Routebysender is not working.

@jdixon 

the default send connector is to a service that does email encryption. there are a few domains that dont use it so they go out a different service that is just filtering spam. 

<?xml version=”1.0″ encoding=”utf-8″?>
<configuration>
<startup>
<supportedRuntime version=”v4.0″ sku=”.NETFramework,Version=v4.5.1″/>
</startup>
<appSettings>
<add key=”DebugOn” value=”False”/>
<add key=”InternalRoutingOn” value=”True”/>
<add key=”GlobalRouteSameDomainOn” value=”False” />
<add key=”RoutingDomain” value=”nexthopdomain.com”/>
<add key=”BlockInternalOutOfOffice” value=”True”/>
<add key=”domain.com” value=”nospam.domain.com”/>
</appSettings>
</configuration>

Ok, in that config it should only route outbound if it is to another domain on your network. Check out this document: https://kb.knowmoreit.com/docs/integrations/route-by-sender/prepare-install/

I think you want to change it to:

<add key=”domain.com” value=”nospam.domain.com;true“/>

That should force all outbound to that send connector, but you need a smart host and not use MX lookups

@jdixon 

still doesnt seem to have fixed it. so basically my server is sending mail to two different smart hosts depending on the domain. everything no matter what tries to send to smart host a even if i try to put in the information here for them to use smart host b. 

Can you send me the output of “Get-Sendconnector | fl” and “Get-TransportAgent | fl”

It’s possible you may need to move up the transport agent in priority. Also enable debug mode in the config file so we can see if it logs anything to the event log.

You did deploy this on all transport servers right?

@jdixon 

the command didnt work the way had it typed.

hopefully this is what you are looking for. 

[PS] C:\Windows\system32>get-sendconnector

Identity AddressSpaces Enabled
——– ————- ——-
** {SMTP:*;1} True
*** {SMTP:*;1} True
*** {SMTP:*;1} True
** {SMTP:*;1} True
** {SMTP:*;1} True
*** {SMTP:*;1} True
*** {SMTP:*;1} True

[PS] C:\Windows\system32>get-transportagent

Identity Enabled Priority
——– ——- ——–
Transport Rule Agent True 1
DLP Policy Agent True 2
Retention Policy Agent True 3
Supervisory Review Agent True 4
Malware Agent True 5
Text Messaging Routing Agent True 6
Text Messaging Delivery Agent True 7
System Probe Drop Smtp Agent True 8
System Probe Drop Routing Agent True 9
CloudPanel Agent True 10

I commented out your domains but the problem is none of those send connectors are scoped to the custom domain. In your example above you had value “nospam.domain.com” and that needs to be scoped on the send connector. For example it shows {SMTP:*;1} above but it should show {SMTP:nospam.domain.com;1} for the one you want it to use.

@jdixon 

so i changed it. still doesnt seem to work.

my understanding is ultimately i should have two connectors one for each place i am sending out to and then scope the domain that uses which ever send connector. my understanding of the send connector is the destination, not the source. it would see this is telling the source. this is the first domain i am putting in that uses a different one and that domain is supposed to go live tuesday.

Identity AddressSpaces Enabled
——– ————- ——-
bca-architects.com {SMTP:*;1} True
streatorhome.com {SMTP:*;1} True
police.spring-valley.il.us {SMTP:*;1} True
gnbonline.com {SMTP:*;1} True
nospam.cpointcc.com {SMTP:leeman.cloud;1} True

@jdixon 

i have done this every which way possible. did services and full server reboots. your previous suggestion does not work or make sense. your route by sender is supposed to control where to route source domains is how i am to understand it. i have re ran through the install processes, verified everything looks like examples. had no responses today at all. not liking that a paid product only support is forum posts. 

This is from microsoft. the address space is for the destination domain not the source domain. 

The address space specifies the destination domains that are serviced by the Send connector. Mail is routed through a Send connector based on the domain of the recipient’s email address.

The purpose of the address space is the transport agent is overwriting the routing domain which is how it sends it to the correct send connector. It essentially is modifying the internal routing to make it think it is going to a different domain name which is why it still is based on the destination address.

Earlier I asked if you saw anything in the event viewer, did you see any messages? It sounds like we will need to schedule a remote session so I can login to take a look. We are in CST and if you can email me some available times that would work: jdixon@knowmoreit.com.

In this example configuration:

<?xml version="1.0" encoding="utf-8"?>
<configuration>
<startup>
<supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.5.1"/>
</startup>
<appSettings>
<add key="DebugOn" value="False"/>
<add key="InternalRoutingOn" value="True"/>
<add key="GlobalRouteSameDomainOn" value="False" />
<add key="RoutingDomain" value="nexthopdomain.com"/>
<add key="BlockInternalOutOfOffice" value="True"/>
<add key="customerdomain.com" value="custom-customerdomain.com;true"/>
</appSettings>
</configuration>

You should have two send connectors like this (you can have a fallback MX one scoped to * also if you want):

FirstSendConnector {SMTP:nexthopdomain.com;1} True

CustomerDomain {SMTP:custom-customerdomain.com;1} True

You can see the second one is scoped to “custom-customerdomain.com” which is a completely bogus domain name. I personally prefer just to prefix the customer’s actual domain with “custom-“. The “;true” is telling CloudPanel to send ALL emails outbound.

What is happening behind the scenes is when a user such as user@customerdomain.com sends an email to blah@knowmoreit.com, it is going to tell Exchange the destination for routing is actually “custom-customerdomain.com” instead of “knowmoreit.com”. This is how it forces it to go to the send connector scoped to “custom-customerdomain.com”.

If there is not a custom entry for a domain, then all emails that are sent from an accepted domain in your Exchange to a different accepted domain in your Exchange will have the routing overwritten to “nexthopdomain.com” and it will send it out that send connector instead.

I hope this makes better sense than the earlier explanations. We can still schedule time for a remote session though.

@jdixon 

re wrote the entire config file based on above explanation and example. whether needed or not but i wrote a send connector for every domain and set every domain up as the examples above and put every domain with their own key setup. restarted msexchangetransport and tested. it seems to be working. thanks for the reply late last night so i could do some more trials today when server has less active users. 

How has it been going?