Go Premium for a chance to win a PS4. Enter to Win

  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1337
  • Last Modified:

MTA Queues showing not available in 4 Node cluster

4 Compaq DL740's
Fibre attaced to HP EVA
3 Active Nodes/1 Passive
Windows 2003 Enterprise
Exchange 2003 Enterprise SP1
32TB of Storage available
Each node has 4SG's with 5DB's each
Currently in the progress of upgrading from Exchange 5.5
Antigen 7.5 SR3

When installing Exchange on a cluster, the first node in owns the MTA Instance.  It is responsible for mail flow to and from the 5.5 servers.  We moved the first mailbox to the cluster yesterday for testing purposes, and no messages are flowing between the Exchange 2003 servers and the Exchange 5.5 servers.  Upon investigation we found the MTA queue showing as Not Available on the first node.  We checked the service and it is started and showing Online in Cluster Admin.  We then failed the Virtual server to another node and the messages that we were expecting came through.  We then tried sending messages again and they did not come through.  Whenever we restart the service or fail it over to another node, the messages all come through, but after it has started no mail flows.  We get the following error in the event log:

Event Type:     Warning
Event Source:     MSExchangeIS Mailbox Store
Event Category:     MTA Connections
Event ID:     2000
Date:          8/4/2004
Time:          10:42:45 AM
User:          N/A
Computer:     GNBEXNODE01
Verify that the Microsoft Exchange MTA service has started. Consecutive ma-open calls are failing with error 3051.

For more information, click http://www.microsoft.com/contentredirect.asp.

I have looked this error up and most people indicate a lack of drive space, which is not the problem here.  The drive with the MTA on it has 99.9GB free space.  We have run the MTACheck util and everything comes back fine with it.

We have tried moving the MTA to a different physical drive, and this made no difference.

Mail flow between the 2003 servers is fine obviously because they use SMTP.  

We also have another 2 node cluster that was setup the exact same way, and everything is fine on it.  All the servers are fresh installs with all the latest Windows and Exchange service packs and hotfixes.

Any help would be appreciated.
  • 3
1 Solution
The error you gave just means that the MTA is no longer running, it doesn't really say why.  It almost sounds like something is causing the MTA to silently exit.  Given that this usually happens a minute or two after starting, its probably either something in the MTA database.  You can test this by "wiping" the MTA.  To do this follow these steps:

1.  Make sure the MTA resource is offline
2.  Move all db*.dat files from the mtadata folder
3.  Copy in the contents of the setup\i386\exchange\bootenv folder from the E2k3 CD into mtadata
4.  Remove the read only attribute from these files
5.  Bring the MTA resource online

If this doesn't work, check the system log for a 7031 from the MTA.  This will at least tell us that its crashing.
nbishop1979Author Commented:
I just did this, and it does the same thing.  I have checked the system log and the last entry is

Event Type:      Information
Event Source:      Service Control Manager
Event Category:      None
Event ID:      7036
Date:            8/6/2004
Time:            8:01:27 AM
User:            N/A
Computer:      Computer
The Microsoft Exchange MTA Stacks service entered the running state.

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

It comes online long enough to see all the db's on the servers and then immediately drops them and goes to the "Not Available" status.  The service still shows online in Cluster Admin and the service is running.

Any other ideas?
nbishop1979Author Commented:
OK...Here is the scoop.  We just talked to Microsoft and we have found an undocumented bug with the MTA on a cluster.  There is a limit to the number of queues it can handle, and since it has a queue for each db in each SG, we have exceeded that limit.  They are finding the limit to be around 35 to 40 MTA queues depending on hardware.  We got 57 queues up before it dies.  

This might be a good posting in an FYI section of this site.  Not sure if there is one or not....
Closed, 500 points refunded.
Community Support Moderator
nbishop1979Author Commented:
Just thought i would let you know that Microsoft is bumping the limit up to 70 DB's in case you happen to run into this.  Here is the email they sent me yesterday indicating this.

Per our conversation earlier, below are the steps that should increase the ability of your Exchange 2003 cluster to work with up to 70 databases.  At this time, there is no further information on the possibility of expanding this beyond the 70 database limit.  I should note that I will be out of the office for the next few days, and you can contact my colleague, in my absence.  As we discussed, I will go ahead and archive your case at this point, but if you have questions or need to reopen while I am gone, feel free to contact
Here are the steps below:
Currently in a Microsoft Exchange Server 2003 clustered environment, the MTA service has a limit on the number of databases that are supported. In a  clustered environment, the MTA service supports a maximum of 50 databases.
To increase the number of databases that are supported by the MTA service to 70, you must modify the MTA registry entry for remote procedure call (RPC) threads.  To do this, follow these steps.
Warning Serious problems might occur if you modify the registry incorrectly by using Registry Editor or by using another method. These problems might require that you reinstall your operating system. Microsoft cannot guarantee that these problems can be solved. Modify the registry at your own risk.
Note You must follow these steps on the Exchange 2003 computer that is running the MTA service.

1. Click "Start", click "Run", type "regedit" (without the quotation
   marks), and then click "OK".
2. Locate and then click the following registry subkey:
 " HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\MSExchangeMTA\Parameters"
3. In the right pane, right-click "Max RPC Calls Outstanding", and then click "Modify".
4. In the "Value data" box, type "00000080" (without the quotation marks) (hexadecimal) or "128" (without the quotation marks) (decimal),
   and then click "OK".
Important:  DO *NOT* CHANGE THE GATEWAY IN THREADS registry value OR THE GATEWAY OUT THREADS registry value in this
   scenario. In a clustered environment, changing these registry values may cause the MTA process to become unstable.

Featured Post

Has Powershell sent you back into the Stone Age?

If managing Active Directory using Windows Powershell® is making you feel like you stepped back in time, you are not alone.  For nearly 20 years, AD admins around the world have used one tool for day-to-day AD management: Hyena. Discover why.

  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now