|
|
|
date: Thu, 31 Jan 2008 07:40:04 -0800,
group: microsoft.public.exchange.connectivity
back
Re: Exchange 2007 Email Routing
- The logical topology defined by AD site links is a layer on top of the
physical network. AD replication and Exchange 2007 routing use the logical
links defined.
- When defining AD Site Links, if you have a direct physical link between 2
Sites, it generally makes sense to have a logical Site link in AD as well.
- Having said that, even with the best path that logically traverses thru
another location/AD Site, the first attempt is always to connect directly to
the Hub Transport server(s) in the destination Site (unless a Hub Site is
defined, in which case all messages must go through the HT server(s) in that
Site).
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Bill Jenner" wrote in message
news:437E4A8C-18D4-4B40-A3BD-A8B07525836D@microsoft.com...
> Iâve been reading about how Exchange 2007 routes email and was hoping
> somebody could help clarify my understanding?
>
> I believe Exchange 2007 uses AD sites and the Inter-site Transport links
> defined between sites to determine how to route an email within an AD
> organization. According to the articles Iâve read it uses the costs
> defined
> on the links to determine the least cost path to the destination Hub
> server
> for the site where the mailbox server is located, but what actually
> happens
> if the defined Inter-Site links do not match the actual network links that
> physically exist?
>
> For example if there were three sites in an organization A, B, and C which
> are physically connected together in a triangle, but in AD sites there are
> only Inter-Site connections defined from A to B and then from B to C. If
> an
> email was going from site A to site C, although there is no link defined
> from
> A to C in Sites and Services, surely from a network perspective the data
> would be routed across the WAN link from A to C and not from A to B to C?
>
> Can anybody advise me on this
>
> Thanks,
> Bill
>
date: Thu, 31 Jan 2008 08:31:17 -0700
author: Bharat Suneja [MVP]
Re: Exchange 2007 Email Routing
That is how I understood it to work, but if there had not been a logical
connection defined between site A and C in my example then surely layer 3
network routing will determine how the data reaches the destination server(s)
and not the AD information?
"Bharat Suneja [MVP]" wrote:
> - The logical topology defined by AD site links is a layer on top of the
> physical network. AD replication and Exchange 2007 routing use the logical
> links defined.
> - When defining AD Site Links, if you have a direct physical link between 2
> Sites, it generally makes sense to have a logical Site link in AD as well.
> - Having said that, even with the best path that logically traverses thru
> another location/AD Site, the first attempt is always to connect directly to
> the Hub Transport server(s) in the destination Site (unless a Hub Site is
> defined, in which case all messages must go through the HT server(s) in that
> Site).
>
> --
> Bharat Suneja
> MVP - Exchange
> www.zenprise.com
> NEW blog location:
> exchangepedia.com/blog
> ----------------------------
>
>
>
> "Bill Jenner" wrote in message
> news:437E4A8C-18D4-4B40-A3BD-A8B07525836D@microsoft.com...
> > Iâve been reading about how Exchange 2007 routes email and was hoping
> > somebody could help clarify my understanding?
> >
> > I believe Exchange 2007 uses AD sites and the Inter-site Transport links
> > defined between sites to determine how to route an email within an AD
> > organization. According to the articles Iâve read it uses the costs
> > defined
> > on the links to determine the least cost path to the destination Hub
> > server
> > for the site where the mailbox server is located, but what actually
> > happens
> > if the defined Inter-Site links do not match the actual network links that
> > physically exist?
> >
> > For example if there were three sites in an organization A, B, and C which
> > are physically connected together in a triangle, but in AD sites there are
> > only Inter-Site connections defined from A to B and then from B to C. If
> > an
> > email was going from site A to site C, although there is no link defined
> > from
> > A to C in Sites and Services, surely from a network perspective the data
> > would be routed across the WAN link from A to C and not from A to B to C?
> >
> > Can anybody advise me on this
> >
> > Thanks,
> > Bill
> >
>
>
date: Thu, 31 Jan 2008 09:09:01 -0800
author: Bill Jenner
Re: Exchange 2007 Email Routing
- There is no direct connection between A and C, so based on AD topology HT
in A can get to HT in C through HT in B (best/only route).
- Since the first attempt is direct connect, yes - that will be routed using
the L3 network routing.
- If direct connect fails, messages will queue on HT in B.
- If HT in B is not reachable, messages will queue on HT in A.
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Bill Jenner" wrote in message
news:5ADB2A51-94B5-43CC-A921-578D41FD1610@microsoft.com...
> That is how I understood it to work, but if there had not been a logical
> connection defined between site A and C in my example then surely layer 3
> network routing will determine how the data reaches the destination
> server(s)
> and not the AD information?
>
> "Bharat Suneja [MVP]" wrote:
>
>> - The logical topology defined by AD site links is a layer on top of the
>> physical network. AD replication and Exchange 2007 routing use the
>> logical
>> links defined.
>> - When defining AD Site Links, if you have a direct physical link between
>> 2
>> Sites, it generally makes sense to have a logical Site link in AD as
>> well.
>> - Having said that, even with the best path that logically traverses thru
>> another location/AD Site, the first attempt is always to connect directly
>> to
>> the Hub Transport server(s) in the destination Site (unless a Hub Site is
>> defined, in which case all messages must go through the HT server(s) in
>> that
>> Site).
>>
>> --
>> Bharat Suneja
>> MVP - Exchange
>> www.zenprise.com
>> NEW blog location:
>> exchangepedia.com/blog
>> ----------------------------
>>
>>
>>
>> "Bill Jenner" wrote in message
>> news:437E4A8C-18D4-4B40-A3BD-A8B07525836D@microsoft.com...
>> > Iâve been reading about how Exchange 2007 routes email and was hoping
>> > somebody could help clarify my understanding?
>> >
>> > I believe Exchange 2007 uses AD sites and the Inter-site Transport
>> > links
>> > defined between sites to determine how to route an email within an AD
>> > organization. According to the articles Iâve read it uses the costs
>> > defined
>> > on the links to determine the least cost path to the destination Hub
>> > server
>> > for the site where the mailbox server is located, but what actually
>> > happens
>> > if the defined Inter-Site links do not match the actual network links
>> > that
>> > physically exist?
>> >
>> > For example if there were three sites in an organization A, B, and C
>> > which
>> > are physically connected together in a triangle, but in AD sites there
>> > are
>> > only Inter-Site connections defined from A to B and then from B to C.
>> > If
>> > an
>> > email was going from site A to site C, although there is no link
>> > defined
>> > from
>> > A to C in Sites and Services, surely from a network perspective the
>> > data
>> > would be routed across the WAN link from A to C and not from A to B to
>> > C?
>> >
>> > Can anybody advise me on this
>> >
>> > Thanks,
>> > Bill
>> >
>>
>>
date: Thu, 31 Jan 2008 16:01:21 -0800
author: Bharat Suneja [MVP]
|
|
