On Sun, Aug 26, 2018 at 10:27 AM, Wietse Venema <wie...@porcupine.org> wrote: > Rodrigo Severo - F?brica: >> On Sun, Aug 26, 2018 at 9:29 AM, Wietse Venema <wie...@porcupine.org> wrote: >> > Rodrigo Severo - F?brica: >> >> On Sat, Aug 25, 2018 at 7:08 PM, Wietse Venema <wie...@porcupine.org> >> >> wrote: >> >> > Rodrigo Severo - F?brica: >> >> >> Hi, >> >> >> >> >> >> >> >> >> Is I set: >> >> >> >> >> >> smtp_destination_concurrency_limit = 2 >> >> >> smtp_destination_rate_delay = 1s >> >> >> >> >> >> and send several messages to the same domain will each connection send >> >> >> one message and wait 1 second (so I would have 2 messages delivered >> >> >> per second, one from each connection) or both connections to the same >> >> >> destination domain coordinate themselves and I would have only one >> >> >> message delivered per second to the destination domain? >> >> > >> >> > Instead of guessing, please read the documentation. >> >> > www.postfix.org/postconf.5.html#transport_destination_rate_delay >> >> >> >> I have read the documentation several times. The documentation states >> >> that "With a corresponding per-destination recipient limit > 1, the >> >> rate delay specifies the time between deliveries to the same domain." >> > >> > Delivery 1 starts, delivery 1 ends, N seconds delay, delivery 2 >> > starts, delivery 2 ends, N seconds delay, delivery 3 starts, delivery >> > 3 ends, and so on. >> >> My doubt is, having 2 simultaneous connections making deliveries to >> the same domain, this process you described happens in each >> simultaneous connection (doubling the delivery rate) or all >> simultaneous connections delivering to the same domain coordinate so, >> no matter how many simultaneous connections I have to the same domain, >> only one delivery will occur each N seconds? > > You are asking about the monster behind the trees. There is no such > monster. Postfix delivers mail exactly as described: > > Delivery 1 starts, delivery 1 ends. N seconds delay. Delivery 2 > starts, delivery 2 ends. N seconds delay. Delivery 3 starts, delivery > 3 ends, and so on.
I believe you mean that each simultaneous connection to the same destination will follow the procedure you described. Thanks for your clarification. I think that this info - the delay between deliveries is per connection and not per, or only, per domain - should be stated clearly in the documentation. Regards, Rodrigo Severo