Re: BUG #14109: pg_rewind fails to update target control file in one scenario

Thanks Michael=2C=0A=
=0A=
----------------------------------------=0A=
> Date: Tue=2C 26 Apr 2016 08:04:58 +0900=0A=
> Subject: Re: [BUGS] BUG #14109: pg_rewind fails to update target control =
file in one scenario=0A=
> From: michael.paquier@gmail.com=0A=
> To: johnlumby@hotmail.com=0A=
> CC: julien.rouhaud@dalibo.com=3B pgsql-bugs@postgresql.org=0A=
>=0A=
> On Tue=2C Apr 26=2C 2016 at 7:15 AM=2C John Lumby <johnlumby@hotmail.com>=
 wrote:=0A=
>> So=2C assuming this is reliable (will work without requiring archiving)=
=0A=
>> then my problem is solved.=0A=
>=0A=
> Depending on the checkpoint frequency and the activity on your=0A=
> systems=2C you may face problems with missing WAL segments at some point=
=0A=
> because past WAL segments need to be recycled or removed by the server=0A=
> to move on with its life.=0A=
=0A=
Yes=2C=A0 I fear I could be caught out by that --=0A=
in fact that is why I now always "stabilize" the replication by halting ins=
/upd/del activity=0A=
and then shut the current Primary down first before promoting current Stand=
by.=0A=
I *think* that then should guarantee there cannot be any missing WAL segmen=
ts=0A=
when I then rewind the old Primary to become new Standby.=0A=
=0A=
> One way to take care of this class of=0A=
> problems is to use wal_keep_segments. An even better one is called=0A=
> replication slot.=0A=
=0A=
Regarding replication slots=A0 --=A0=A0 Actually I do use them (I think it =
is unsafe to run=0A=
streaming replication without either archiving or a replication slot) =0A=
but even that would still not guarantee success =0A=
if I did not take the precaution of shutting down current primary first bef=
ore flip.=0A=
=0A=
And=A0 ..=A0=A0 we discussed this very point in pqsql-general just a month =
ago=A0 --=0A=
=0A=
http://www.postgresql.org/message-id/COL131-W804D45E77B0D0FB1EF08B1A3890@ph=
x.gbl=0A=
=0A=
I did not get any answer to my suggestion in that post but I think it might=
 be useful.=0A=
=0A=
> This solely depends on how your system is working=2C=0A=
> so perhaps you will not need some extra configuration.=0A=
> --=0A=
> Michael=0A=
=0A=
I think there needs to be some clear instructions on exactly what configura=
tion is needed=0A=
to be able to run streaming replication and always be able to flip=0A=
Standby->Primary=A0=A0 =2C=A0 <some actions>=A0=A0 =2C=A0=A0 Primary-> Stan=
dby=0A=
=0A=
and in those posts in pgsql-general I wrote a suggested addition to the wik=
i page=0A=
but was unable to edit it myself.=0A=
=0A=
Cheers=2C=A0=A0=A0 John=0A=
=0A=
                           =