Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS

От

Craig Ringer

Тема

Дата

29 марта 2018 г. в 08:32:43

Msg-id

CAMsr+YFET=hWD7A5ULSpbbiZ6e-N-W7ajga1zZfzy=mko5qfqA@mail.gmail.com

обсуждение

Ответ на

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS (Thomas Munro)

Список

pgsql-hackers

Дерево обсуждения

PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 28 марта 2018 г. в 05:23:46

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 23 апреля 2018 г. в 23:14:48

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 24 апреля 2018 г. в 03:09:23

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 27 апреля 2018 г. в 04:18:55

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 26 апреля 2018 г. в 05:16:52

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 30 апреля 2018 г. в 07:55:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tom Lane <tgl@sss.pgh.pa.us> 28 марта 2018 г. в 06:53:08

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Michael Paquier <michael@paquier.xyz> 29 марта 2018 г. в 05:30:59

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 29 марта 2018 г. в 05:48:27

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Justin Pryzby <pryzby@telsasoft.com> 29 марта 2018 г. в 08:00:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 29 марта 2018 г. в 08:06:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 29 марта 2018 г. в 08:25:51

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Gasper Zejn <zejn@owca.info> 21 апреля 2018 г. в 22:21:39

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 29 марта 2018 г. в 08:32:43

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 29 марта 2018 г. в 08:35:47

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 29 марта 2018 г. в 08:58:45

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 29 марта 2018 г. в 15:07:56

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Catalin Iacob <iacobcatalin@gmail.com> 29 марта 2018 г. в 19:20:00

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 30 марта 2018 г. в 00:18:14

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 31 марта 2018 г. в 16:24:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 31 марта 2018 г. в 19:13:09

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tom Lane <tgl@sss.pgh.pa.us> 31 марта 2018 г. в 19:38:12

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Michael Paquier <michael@paquier.xyz> 1 апреля 2018 г. в 03:20:38

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 1 апреля 2018 г. в 04:14:46

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 2 апреля 2018 г. в 21:13:46

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 2 апреля 2018 г. в 21:53:20

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 2 апреля 2018 г. в 22:32:45

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 2 апреля 2018 г. в 23:38:06

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Stephen Frost <sfrost@snowman.net> 2 апреля 2018 г. в 23:58:08

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 3 апреля 2018 г. в 02:05:44

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 3 апреля 2018 г. в 02:23:24

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 3 апреля 2018 г. в 02:27:35

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 3 апреля 2018 г. в 03:03:39

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 3 апреля 2018 г. в 03:05:09

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 3 апреля 2018 г. в 03:07:41

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Peter Geoghegan <pg@bowt.ie> 3 апреля 2018 г. в 03:48:00

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 3 апреля 2018 г. в 05:54:26

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 3 апреля 2018 г. в 13:35:39

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 4 апреля 2018 г. в 00:47:01

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 03:56:37

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 4 апреля 2018 г. в 04:54:50

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 05:05:19

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 05:14:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 4 апреля 2018 г. в 05:44:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 4 апреля 2018 г. в 08:29:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 09:00:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 4 апреля 2018 г. в 10:32:04

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 16:49:38

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Gasper Zejn <zejn@owca.info> 4 апреля 2018 г. в 20:23:58

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 18:23:51

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 20:51:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andrew Gierth <andrew@tao11.riddles.org.uk> 6 апреля 2018 г. в 02:37:42

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 6 апреля 2018 г. в 04:27:05

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 6 апреля 2018 г. в 05:53:56

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 6 апреля 2018 г. в 06:20:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 8 апреля 2018 г. в 05:16:07

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 8 апреля 2018 г. в 05:33:37

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 8 апреля 2018 г. в 05:37:47

Re: PostgreSQL's handling of fsync() errors is unsafe and risks dataloss at least on XFS Craig Ringer <craig@2ndquadrant.com> 8 апреля 2018 г. в 06:27:45

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 8 апреля 2018 г. в 06:46:17

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 8 апреля 2018 г. в 13:30:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 8 апреля 2018 г. в 19:38:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 9 апреля 2018 г. в 01:29:16

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 02:16:25

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 9 апреля 2018 г. в 02:27:57

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 04:55:10

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 9 апреля 2018 г. в 05:00:41

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 05:06:12

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 9 апреля 2018 г. в 06:15:01

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 16:54:19

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 16:42:35

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Abhijit Menon-Sen <ams@2ndQuadrant.com> 9 апреля 2018 г. в 16:47:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 00:29:17

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 18 апреля 2018 г. в 00:34:53

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 12:52:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 00:32:45

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 18 апреля 2018 г. в 00:41:42

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 14:56:57

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Gasper Zejn <zejn@owca.info> 9 апреля 2018 г. в 21:02:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 9 апреля 2018 г. в 02:10:24

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 9 апреля 2018 г. в 04:35:06

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andreas Karlsson <andreas@proxel.se> 8 апреля 2018 г. в 12:41:06

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 8 апреля 2018 г. в 13:31:24

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Peter Geoghegan <pg@bowt.ie> 8 апреля 2018 г. в 06:37:06

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 9 апреля 2018 г. в 00:23:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 00:47:04

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 9 апреля 2018 г. в 11:45:40

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 13:50:41

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Geoff Winkless <pgsqladmin@geoff.dj> 9 апреля 2018 г. в 15:03:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 15:31:27

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 16:33:18

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 17:22:06

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 22:47:44

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 10 апреля 2018 г. в 01:33:16

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 10 апреля 2018 г. в 03:32:20

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 9 апреля 2018 г. в 18:29:36

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 22:26:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 22:29:16

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 22:44:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 22:37:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 22:51:12

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 22:54:05

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 23:04:20

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 23:30:00

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 23:37:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 04:59:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 10 апреля 2018 г. в 05:00:59

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 10 апреля 2018 г. в 19:54:40

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 00:19:53

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 18 апреля 2018 г. в 13:04:30

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 14:46:15

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 18 апреля 2018 г. в 15:45:53

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 20 апреля 2018 г. в 23:49:08

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Mark Kirkwood <mark.kirkwood@catalyst.net.nz> 19 апреля 2018 г. в 02:31:50

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 19 апреля 2018 г. в 03:44:33

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS "Joshua D. Drake" <jd@commandprompt.com> 10 апреля 2018 г. в 21:58:37

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS "Joshua D. Drake" <jd@commandprompt.com> 10 апреля 2018 г. в 22:51:01

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS "Joshua D. Drake" <jd@commandprompt.com> 10 апреля 2018 г. в 23:57:34

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 11 апреля 2018 г. в 15:23:49

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 9 апреля 2018 г. в 15:16:38

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 9 апреля 2018 г. в 15:54:16

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Justin Pryzby <pryzby@telsasoft.com> 9 апреля 2018 г. в 22:41:19

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 22:59:34

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 04:44:59

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 10 апреля 2018 г. в 04:52:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andreas Karlsson <andreas@proxel.se> 10 апреля 2018 г. в 03:41:10

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 05:02:48

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 9 апреля 2018 г. в 19:45:00

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS "Joshua D. Drake" <jd@commandprompt.com> 9 апреля 2018 г. в 20:26:24

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Mark Dilger <hornschnorter@gmail.com> 9 апреля 2018 г. в 21:29:42

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 22:22:58

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 9 апреля 2018 г. в 22:02:11

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 22:13:14

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 10 апреля 2018 г. в 18:15:46

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 10 апреля 2018 г. в 18:40:05

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Jonathan Corbet <corbet@lwn.net> 11 апреля 2018 г. в 15:05:27

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 11 апреля 2018 г. в 17:29:09

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Jonathan Corbet <corbet@lwn.net> 11 апреля 2018 г. в 17:40:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 10 апреля 2018 г. в 19:38:27

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Peter Geoghegan <pg@bowt.ie> 9 апреля 2018 г. в 22:25:33

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 18 апреля 2018 г. в 00:49:42

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Mark Dilger <hornschnorter@gmail.com> 9 апреля 2018 г. в 23:25:54

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 9 апреля 2018 г. в 23:34:15

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 04:54:30

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 9 апреля 2018 г. в 23:43:03

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Mark Dilger <hornschnorter@gmail.com> 9 апреля 2018 г. в 23:55:29

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Andres Freund <andres@anarazel.de> 10 апреля 2018 г. в 00:08:29

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Tomas Vondra <tomas.vondra@2ndquadrant.com> 10 апреля 2018 г. в 00:25:52

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Mark Dilger <hornschnorter@gmail.com> 10 апреля 2018 г. в 00:33:29

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Michael Paquier <michael@paquier.xyz> 10 апреля 2018 г. в 08:04:13

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 08:37:19

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Michael Paquier <michael@paquier.xyz> 10 апреля 2018 г. в 09:10:21

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 10 апреля 2018 г. в 15:15:15

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 18 апреля 2018 г. в 13:19:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Robert Haas <robertmhaas@gmail.com> 10 апреля 2018 г. в 18:28:07

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Christophe Pettus <xof@thebuild.com> 9 апреля 2018 г. в 00:28:43

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 9 апреля 2018 г. в 04:31:56

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 10:51:53

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 17:00:15

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 5 апреля 2018 г. в 00:28:09

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 5 апреля 2018 г. в 01:14:24

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 17:09:09

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 17:25:47

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 17:42:18

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Antonis Iliopoulos <ailiop@altatus.com> 4 апреля 2018 г. в 18:23:31

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 4 апреля 2018 г. в 05:40:16

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 4 апреля 2018 г. в 16:53:01

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 3 апреля 2018 г. в 14:26:05

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 3 апреля 2018 г. в 16:36:47

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Greg Stark <stark@mit.edu> 3 апреля 2018 г. в 17:37:30

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 3 апреля 2018 г. в 19:52:07

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Peter Geoghegan <pg@bowt.ie> 3 апреля 2018 г. в 06:45:30

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 3 апреля 2018 г. в 17:29:10

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Anthony Iliopoulos <ailiop@altatus.com> 1 апреля 2018 г. в 03:58:22

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 1 апреля 2018 г. в 21:24:51

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 2 апреля 2018 г. в 18:03:42

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 3 апреля 2018 г. в 04:29:28

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Thomas Munro <thomas.munro@enterprisedb.com> 4 апреля 2018 г. в 02:59:27

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 29 марта 2018 г. в 16:15:10

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 5 апреля 2018 г. в 10:09:57

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Bruce Momjian <bruce@momjian.us> 5 апреля 2018 г. в 22:33:14

Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS Craig Ringer <craig@2ndquadrant.com> 5 апреля 2018 г. в 11:46:08

On 29 March 2018 at 10:48, Thomas Munro <thomas.munro@enterprisedb.com> wrote:

On Thu, Mar 29, 2018 at 3:30 PM, Michael Paquier <michael@paquier.xyz> wrote:
> On Tue, Mar 27, 2018 at 11:53:08PM -0400, Tom Lane wrote:
>> Craig Ringer <craig@2ndquadrant.com> writes:
>>> TL;DR: Pg should PANIC on fsync() EIO return.
>>
>> Surely you jest.
>
> Any callers of pg_fsync in the backend code are careful enough to check
> the returned status, sometimes doing retries like in mdsync, so what is
> proposed here would be a regression.

Craig, is the phenomenon you described the same as the second issue
"Reporting writeback errors" discussed in this article?

https://lwn.net/Articles/724307/

A variant of it, by the looks.

The problem in our case is that the kernel only tells us about the error once. It then forgets about it. So yes, that seems like a variant of the statement:

"Current kernels might report a writeback error on an fsync() call,
but there are a number of ways in which that can fail to happen."

That's... I'm speechless.

Yeah.

It's a bit nuts.

I was astonished when I saw the behaviour, and that it appears undocumented.

Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

В списке pgsql-hackers по дате отправления

Предыдущее

От: Craig Ringer

Дата: 29 марта 2018 г. в 08:25:51

Сообщение: Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS

Следующее

От: Craig Ringer

Дата: 29 марта 2018 г. в 08:35:47

Сообщение: Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS