4.7.500 Server busy Microsoft

[d3m0n]

Hello Community,

We are experiencing a very heavy throttling across 80% of our domains sending to Microsoft outlook in the last month.
We are members of Smart Network Data Services program (SNDS) and Junk Email Reporting Program (JMRP)
We send legit b2b corporate email not more then 3000 emails per day. there is no volume increase that could trigger
Microsoft support and they response with " I do not see anything offhand with the IP’s that would be preventing your mail from reaching our customers. "
yet we still get "smtp;451 4.7.500 Server busy that fill the queue and rotate there for hours until bounce"

smtp;451 4.7.500 Server busy. Please try again later from
[x.x.x.x]. (S77719) [SN1PEPF00026368.namprd02.prod.outlook.com
2023-10-31T01:07:33.490Z 08DBD852245041A3]
smtp;451 4.7.500 Server busy. Please try again later from
[x.x.x.x]. (S77719)

everything is set correctly in terms of DKMI, SPF , rDNS , we are hosted on , Hetzner and OVH.

abit frustrated and don't know where to look for clues .
does any one else has experienced or experiencing something similar ?
thx

 

[Dinaras]

it occurred to me lately and solution was to slow down a bit how fast i send emails to outlook from my smtp.
In my case i am using separate transport to send emails to big esp's. Problem arises when a custom domain uses as its mx a big ESP e.g outlook
In this case although you may be sending mails to outlook in a certain pace, more custom domains that use outlook might be in your list and this may cause outlook to reply like server busy try later
add a delay to your smtp between successive sends to outlook. That should fix it

 

[d3m0n]

@Dinaras
thank you very much for sharing experience , we have introduced speed throttling to 250 msg per hour per IP to outlook , even aded a "Pause after send" in mailwizz adding 2 seconds between messages

90% ips go to infinite loop with smtp;451 4.7.500 Server busy. Please try again later yet they all are "All of the specified IPs have normal status." in Smart Network Data Services program (SNDS)

 

[Dinaras]

exactly this is what i try ti explain. You throttled hotmail.com and outlook.com and maybe windowslive.com (all these domains are outlook).

Did you also throttle someotherdomain.com that has outlook as its email provider? same goes for yahoo and gmail. If outlook can barely bare your 250/hour msg/ip can it bare the rest of hidden outlook emails that you send on top? Do you throtle these or they just fall to your default transport?

i am sure your mta must have some sort of min and max backoff time for deferred messages? Increase that also if current setting is not enough.
In order to solve this problem i query mx domains and update my MTA transports so MTA knows how to keep the correct pace for each domain it delivers