On Wed, Feb 24, 2010 at 2:31 AM, Simon Riggs <simon@2ndquadrant.com> wrote:
> On Wed, 2010-02-24 at 08:30 +0200, Heikki Linnakangas wrote:
>> Simon Riggs wrote:
>> > On Mon, 2010-02-22 at 11:47 +0000, Heikki Linnakangas wrote:
>> >> Log Message:
>> >> -----------
>> >> Move documentation of all recovery.conf option to a new chapter.
>> >> They used to be scattered between the "backup and restore" and "streaming
>> >> replication" chapters.
>> >
>> > It's just taken me 15 minutes to locate the settings for
>> > primary_conninfo to better understand Stefan's recent post.
>> >
>> > The commit referred to here is an extremely bad change.
>> >
>> > If you intended this to be a heading within the "High Availability"
>> > chapter then I would agree. This is what I thought you had done.
>>
>> The idea was to have one chapter that describes all the options in
>> recovery.conf. Some of them are specific to streaming replication
>> (primary_conninfo), some are specific to PITR (recovery_target_*), some
>> are common (restore_command, restore_end_command). We need a reference
>> page to list them all.
>
> We need a section, not a new chapter, if at all. A chapter on its own
> for this makes no sense at all. The new section should either be in the
> Backup, HA or Server Config chapters.
>
> Please revert the change to create a new chapter.
Just to mention it, this change was discussed and apparently agreed to
5 days ago on hackers. That's not to say that it necessarily
shouldn't be reverted, but I think that will take some discussion (and
more than 1 vote).
Looking at it, I think I do agree that this shouldn't be a whole
chapter by itself. It's not nearly as high-level a concept as the
chapters that surround it. But I do think it's good to list all the
recovery.conf settings in one central place. The server config
chapter looks like the best fit to me on first glance, but I wonder
what others think?
...Robert
