Exchange virtual server limitations on Exchange 2000 clusters and Exchange 2003 clusters that have more than two nodes

View products that this article applies to.

This article was previously published under Q329208

Expand all | Collapse all

SUMMARY

Microsoft Exchange 2000 Server and Microsoft Exchange Server 2003 enforce Active/Passive scenarios on Exchange clusters that have more than two nodes. This means that on a four-node Exchange 2000 cluster or a four-node Exchange 2003 cluster, only three Exchange virtual servers can be created, and only one Exchange virtual server can be hosted per node.

MORE INFORMATION

For Datacenter clusters, you must update Exchange 2000 to Exchange 2000 Service Pack 1 (SP1) or later before you create the System Attendant resources (this does not apply to Exchange 2003). If an earlier version of Exchange 2000 is installed on a Datacenter cluster, and you try to create System Attendant resources, you may receive the following error message:

An error occurred attempting to add pages from the '{30C5DA98-BA0C-11D2-BC65-00C04FA32177}' extension DLL. Catastrophic failure. Error ID: -2147418113 (8000ffff).

You may see the following error in the application event log:

Event Type: Error
Event Source: MSExchangeCluster
Event ID: 2076
Description:<SAResourceName>: Cluster Service failed resource control function 
 
For more information, click http://search.support.microsoft.com/search/?adv=1.

You may also find the following entries that correspond to the failure in the cluster log:

Microsoft Exchange System Attendant <SA>:   : Cluster Service failed resource control function
For more information, click http://search.support.microsoft.com/search/?adv=1. Error Code: 5007.
[RM] s_RmResourceControl: Resource <SA> control operation 0x01500006 gives status=5007...
[FM] FmpRmResourceControl: RmResourceControl returned 5007 for resource 76a4722c-c78e-4661-aa03-c2c006c6d9bd, resid=1164336...

Note

Error 5007 = ERROR_RESOURCE_NOT_FOUND

On Exchange 2000 clusters and Exchange 2003 clusters that have more than two nodes, only one Exchange virtual server at a time can come online on each node. In a configuration that has a single passive node (N+1), this means that if two nodes fail, only one Exchange Virtual Server can come online, and the second Exchange Virtual Server cannot come online.

During the failover, you may receive the following error message (where <SA> is the System Attendant resource name:

An error occurred attempting to bring the '<SA>' resource online: The cluster resource could not be brought online by the resource monitor. Error ID: 5018 (0000139a).

The following events are logged in the system event log:

Event ID: 1065
Event Type: Error
Event Source: ClusSvc
Description:
Cluster resource SA failed to come online. 

Event ID: 1069
Event Type: Error
Event Source: ClusSvc
Description:
Cluster resource 'SA' failed.

The following messages appear in the Exchange 2000 Cluster log:

Microsoft Exchange System Attendant <SA>: [EXRES] Online requested.
Microsoft Exchange System Attendant <SA>: [EXRES]Online: Cannot start System Attendant resource because there is already another one running on this node.
[RM] OnlineResource failed, resource SA, status =  5071.
[CP] CppResourceNotify for resource SA
[CP] CppRundownCheckpoints removing RNB for System\CurrentControlSet\Services\MSExchangeSA\Parameters\DELLEVS1
[FM] FmpRmOnlineResource: InterlockedDecrement on gdwQuoBlockingResources for resource d213bc2a-790e-4e8b-9c6f-f546359874e6
[FM] FmpRmOnlineResource: RmOnlineResource Failed. Resource d213bc2a-790e-4e8b-9c6f-f546359874e6, status 5071
[FM] FmpRmOnlineResource: Failed. Resource d213bc2a-790e-4e8b-9c6f-f546359874e6, state 4
[FM] FmpRmOnlineResource: Returning. Resource d213bc2a-790e-4e8b-9c6f-f546359874e6, state 4, status 5018.

The following messages appear in the Exchange 2003 Cluster log:

Microsoft Exchange System Attendant <EVS2 SA>: [EXRES] ExchangeOnline: Online requested.
Microsoft Exchange System Attendant <EVS2 SA>: [EXRES] EventLogging: EVS2 SA: The Exchange cluster is operating in an Active/Passive
   mode and cannot have more than one System Attendant resource online on the same node. Error Code: 0.	
Microsoft Exchange System Attendant <EVS2 SA>: [EXRES] EventLogging: EVS2 SA: Failed to bring the resource online. Error Code: 5071.
   [RM] OnlineResource failed, resource EVS2 SA, status = 5071.

Note

Error 5071 = ERROR_NODE_CANT_HOST_RESOURCE
Error 5018 = ERROR_RESMON_ONLINE_FAILED

By default, all the resources that are in the cluster are configured to permit "possible owner" of all nodes that are in the cluster, and no "preferred owner" values are set. If a failure is experienced, the Exchange 2000 and Exchange 2003 resources may try to come online on a node that is already running an Exchange virtual server. As described earlier in this article, the resources cannot come online.

To work around this issue, make sure that you correctly configure the failover policies on the cluster to target any failure to a passive node.

197047 (http://support.microsoft.com/kb/197047/ ) Failover/failback policies on Microsoft Cluster Server

Exchange 2000 clusters can host only a single MAPI public folder store across all Exchange virtual servers in the cluster. By default, only the first Exchange virtual server that is created contains a public folder store that is associated with the MAPI public folder hierarchy. When you try to create additional public folder stores in the MAPI hierarchy, you receive the following error:

All the public folder trees already have an associated public store on this server. You will need to create a new public folder tree before creating this new public store.

For more information, click the following article number to view the article in the Microsoft Knowledge Base:

255077 (http://support.microsoft.com/kb/255077/ ) How to create a new public folder tree and store in Exchange 2000 Server and in Exchange Server 2003