Re: allow online change primary_conninfo

On 2020-03-28 02:39, Alvaro Herrera wrote:
> On 2020-Mar-27, Alvaro Herrera wrote:
> 
>> On 2020-Mar-27, Alvaro Herrera wrote:
>> 
>> > I pushed the wal_receiver_create_temp_slot bugfix, because I realized
>> > after looking for long enough at WalReceiverMain() that the code was
>> > beyond saving.  I'll be pushing the rest of this later today.
>> 
>> So here's the next one.  I still have to go over the doc changes here,
>> but at least the tests are passing for me.
> 
> Pushed with some documentation tweaks.
> 
> Many thanks for working on this!

Nice work! What do you think of possibility to add restore_command?

Is it a good idea to make a restore_command to be reloadable via SUGHUP 
too?

On the one hand it looks reasonable since primary_conninfo got such 
ability. On the other hand we don't exactly know whether the actual 
process after reload config would use "new" command or "old"  since it 
may take a long time to finish running old command (in case of scp or 
smth).

Also setting faulty restore_command lead to server termination, which is 
rather unexpectedly.
If anyone makes some minor typo in restore_command, say write 
restore_command='cp1 /mnt/srv/%f %p' instead of restore_command='cp 
/mnt/srv/%f %p' and do SELECT pg_reload_conf() then server will 
terminate.

Do you consider this feature useful? Any suggestions?

---
Best regards,
Maxim Orlov.