MSExchange TransportService Not starting - Event 1016

[patilp]

Exchange 2007 - Transport service not starting

I have two HT/CAs server / which one fine day the tarsport service stooped
and is not starting. The following two events are logged:

-------------------------------------------------------------------------
Event Type: Error
Event Source: MSExchange TransportService
Event Category: ProcessManager
Event ID: 1016
Date: 4/9/2008
Time: 4:01:51 PM
User: N/A
Computer: DBDAN60
Description:
The worker process crashes continuously on startup: C:\Program
Files\Microsoft\Exchange Server\Bin\edgetransport.exe. The service will be
stopped.

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.

Event Type: Error
Event Source: MSExchange Common
Event Category: General
Event ID: 4999
Date: 4/9/2008
Time: 4:01:50 PM
User: N/A
Computer: DBDAN60
Description:
Watson report about to be sent to dw20.exe for process id: 2428, with
parameters: E12, c-RTL-AMD64, 08.01.0263.000, edgetransport, M.E.Transport,
M.E.T.C.RoutingTables.CalculateTiServerRoutes, System.ArgumentException,
c249, 08.01.0263.000. ErrorReportingEnabled: False

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.
-------------------------------------------------------------------------
I have tried to enable logging to expert / high etc but i do not get any
more info. OUr mail routing is stuck

I try to remove and reinstall but the setup process complains about not able
to start the trasport service and fails.

Any insight inot what could be wrong or how ti get more info on what could
be happening -

TIA
--
Patilp

[John Fullbright]

http://support.microsoft.com/kb/944752/en-us

"patilp" <pat...@discussions.microsoft.com> wrote in message
news:4F76A9BF-DA31-40FF...@microsoft.com...

[patilp]

This did not help. The actual error is different.

In working with MS, we found the problem - The solution is listed below.

It seems like the transport service scans through a list of available
servers in the org and constructs a routing table (might be only in a interop
scenario). CUSTOMER has implemented an EMC Replistor based solution where I
believe it is required to create a virtual name and add it to the network
property in AD for both the server objects (production and recovery) / there
is a attribute called ‘NetworkAddress’ on the object servername in the
exchange org hierarchy under services in the config container.

Here CUSTOMER has added a virtual name called VEXCHANGE for both the server
objects (SERVER1 and SERVER2).

When the transport service is building its topology, (actually I believe the
edgetransport.exe process thread) it adds each server object in a table and
when it finds another one with the same name it throws an error called
“Exception Message=An item with the same key has already been added” and
stops the transport service.

When we removed the VEXCHANGE value for the recovery server through
adsiedit, the transport service was able to start just fine.

MS seems to think that this logic was only added in SP1 for Exchange 2007
and should not have worked from the beginning if this attribute had these
values. They are not 100% sure but are going to check.

--
Patilp

[John Fullbright]

I can't even contemplate speaking for EMC (I work for the competition). I
would contact them directly and ask for a patch or update to permanently
resolve the issue based on what MS has communicated to you.

In a multivendor solution, compatability issues become more complex. For
any vendor, before applyiing a service pack check the supportability matrix
for other vendors' products in your environment. It a good idea to have a
test or QA environment (even a virtual one) available to test updates as
well.

John

"patilp" <pat...@discussions.microsoft.com> wrote in message

news:86751B18-B8FE-4F1F...@microsoft.com...