Re: PG Upgrade with hardlinks, when to start/stop master and replicas - Mailing list pgsql-general

From Martín Fernández
Subject Re: PG Upgrade with hardlinks, when to start/stop master and replicas
Date
Msg-id CAH0BA3kasnHosSw1LB-YXRANc15nKz7knRt9oHSqzh44oc7UvQ@mail.gmail.com
Whole thread Raw
In response to Re: PG Upgrade with hardlinks, when to start/stop master and replicas  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general

On Fri, Feb 22, 2019 at 2:03 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
Bruce Momjian <bruce@momjian.us> writes:
> On Thu, Feb 21, 2019 at 09:31:32PM -0500, Stephen Frost wrote:
>> * Bruce Momjian (bruce@momjian.us) wrote:
>>> There was too much concern that users would accidentally start the old
>>> server at some later point, and its files would be hard linked to the
>>> new live server, leading to disaster.


I think this is a great solution. Knowing that neither a human nor a supervisor can mess up the hardlinks is something I really appreciate. 
 
>> Sure, I understand that concern, just wish there was a better approach
>> we could use for "DO NOT START THIS SERVER" rather than moving of the
>> pg_control file.

> As ugly as it is, I have never heard of a better solution.

system("rm -rf $OLDPGDATA") ... nah, that is not a better idea.

                        regards, tom lane

pgsql-general by date:

Previous
From: Zachary Hanson-Hart
Date:
Subject: streaming replication authentication
Next
From: Sergey Burladyan
Date:
Subject: Re: Running from 9.6 backups sometimes fails with fatal error