Re: Improving tracking/processing of buildfarm test failures

Hello hackers,

Please take a look at the July report on buildfarm failures:
# SELECT br, COUNT(*) FROM failures WHERE dt >= '2025-07-01' AND
  dt < '2025-08-01' GROUP BY br;
REL_13_STABLE: 4
REL_14_STABLE: 2
REL_15_STABLE: 4
REL_16_STABLE: 3
REL_17_STABLE: 4
REL_18_STABLE: 9
master: 42
-- Total: 68
(Counting test failures only, excluding indent-check, Configure, Build
errors.)

# SELECT COUNT(*) FROM (SELECT DISTINCT issue_link FROM failures WHERE
  dt >= '2025-07-01' AND dt < '2025-08-01');
20

# SELECT issue_link, COUNT(*) FROM failures WHERE dt >= '2025-07-01' AND
  dt < '2025-08-01' GROUP BY issue_link ORDER BY 2 DESC LIMIT 5;
https://www.postgresql.org/message-id/202507041645.afjl5rssvrgu%40alvherre.pgsql : 7
-- Fixed

https://www.postgresql.org/message-id/6a1ccdf2-3e5f-4b08-984a-8ceb40e36ec1%40gmail.com : 6
-- Fixed

https://www.postgresql.org/message-id/90506.1751754464%40sss.pgh.pa.us : 6
-- Fixed

https://www.postgresql.org/message-id/202507071504.xm2r26u7lmzr@alvherre.pgsql : 6
-- Fixed

https://www.postgresql.org/message-id/CA+hUKGL0bikWSC2XW-zUgFWNVEpD_gEWXndi2PE5tWqmApkpZQ@mail.gmail.com : 5

# SELECT COUNT(*) FROM failures WHERE dt >= '2025-07-01' AND
  dt < '2025-08-01' AND issue_link IS NULL; -- Unsorted/unhelpful failures
15

Short-lived failures:
34

Best regards,
Alexander