On Sat, Jul 6, 2013 at 3:49 AM, Josh Berkus <josh(at)agliodbs(dot)com> wrote:
> Robert, Simon, All,
>
> On 04/01/2013 04:51 AM, Robert Haas wrote:> On Thu, Mar 28, 2013 at
> 11:48 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>>> a) recovery parameters are made into GUCs (for which we have a patch
>>> from Fujii)
>>> b) all processes automatically read recovery.conf as the last step in
>>> reading configuration files, if it exists, even if data_directory
>>> parameter is in use (attached patch)
>>> c) we trigger archive recovery by the presence of either
>>> recovery.conf or recovery.trigger in the data directory. At the end,
>>> we rename to recovery.done just as we do now. This means that any
>>> parameters put into recovery.conf will not be re-read when we SIGHUP
>>> after end of recovery. Note that recovery.trigger will NOT be read for
>>> parameters and is assumed to be zero-length.
>>> (minor patch required)
>>
>> I still prefer Greg Smith's proposal.
>>
>> http://www.postgresql.org/message-id/4EE91248.8010505@2ndQuadrant.com
>
> So, this seems to still be stalled. Is there a way forward on this
> which won't cause us to wait *another* version before we have working
> replication GUCs?
Yeah, it would be good to revive this thread now, which is the
beginning of the development cycle. As of now, just to recall
everybody, an agreement on this patch still needs to be found... Simon
is concerned with backward compatibility. Greg presented a nice spec
some time ago (Robert and I liked it) which would break backward
compatibility but allow to begin with a fresh infrastructure.
--
Michael