www.postgresql.org/message-id/CAPpHfdvht%3D_VcpDybYvwon96Enh%3D-S-9wGLGPzbu%2B6_5QmwmwQ%40mail.gmail.com : 105
-- Fixed

https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=6f0bff33d : 23
-- Fixed

https://www.postgresql.org/message-id/c64cbda0-7ef2-4762-8e70-9d0dedccc9cf%40gmail.com : 20
-- Fixed

https://www.postgresql.org/message-id/CAD21AoBmdmBQ-+Jga+jSKKq5OPGEP1pEjSJfRPT6MCwVHLD6og@mail.gmail.com : 11
-- Fixed

https://www.postgresql.org/message-id/CA+hUKGL0bikWSC2XW-zUgFWNVEpD_gEWXndi2PE5tWqmApkpZQ@mail.gmail.com : 10


https://www.postgresql.org/message-id/agr6-cIQ4EUA86Cs%40paquier.xyz : 8
-- Fixed

# SELECT COUNT(*) FROM failures WHERE dt >= '2026-05-01' AND
  dt < '2026-06-01' AND issue_link IS NULL; -- Unsorted/unhelpful failures
44

Short-lived failures: 158

Best regards,
Alexander