Re: FSM corruption and standby servers

I have some comments to the documentation. 

I worked with PostgreSQL for some time ago( 5-8 years)  setting up a few Postgres databases using version 8.4 and RHEL 5:* in production and still running.

A LOT have happened since then and I am going to setup Postgres 9+ with everything needed to secure data, backup(PITR) 

, cluster with zero dataloss. OS is RHEL 7.2.

I my opinion the documentation is not very straightforward with examples omn exactly how to make such a complete installation using

2+ servers to make data available 24/7. One has to pick up a little here and a litte there. And one always have to do a litte there and here to make things switch

in order to make data available 24/7. 

Creating of tablespace can't be scriptet, Restore(I know this another list(pg_rman) can't handle symbolic links. 

Sorry about my belching as vocabulary and my bad english!

/Poul

     

 

2016-10-31 19:19 GMT+01:00 Tom Lane <tgl@sss.pgh.pa.us>:

"David G. Johnston" <david.g.johnston@gmail.com> writes:
> On Mon, Oct 31, 2016 at 9:55 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> I would have thought that the referenced page is clear enough about
>> needing to check the standbys; do you think it isn't?

> ​I can ​see how the following is a bit loose for someone not super-familiar
> with WAL.

Yeah.  On the other hand, I don't want people who aren't running
replication to stop reading as soon as they see something about standby
servers.  I tweaked the existing wording a bit to emphasize that standbys
can be corrupt even with a valid master; hopefully that's enough.

                        regards, tom lane


--
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin

--

Med venlig hilsen / Best regards
Poul Kristensen
Linux-OS/Virtualizationexpert and Oracle DBA