Re: allow online change primary_conninfo

Hi

>>  > I think we unfortunately got to mark this as returned with
>>  > feedback. I've not done so, but just switched the entry to waiting on
>>  > author.
>>
>>  Why returned with feedback? Why waiting on author? I didn't receive a
>>  feedback for latest published patch version. What can I do as author?
>>  Patch still applied (thanks cf bot) Obviously too late for pg12, but
>>  why can not be target pg13 and therefore be moved to next CF?
>
> Well, my impression was that the patch didn't yet really address the
> feedback. And thus should have been marked as waiting on author for a
> while.

Not agree. Latest patch version perform walreceiver restart without switch to a different method as discussed. Here is no race condition between startup process and walreceiver because conninfo passed via WalRcvData struct as currently. I miss something important?
Michael Paquier had no possibility to review latest implementation, but did not say this is totally wrong, just asked wait a rather close lookup.

Of cource we can close this cf entry. I would be happy if someone else post proper implementation. And I can rework my implementation again, but I don’t know how the correct implementation should look or why latest implementation is wrong.

regards, Sergei