I have just introduced Exchange 2010 into a Ex2003/2007 environment. Subsequently, one of the users who was migrated from 2003 to 2010 was was having problem receiving email from a particular external sender. The email content was in garbage code. After some checking, it was found that the contents (displayed in Outlook and OWA) was actually Base64 coded. However from the email header, I found this "Content-Transfer-Encoding: quoted-printable". So it appears that the header says that the email should be quoted-printable coded and thus exchange does not convert it to readable text? Anyone encountered a same/similar problem?

this is the email header From: "someone" <someone@somedomain.com> To: recipient <recipient@recipientdomain.com> Subject: Some Subject Thread-Topic: Some Subject Thread-Index: AQHL+jqYh6CaQ4EDrkq+hPMiepxbgA== Date: Thu, 14 Apr 2011 00:28:56 +0000 Message-ID: <b669d7d30002185c@somedomain.com> Content-Language: en-SG X-MS-Has-Attach: X-MS-Exchange-Organization-SCL: 0 X-MS-TNEF-Correlator: x-originalarrivaltime: 14 Apr 2011 00:28:56.0859 (UTC) FILETIME=[F100A6B0:01CBFA3A] x-tm-as-result: No--28.402700-8.000000-31 x-tm-as-product-ver: SMEX-10.0.0.1412-6.500.1024-18072.001 x-tm-as-user-blocked-sender: No x-tm-as-user-approved-sender: No x-imss-scan-details: No--13.503-5.0-31-1 x-tm-imss-message-id: <b669d7d30002185c@somedomain.com> Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0

On Mon, 30 May 2011 06:16:40 +0000, hayami.wai wrote: > > >this is the email header > > > >From: "someone" <someone@somedomain.com> >To: recipient <recipient@recipientdomain.com> >Subject: Some Subject >Thread-Topic: Some Subject >Thread-Index: AQHL+jqYh6CaQ4EDrkq+hPMiepxbgA== >Date: Thu, 14 Apr 2011 00:28:56 +0000 >Message-ID: <b669d7d30002185c@somedomain.com> >Content-Language: en-SG >X-MS-Has-Attach: >X-MS-Exchange-Organization-SCL: 0 >X-MS-TNEF-Correlator: >x-originalarrivaltime: 14 Apr 2011 00:28:56.0859 (UTC) > FILETIME=[F100A6B0:01CBFA3A] >x-tm-as-result: No--28.402700-8.000000-31 >x-tm-as-product-ver: SMEX-10.0.0.1412-6.500.1024-18072.001 >x-tm-as-user-blocked-sender: No >x-tm-as-user-approved-sender: No >x-imss-scan-details: No--13.503-5.0-31-1 >x-tm-imss-message-id: <b669d7d30002185c@somedomain.com> >Content-Type: text/plain; charset="iso-8859-1" >Content-Transfer-Encoding: quoted-printable >MIME-Version: 1.0 Enable pipeline tracing and content conversion tracing. You'll see the message in its original form. http://technet.microsoft.com/en-us/library/bb125018.aspx http://technet.microsoft.com/en-us/library/bb397226.aspx Since the message is delivered I don't think there's any content conversion taking place. I think the message really contains base64-encoded data and the client is either getting it delivered as quoted-printable or as plain text as the message preamble. FWIW, the MIME headers in the message headers may, or may not, be correct. Are there other MIME body parts in the message or is the message supposed to be interpreted as being composed only of the MIME preamble (i.e. text that's supposed to be accessible to clients that don't understand MIME -- it's what you see between the 1st blank line in the message and the first MIME body part)? Exchange 2003 (and earlier) allowed an awful lot of badly formed MIME (and UUENCODE) messages to be delivered to users. --- Rich Matheisen MCSE+I, Exchange MVP --- Rich Matheisen MCSE+I, Exchange MVP

How the inbound mail flow looks like? Internet->Exchange 2010 Edge->Exchange 2010 Hub? Does the symptom happen to all internal exchange 2010 mailbox users when the mail comes from this specific sender? Please set the value of “ByteEncoderTypeFor7BitCharsets” to 0 on the Hub server, and check the issue again Change the Method for 7-bit Transfer EncodingPlease remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.

Hi, thanks for the reply. I tried decoding the Base64 body and I do find many some email header appearing from within the email body. Which is now another mystery why the headers appears in the body. Below is a sample of the body of the email Message-Id: <eventum.4m8phyprc.dg3l88@senderdomain.com> Content-Type: multipart/alternative; boundary="=_f23c68dee193aa7e33e16d24a17b60e5" From: "Sender" <email@senderdomain.com> Reply-To: "Sender" <email@senderdomain.com> Cc: recipient_1 <recipient_1@recipientdomain.com>, recipient_2 <recipient_2@recipientdomain.com> Subject: <Subject> Date: Thu, 14 Apr 2011 00:25:32 +0000 <<<<< some contents >>>>> --=_f23c68dee193aa7e33e16d24a17b60e5 Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable <<<<< some contents >>>>> --=_f23c68dee193aa7e33e16d24a17b60e5-- For the same email, a user who is in Ex2003 does not have any problem with it. Can I get some idea on what i will be expecting from the pipeline tracing? Will I be able to trace past emails or it is a real time tracing?

There is no Edge server, its a all in one box with CAS, HUB and Mailbox. The mail flow will be Internet -> Exchange 2003 -> Exchange 2010 May I know the effect of changing the method for transfer encoding? I have read the technet article, but what am I going to expect on the client side?

On Tue, 31 May 2011 03:53:34 +0000, hayami.wai wrote: >I tried decoding the Base64 body and I do find many some email header appearing from within the email body. Which is now another mystery why the headers appears in the body. Probably because the message body includes other messages that are part of an e-mail thread. >Below is a sample of the body of the email > > >Message-Id: <eventum.4m8phyprc.dg3l88@senderdomain.com> >Content-Type: multipart/alternative; boundary="=_f23c68dee193aa7e33e16d24a17b60e5" >From: "Sender" <email@senderdomain.com> >Reply-To: "Sender" <email@senderdomain.com> >Cc: recipient_1 <recipient_1@recipientdomain.com>, recipient_2 <recipient_2@recipientdomain.com> >Subject: <Subject> >Date: Thu, 14 Apr 2011 00:25:32 +0000 > ><<<<< some contents >>>>> > >--=_f23c68dee193aa7e33e16d24a17b60e5 >Content-Type: text/html; charset="ISO-8859-1" >Content-Transfer-Encoding: quoted-printable > ><<<<< some contents >>>>> > >--=_f23c68dee193aa7e33e16d24a17b60e5-- There isn't much useful information in that. However, if the MIME body part is supposed to be interpreted as "quoted-printable" and it contains base64 encoded data I'd expect base64 encoded data to be what the user sees -- IOW, a mess. Most of the time users complain of seeing "garbled data" when this happens. >For the same email, a user who is in Ex2003 does not have any problem with it. Without the original message it's impossible to know why that would be. >Can I get some idea on what i will be expecting from the pipeline tracing? Will I be able to trace past emails or it is a real time tracing? You'll only see new messages (one copy for each part of the transport pipeline), and only for the address of the sender that you specify. --- Rich Matheisen MCSE+I, Exchange MVP --- Rich Matheisen MCSE+I, Exchange MVP

Quote: “May I know the effect of changing the method for transfer encoding?” The behavior should be the same as in exchange 2003 “The default value of 0 for 7-bit encoding is used if the registry subkey value contains a value that is not listed in this table or if the TransferEncodingFor7bit registry entry is not present” ----------Refer to <An Exchange Server 2003 recipient may receive a non-delivery report or an e-mail message that is not formatted correctly>Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.