Expand my Community achievements bar.

Announcing the launch of new sub-community for Campaign Web UI to cater specifically to the needs of Campaign Web UI users!
SOLVED

Single IP address per Affinity and MTA - dual email delivery issue

Avatar

Level 3

Masters of Adobe Campaign,

 

I appreciate the past hints and guidelines, and as I am discovering Adobe Campaign Classic and its possibilities I have one more question regarding Affinity and IP address setup.

 

We have a few IP addresses and affinities providing redundancy for our email sends. This pertains to my last question on the dual email sent to the same recipient and SQL deadlocks - logs show only one delivery despite recipients' complaints about receiving 2 emails. 

 

a) do you keep one dedicated IP address per affinity?
b) does it make a difference? Will it change the issue with a dual email sent?
c) also, are the SQL deadlocks a possible reason for dual emails sent in case MTA uses more than one IP address and affinities for redundancy?

 

I could not find a clear answer to the above, and perhaps some of you occurred to have similar experiences or questions.


Anything to help understand the pros and cons of a single IP address/affinity per MTA will help out.

 

Thank you very much in advance.

1 Accepted Solution

Avatar

Correct answer by
Level 2

Hi 

 

a) do you keep one dedicated IP address per affinity? - Not necessarily, it depends on how many you have requested/is included in your contract. You may have multiple IPs associated to same affinity as well. 

 

NOTE : For more details now how IP Affinities are set read here . 

b) does it make a difference? Will it change the issue with a dual email sent? - No, Dual email sent as explained in the other post here, it could be related to delivery settings or exclusions not happening on delivery level. You must check the delivery settings to get this issue rectified. 

c) also, are the SQL deadlocks a possible reason for dual emails sent in case MTA uses more than one IP address and affinities for redundancy? SQL Dead locks has no relation with IP/Affinities. 

View solution in original post

2 Replies

Avatar

Correct answer by
Level 2

Hi 

 

a) do you keep one dedicated IP address per affinity? - Not necessarily, it depends on how many you have requested/is included in your contract. You may have multiple IPs associated to same affinity as well. 

 

NOTE : For more details now how IP Affinities are set read here . 

b) does it make a difference? Will it change the issue with a dual email sent? - No, Dual email sent as explained in the other post here, it could be related to delivery settings or exclusions not happening on delivery level. You must check the delivery settings to get this issue rectified. 

c) also, are the SQL deadlocks a possible reason for dual emails sent in case MTA uses more than one IP address and affinities for redundancy? SQL Dead locks has no relation with IP/Affinities. 

Avatar

Level 6

Never setup 1 IP per affinity... You need at least 2 per affinity
An "affinity" is a dedicated pool of IPs that will be used for dedicated sendings (see

 doc from Anshar)

Why never using just one? In case you have a deliverability issue with it, you'll then lose all your delivery messages.
You have to calculate the amount of IPs you'll need dedicated per use.
For example, imagine you have some sendings dedicated for China and others dedicated for Europe.
You'll then need at least 2 affinities: one for China, one for Eu.
Generally speaking, an IP can deliver up to 150K emails per hour with a really good warm up for reputation and following deliverability best practices.
As china ISPs are different from Europe ones, you'll then prefer to dedicate a pool of IP specially for China and another one specially for Europe. To isolate those both pools will permit you to isolate also any issue that occurs on one pool from the other and let you warm up efficiently your IP reputation.
But now, you know that you send 2 millions of emails between 10am and 12am for China and 1 million for Europe.
It means, that you'll then need 1 million per hour for China (so 7 ips) and 500K for Europe (so 4 ips)
you'll then create your China affinity containing 7 ips and Europe affinity with only 4

if you want to know which Ips had mostly an issue (generally this is not just one but the all pool), you can extend the broadlogRcp schema in order to add the "publicId" that you'll declare in the config-instance.xml file