| From: | Michael Paquier <michael(dot)paquier(at)gmail(dot)com> |
|---|---|
| To: | Patrick B <patrickbakerbr(at)gmail(dot)com> |
| Cc: | Venkata B Nagothi <nag1010(at)gmail(dot)com>, pgsql-general <pgsql-general(at)postgresql(dot)org> |
| Subject: | Re: WAL history files - Pgsql 9.2 |
| Date: | 2016-12-12 03:08:04 |
| Message-ID: | CAB7nPqRW2L_PoGWfK5gDK_QDhPjm53n7AYjX0=RnOLUGcbv4aA@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Mon, Dec 12, 2016 at 9:31 AM, Patrick B <patrickbakerbr(at)gmail(dot)com> wrote:
> No.. it didn't copy. i tested here. I had to manually copy the history file
> from /var/lib/pgsql/9.2/data/pg_xlogs from the new master to the same
> directory on the slaves.
An archive command is able to properly fetch the history files to
define a new timeline. If you rely only on streaming replication, only
9.3 and newer versions are able to do that: the replication protocol
has been extended at this point to allow a standby to fetch from a
primary history files.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michael Paquier | 2016-12-12 04:08:08 | Re: logical decoding output plugin |
| Previous Message | Michael Paquier | 2016-12-12 02:23:03 | Re: [GENERAL] Re: Would like to below scenario is possible for getting page/block corruption |