Thread: Small fixes about backup history file in doc and pg_standby
Hi, While looking into the backup and recovery code, I found small documentation bugs. The documatation says that the backup history files can be requested for recovery, but it's not used by the system and not requested anymore since PG 9.0 (commit 06f82b29616cd9effcaefd99c6b6e2e80697482f) and never be requested. Attached patch (doc_backup_history_file.patch) corrects the description about this. In addition, the current pg_standby still can handle a backup history file that are never requested. It is harmless but unnecessary code. Another attached patch (pg_standby.patch) removes this part of code. Regards, -- Yugo Nagata <nagata@sraoss.co.jp>
Attachment
Hello. Good catch! At Tue, 26 Jun 2018 17:47:52 +0900, Yugo Nagata <nagata@sraoss.co.jp> wrote in <20180626174752.0ce505e3.nagata@sraoss.co.jp> > Hi, > > While looking into the backup and recovery code, I found small documentation bugs. > The documatation says that the backup history files can be requested for recovery, > but it's not used by the system and not requested anymore since PG 9.0 > (commit 06f82b29616cd9effcaefd99c6b6e2e80697482f) and never be requested. > > Attached patch (doc_backup_history_file.patch) corrects the description about this. > > In addition, the current pg_standby still can handle a backup history file that are > never requested. It is harmless but unnecessary code. Another attached patch > (pg_standby.patch) removes this part of code. The comment fix seems fine and they seem to be all occurances of the word ".backup" in the context of recovery_command. The definition of the symbol XLOG_BACKUP_LABEL is no longer useful after your patch applied. Removing the symbol makes XLOG_DATA and the variable nextWALFileName useless and finally we can remove all branching using it. regards. -- Kyotaro Horiguchi NTT Open Source Software Center
On Tue, Jun 26, 2018 at 5:47 PM, Yugo Nagata <nagata@sraoss.co.jp> wrote: > Hi, > > While looking into the backup and recovery code, I found small documentation bugs. > The documatation says that the backup history files can be requested for recovery, > but it's not used by the system and not requested anymore since PG 9.0 > (commit 06f82b29616cd9effcaefd99c6b6e2e80697482f) and never be requested. > > Attached patch (doc_backup_history_file.patch) corrects the description about this. Pushed. Thanks! > In addition, the current pg_standby still can handle a backup history file that are > never requested. It is harmless but unnecessary code. Another attached patch > (pg_standby.patch) removes this part of code. Since this is not bug fix, let's discuss this in 12dev cycle. Regards, -- Fujii Masao
On Wed, 27 Jun 2018 00:58:18 +0900 Fujii Masao <masao.fujii@gmail.com> wrote: > On Tue, Jun 26, 2018 at 5:47 PM, Yugo Nagata <nagata@sraoss.co.jp> wrote: > > Hi, > > > > While looking into the backup and recovery code, I found small documentation bugs. > > The documatation says that the backup history files can be requested for recovery, > > but it's not used by the system and not requested anymore since PG 9.0 > > (commit 06f82b29616cd9effcaefd99c6b6e2e80697482f) and never be requested. > > > > Attached patch (doc_backup_history_file.patch) corrects the description about this. > > Pushed. Thanks! Thanks! > > > In addition, the current pg_standby still can handle a backup history file that are > > never requested. It is harmless but unnecessary code. Another attached patch > > (pg_standby.patch) removes this part of code. > > Since this is not bug fix, let's discuss this in 12dev cycle. Certainly. Regards, > > Regards, > > -- > Fujii Masao > -- Yugo Nagata <nagata@sraoss.co.jp>
On Tue, 26 Jun 2018 20:19:42 +0900 (Tokyo Standard Time) Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> wrote: > Hello. > > Good catch! > > At Tue, 26 Jun 2018 17:47:52 +0900, Yugo Nagata <nagata@sraoss.co.jp> wrote in <20180626174752.0ce505e3.nagata@sraoss.co.jp> > > Hi, > > > > While looking into the backup and recovery code, I found small documentation bugs. > > The documatation says that the backup history files can be requested for recovery, > > but it's not used by the system and not requested anymore since PG 9.0 > > (commit 06f82b29616cd9effcaefd99c6b6e2e80697482f) and never be requested. > > > > Attached patch (doc_backup_history_file.patch) corrects the description about this. > > > > In addition, the current pg_standby still can handle a backup history file that are > > never requested. It is harmless but unnecessary code. Another attached patch > > (pg_standby.patch) removes this part of code. > > The comment fix seems fine and they seem to be all occurances of > the word ".backup" in the context of recovery_command. > > The definition of the symbol XLOG_BACKUP_LABEL is no longer > useful after your patch applied. Removing the symbol makes > XLOG_DATA and the variable nextWALFileName useless and finally we > can remove all branching using it. Thank you for your reviewing my patch. I've also removed XLOG_BACKUP_LABEL, but I left nextWALFileName since this is still referred in CustomizableCleanupPriorWALFiles(). Attached is the updated patch. Regards, > > regards. > > -- > Kyotaro Horiguchi > NTT Open Source Software Center > > -- Yugo Nagata <nagata@sraoss.co.jp>
Attachment
On Wed, Jun 27, 2018 at 05:42:07PM +0900, Yugo Nagata wrote: > On Wed, 27 Jun 2018 00:58:18 +0900 > Fujii Masao <masao.fujii@gmail.com> wrote: >>> In addition, the current pg_standby still can handle a backup history file that are >>> never requested. It is harmless but unnecessary code. Another attached patch >>> (pg_standby.patch) removes this part of code. >> >> Since this is not bug fix, let's discuss this in 12dev cycle. +1. -- Michael
Attachment
On Wed, 27 Jun 2018 18:36:46 +0900 Michael Paquier <michael@paquier.xyz> wrote: > On Wed, Jun 27, 2018 at 05:42:07PM +0900, Yugo Nagata wrote: > > On Wed, 27 Jun 2018 00:58:18 +0900 > > Fujii Masao <masao.fujii@gmail.com> wrote: > >>> In addition, the current pg_standby still can handle a backup history file that are > >>> never requested. It is harmless but unnecessary code. Another attached patch > >>> (pg_standby.patch) removes this part of code. > >> > >> Since this is not bug fix, let's discuss this in 12dev cycle. > > +1. I added this to CF. > -- > Michael -- Yugo Nagata <nagata@sraoss.co.jp>
On 27.06.18 18:22, Yugo Nagata wrote: > On Wed, 27 Jun 2018 18:36:46 +0900 > Michael Paquier <michael@paquier.xyz> wrote: > >> On Wed, Jun 27, 2018 at 05:42:07PM +0900, Yugo Nagata wrote: >>> On Wed, 27 Jun 2018 00:58:18 +0900 >>> Fujii Masao <masao.fujii@gmail.com> wrote: >>>>> In addition, the current pg_standby still can handle a backup history file that are >>>>> never requested. It is harmless but unnecessary code. Another attached patch >>>>> (pg_standby.patch) removes this part of code. >>>> >>>> Since this is not bug fix, let's discuss this in 12dev cycle. >> >> +1. > > I added this to CF. committed -- Peter Eisentraut http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services