Outlook won't send e-mail to one paticular contact

J

Jasper

When we try to send an email to this contact she does not receive it
but she has no problem sending to us.

No errors come up and it looks like it was sent.

Thanks for any help on this.
 
V

VanguardLH

Jasper said:
When we try to send an email to this contact she does not receive it
but she has no problem sending to us.

No errors come up and it looks like it was sent.

What has the recipient tried so far (so we don't end up duplicating
suggestions to do the same actions that she already performed)? Looked
in the trash or spam folders? Disabled or deleted any server-side rules
or filters in her account? Sees e-mails from other senders okay but
just not you? Did she disable her anti-spam program or anything else
that interrogates her e-mails?

What have you as the sender tried so far (so we don't end up duplicating
suggestions to do the same actions you already performed)? Send from a
different domain? BCC'ed yourself (and at a *different* domain than
from where you send) to ensure your e-mail provider's SMTP server is
actually sending the e-mails (and not just simply accepting them from
you)? Did you try disabling your anti-spam, anti-virus, or any other
program that interrogates your outbound e-mails?

Some anti-virus programs act as a transparent proxy. They interrogate
your outbound e-mail traffic byte-by-byte as it passes through their
proxy. That's how they interrogate its content to make sure your
outbound e-mails aren't infected. However, some AV products pretend
they are your SMTP server. They don't operate as a proxy but instead as
a mail server. They accept your complete outbound e-mail which means
your client gets a good status back from the phony SMTP server (the one
used by the AV program). Then the AV program inspects the entire e-mail
message for malware. If it's okay then it pretends it is your e-mail
client and connects to your real SMTP server to finally transmit your
outbound e-mail. If there is an error between the AV program's bogus
SMTP server and the real SMTP server, your e-mail client won't ever know
that because it was NOT involved in the mail session with your real SMTP
server. Only the logs in your AV program might show there was an error
between its intercepting SMTP server and the real SMTP server. So you
won't see a problem in your e-mail client because the AV program's SMTP
server is always returning an OK status on every e-mail you send from
your client. You have to inspect the AV program (or anti-spam program
or whatever intercepts your outbound e-mails) to see if its logs report
a problem. You might then find there was an invalid domain or undefined
account (username) reported by the receiving SMTP server but only your
AV program's bogus SMTP server would know about that status.

You have a mail session between your client and the AV program's SMTP
server and it sends back an OK status. Your client is now out of the
picture an no longer involved. Your AV program's SMTP server after
interrogating your message then starts another mail session between it
and the real SMTP server. That 2nd mail session doesn't involve your
e-mail client. That's why AV, anti-spam, or other e-mail interrogating
software that operates as a proxy to immediately pass on the e-mail
traffic so your e-mail client gets status from the real SMTP server are
preferred. Of course, if your AV program was working well then you
never need to have it inspect your outbound e-mails. After all, if it
can find the pest on your host then how could it possible not already be
detected by the time you attach it to an e-mail?

Do you use the recipient's e-mails to define the reply-to e-mail address
or do you manually type it into your new outbound e-mails to her? She
might have the wrong e-mail address in her From or Reply-To headers so
you end up recording (in your contacts) and sending to the wrong e-mail
address on your end. Just because her e-mails get delivered to you
doesn't mean she specified the correct e-mail address for herself in
those e-mails you got from here. That you don't get an error
immediately simply means your SMTP server accepted your outbound e-mail.
It doesn't mandate the recipient's e-mail address is valid. Not until
your SMTP server actually sends the e-mail will it know if the domain is
valid and if there is an account with the specified user name at that
domain.

If the domain was invalid then your SMTP server would send back an NDR
(non-delivery report). Normally (in a good setup) the sending SMTP
server gets an immediate error from the receiving SMTP server if the
account doesn't exist at the destination. That is, *during* the mail
session between sending and receiving servers an error is sent by the
receiving server and the sending server immediately sends you back an
NDR (but *not* during your mail session with your SMTP server but later
during the mail session with the other SMTP server). So it's quite
possible that your SMTP server cannot send (invalid domain) or gets an
error (no such account there) and sends back an NDR but *you* are
filtering out those status e-mails. That means you need to check your
server- and client-side filtering to make sure you get the NDRs.

Some mail servers are [deliberately] misconfigured, especially those
that merely forward e-mails rather than used as a normal e-mail account.
When they are the receiving server and instead of immediately sending
back an ERR status when no such account is defined on their end, they
merely accept all e-mails. After the mail session is between the
sending and receiving mail servers, the receiving mail server then
decides later to process the received e-mail whereupon then the check is
made to determine if the recipient is defined there. When they find
there is no such user, the user's quota is exceeded (they used up their
disk space for their account), the size of the message exceeds the max
permitted for the recipient's account, or some other problem, they send
back an NDR to the sender. However, at that point, the receiving server
has no idea who was the real sender. All they have is what the sender
claimed was their e-mail address - and we all know that spammers don't
specify their real e-mail addresses. During the mail session, the
receiving mail server can send back status immediately to the sending
mail server without ever knowing the originating e-mail address. It
doesn't care. The sending server is supposed to know which account was
used at it to send the message and issues an NDR to its user's account.
But after the mail session is over, an NDR only has what the sender said
was their e-mail address and the NDR gets sent there. So maybe you are
specifying a From, Reply-To, or Return-Path header with an e-mail
address that doesn't point back to the account where you expect the NDRs
to get delivered, or your return e-mail address is invalid so the
receiving mail server tries to send the NDR to an invalid address which
fails but obviously it has nowhere else to try to send the NDR e-mail.

You need to look at what is in the headers of your outbound e-mails.
You also need to check that your real SMTP server is actually sending
your outbound e-mails. Sending e-mails to yourself (i.e., you send from
your e-mail provider to your own account at that same e-mail provider)
is not a valid test. Most e-mail services will short-circuit e-mails
sent from yourself to yourself. They simply redirect your message back
into your account without ever touching their SMTP server. In fact,
some providers will automatically filter out such e-mails since few
users send e-mails to themselves (e.g., Gmail filters out e-mails sent
by you to yourself at Gmail). You might have your own client- or
server-side filters that do the same. To ensure your SMTP server is
actually sending out your e-mails, send a test e-mail to the problematic
recipient while also including yourself as a BCC recipient but where the
target e-mail address for yourself is at some OTHER e-mail provider.
Saying you send e-mails using your Hotmail account. BCC yourself in the
test e-mail to your Gmail account. Then if you see your test e-mail
from Hotmail get delivered to your Gmail account then you can be sure
that your Hotmail SMTP server actually sent out your test e-mail, plus
you can inspect the headers (as they will be full headers rather than
some internal redirection from your account back into your account).
 
J

Jasper

"VanguardLH" wrote in message
When we try to send an email to this contact she does not receive it
but she has no problem sending to us.

No errors come up and it looks like it was sent.

What has the recipient tried so far (so we don't end up duplicating
suggestions to do the same actions that she already performed)? Looked
in the trash or spam folders? Disabled or deleted any server-side rules
or filters in her account? Sees e-mails from other senders okay but
just not you? Did she disable her anti-spam program or anything else
that interrogates her e-mails?

What have you as the sender tried so far (so we don't end up duplicating
suggestions to do the same actions you already performed)? Send from a
different domain? BCC'ed yourself (and at a *different* domain than
from where you send) to ensure your e-mail provider's SMTP server is
actually sending the e-mails (and not just simply accepting them from
you)? Did you try disabling your anti-spam, anti-virus, or any other
program that interrogates your outbound e-mails?

Some anti-virus programs act as a transparent proxy. They interrogate
your outbound e-mail traffic byte-by-byte as it passes through their
proxy. That's how they interrogate its content to make sure your
outbound e-mails aren't infected. However, some AV products pretend
they are your SMTP server. They don't operate as a proxy but instead as
a mail server. They accept your complete outbound e-mail which means
your client gets a good status back from the phony SMTP server (the one
used by the AV program). Then the AV program inspects the entire e-mail
message for malware. If it's okay then it pretends it is your e-mail
client and connects to your real SMTP server to finally transmit your
outbound e-mail. If there is an error between the AV program's bogus
SMTP server and the real SMTP server, your e-mail client won't ever know
that because it was NOT involved in the mail session with your real SMTP
server. Only the logs in your AV program might show there was an error
between its intercepting SMTP server and the real SMTP server. So you
won't see a problem in your e-mail client because the AV program's SMTP
server is always returning an OK status on every e-mail you send from
your client. You have to inspect the AV program (or anti-spam program
or whatever intercepts your outbound e-mails) to see if its logs report
a problem. You might then find there was an invalid domain or undefined
account (username) reported by the receiving SMTP server but only your
AV program's bogus SMTP server would know about that status.

You have a mail session between your client and the AV program's SMTP
server and it sends back an OK status. Your client is now out of the
picture an no longer involved. Your AV program's SMTP server after
interrogating your message then starts another mail session between it
and the real SMTP server. That 2nd mail session doesn't involve your
e-mail client. That's why AV, anti-spam, or other e-mail interrogating
software that operates as a proxy to immediately pass on the e-mail
traffic so your e-mail client gets status from the real SMTP server are
preferred. Of course, if your AV program was working well then you
never need to have it inspect your outbound e-mails. After all, if it
can find the pest on your host then how could it possible not already be
detected by the time you attach it to an e-mail?

Do you use the recipient's e-mails to define the reply-to e-mail address
or do you manually type it into your new outbound e-mails to her? She
might have the wrong e-mail address in her From or Reply-To headers so
you end up recording (in your contacts) and sending to the wrong e-mail
address on your end. Just because her e-mails get delivered to you
doesn't mean she specified the correct e-mail address for herself in
those e-mails you got from here. That you don't get an error
immediately simply means your SMTP server accepted your outbound e-mail.
It doesn't mandate the recipient's e-mail address is valid. Not until
your SMTP server actually sends the e-mail will it know if the domain is
valid and if there is an account with the specified user name at that
domain.

If the domain was invalid then your SMTP server would send back an NDR
(non-delivery report). Normally (in a good setup) the sending SMTP
server gets an immediate error from the receiving SMTP server if the
account doesn't exist at the destination. That is, *during* the mail
session between sending and receiving servers an error is sent by the
receiving server and the sending server immediately sends you back an
NDR (but *not* during your mail session with your SMTP server but later
during the mail session with the other SMTP server). So it's quite
possible that your SMTP server cannot send (invalid domain) or gets an
error (no such account there) and sends back an NDR but *you* are
filtering out those status e-mails. That means you need to check your
server- and client-side filtering to make sure you get the NDRs.

Some mail servers are [deliberately] misconfigured, especially those
that merely forward e-mails rather than used as a normal e-mail account.
When they are the receiving server and instead of immediately sending
back an ERR status when no such account is defined on their end, they
merely accept all e-mails. After the mail session is between the
sending and receiving mail servers, the receiving mail server then
decides later to process the received e-mail whereupon then the check is
made to determine if the recipient is defined there. When they find
there is no such user, the user's quota is exceeded (they used up their
disk space for their account), the size of the message exceeds the max
permitted for the recipient's account, or some other problem, they send
back an NDR to the sender. However, at that point, the receiving server
has no idea who was the real sender. All they have is what the sender
claimed was their e-mail address - and we all know that spammers don't
specify their real e-mail addresses. During the mail session, the
receiving mail server can send back status immediately to the sending
mail server without ever knowing the originating e-mail address. It
doesn't care. The sending server is supposed to know which account was
used at it to send the message and issues an NDR to its user's account.
But after the mail session is over, an NDR only has what the sender said
was their e-mail address and the NDR gets sent there. So maybe you are
specifying a From, Reply-To, or Return-Path header with an e-mail
address that doesn't point back to the account where you expect the NDRs
to get delivered, or your return e-mail address is invalid so the
receiving mail server tries to send the NDR to an invalid address which
fails but obviously it has nowhere else to try to send the NDR e-mail.

You need to look at what is in the headers of your outbound e-mails.
You also need to check that your real SMTP server is actually sending
your outbound e-mails. Sending e-mails to yourself (i.e., you send from
your e-mail provider to your own account at that same e-mail provider)
is not a valid test. Most e-mail services will short-circuit e-mails
sent from yourself to yourself. They simply redirect your message back
into your account without ever touching their SMTP server. In fact,
some providers will automatically filter out such e-mails since few
users send e-mails to themselves (e.g., Gmail filters out e-mails sent
by you to yourself at Gmail). You might have your own client- or
server-side filters that do the same. To ensure your SMTP server is
actually sending out your e-mails, send a test e-mail to the problematic
recipient while also including yourself as a BCC recipient but where the
target e-mail address for yourself is at some OTHER e-mail provider.
Saying you send e-mails using your Hotmail account. BCC yourself in the
test e-mail to your Gmail account. Then if you see your test e-mail
from Hotmail get delivered to your Gmail account then you can be sure
that your Hotmail SMTP server actually sent out your test e-mail, plus
you can inspect the headers (as they will be full headers rather than
some internal redirection from your account back into your account).

On our side we have tried to reply, forward and manually type in the email
address. Other people in our company send emails to her without a problem.
I have sent a test message from our email service's webmail and it also
fails.

We do not get any indication that the email has failed (no NDR) and the
outbox is empty.

She has had her ISP check into this and they say it is not on her end.

We use a pop/smtp server hosted by USA.net they have checked their filters
and found no problems on that side. I have our email accounts set up to send
via USA.net's
smtp not our ISP's. I did try to use our ISP's smtp server but the result is
the same.

But here is another interesting thing, He can reply to her from his phone
without a problem
the phone is setup the same as our internal email clients.
 
V

VanguardLH

Note: Due to use of WLM v15 by Jasper with its loss of proper quoting,
and to trim the size of my reply, I deleted the non-quoted portion of
Jasper's reply which was my content.
On our side we have tried to reply, forward and manually type in the email
address. Other people in our company send emails to her without a problem.
I have sent a test message from our email service's webmail and it also
fails.

We do not get any indication that the email has failed (no NDR) and the
outbox is empty.

She has had her ISP check into this and they say it is not on her end.

We use a pop/smtp server hosted by USA.net they have checked their filters
and found no problems on that side. I have our email accounts set up to send
via USA.net's
smtp not our ISP's. I did try to use our ISP's smtp server but the result is
the same.

But here is another interesting thing, He can reply to her from his phone
without a problem
the phone is setup the same as our internal email clients.

Note to Jasper: Because you are using WLM v15 which lost the ability to
properly quote & indented the cited content of the parent post to which
you reply, please either do the quoting yourself (indent the lines with
the ">" quoting character) or provide some delimiter line between what
you quoted and what is your new content.

If others in your same company (all using the same Exchange or SMTP
server as yourself) can get e-mails successfully delivered to the same
recipient then the problem is not your domain being blacklisted at the
recipient's mail server but it could be your username is being filter by
the recipient's service or by the recipient. If someoneElse@yourDomain
can get their e-mails delivered to her but you@yourDomain fails then the
problem is with your account or with filtering on their end.

You'll get an NDR if: (1) The receiving mail server checks *during* the
mail session with your sending mail server that the destination account
exists and, if not, issues an ERR status to your sending mail server who
then gives you the NDR message; or, (2) The receiving mail server
accepts all e-mails without doing any checks during its mail session
with your sending mail server, finds out later your message is
undeliverable, and then uses your headers to decide where to mail back
the NDR message (so the return-path headers in your e-mail have to
correctly point back to you; else, the NDR goes elsewhere or fails to
get delivered). In case #1, not getting an NDR means the receiving
server found the recipient's account okay and accepted your message to
deliver there. In case #2, you need to make sure your return headers
point to you to get the NDR.

Negative confirmation via NDR is how most e-mail servers function. They
figure that telling you that a message was undeliverable equates to
implying your e-mail was deliverable because you didn't get an NDR. The
number of undeliverable e-mails is far smaller than the number of
deliverable e-mails so they consume less resources sending negative
feedback regarding failures. You can try to use positive feedback by
configuring your e-mail client to either request a read receipt from the
recipient (most users disable this in their e-mail clients because
spammers use this trick to harvest active e-mail addresses) or request
delivery confirmation. The former has the *recipient* (their e-mail
client) send you confirmation when they open your message. The latter
only requests the receiving mail server send back a confirmation
message. Most mail servers ignore requests for delivery confirmation
because if they didn't error during the mail session or don't later send
an NDR then the lack of failure notification is considered sufficient
positive feedback. So while you can ask for a read receipt, the
recipient may not send you one. While you can ask the recipient's mail
server to confirm your message got to that server, very few bother with
sending that positive feedback. Most mail servers just send the lesser
number of negative feedback via NDRs.

That your e-mail provider indicates by status to your e-mail client or
to their webmail client that they sent your e-mail, all that means is
that they *accepted* your e-mail which they will send later (perhaps in
a couple seconds). The mail session you have between you (or your
client) and their server is NOT the mail session their server has with
the recipient's mail server. Those are two separate mail sessions. So
them giving you an OK status only means they accepted your message and
your request to send it out. That's why I mention BCC'ing yourself to a
*different* e-mail provider to ensure that they really are sending out
your message.

Since the e-mail app on the smart phone works but not your e-mail client
on a workstation, I have to wonder what additional content your sender
is adding to their outbound e-mails. Perhaps you are adding a spammy
signature or a huge-sized one (i.e., bloated with binary content, like
images). Your company my be adding some "confidential" or legal
disclaimer text (which is not enforceable in court since the disclaimer
is not sent previously and separately of the content being disclaimed or
restricted) to all their outbound e-mails. That could result in the
majority of each message being this appended text with little in the
body that is new content. One way to deflect spam is to look for a
large number of incoming messages that have nearly similar content. The
spammer uses a template for their content so a domain sees a lot of
incoming e-mails that have nearly identical content.

http://www.rhyolite.com/dcc/

I've used anti-spam clients in the past that can make use of the DCC
database. Users of DCC send in a hash code of e-mails that they
receive. Fuzzy logic determines how similar are the received e-mails.
The DCC database gets updated to reflect how many users are getting the
nearly-same e-mails. A user of the DCC filter can set a threshold so,
for example, if they don't want to receive e-mails that have already
been received by another 100 recipients then that e-mail gets block.
That the same e-mail was mass mailed to lots of recipients qualifies it
as spam; however, there are many companies that use templates in their
e-mails that will trigger this similarity test. So check if anything is
getting consistently added to your outbound e-mails, like signatures
(with images) or disclaimer/restriction notices, that is always the same
an could comprise a majority of the content of your e-mails and if
you're doing this to lots of recipients at the same domain.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top