SharePoint 2010 multi farm high availability consideration

With SharePoint 2010 the idea of a “farm of farms” got introduced by this an SharePoint 2010 deployment is described which not only consists of a single farm but of multiple farms sharing special service applications among each other.

A very common scenario is to have a “resource- or service farm” hosting the shareable services and multiple “consumer- farms” consuming the services from it.

As you might know, not all SharePoint2010 service applications can be shared across farms.

These are the service applications shareable across farms:

• User Profile
• Managed Metadata
• Business Connectivity
• Search
• Secure Store
• Webanalytics

Please refer to this TechNet Article for more details.

If you are looking for details on how to configure it this article can give you a good indication on how to do it.

So far so well known… I encountered an interesting thing with one of my clients who is running a mulit-farm environment and has several services shared across multiple farms.

During maintenance work they had to stop the SharePoint services on their “service farm” (that be the farm providing services for other farms) and when trying to access any webapplications hosted on the other farms they encountered an error message with a corelation id.

After checking the ULS logs I found:

Exception occured while connecting to WCF endpoint: System.ServiceModel.EndpointNotFoundException: Could not connect to https://servername:port/serviceapplicationGUID/ProfilePropertyService.svc. TCP error code 10061: No connection could be made because the target machine actively refused it <IP Adress>.  —> System.Net.WebException: Unable to connect to the remote server —> System.Net.Sockets.SocketException: No connection could be made because the target machine actively refused it <IP Adress>.      at System.Net.Sockets.Socket.DoConnect(EndPoint endPointSnapshot, SocketAddress socketAddress)     at System.Net.ServicePoint.ConnectSocketInternal(Boolean connectFailure, Socket s4, Socket s6, Socket& socket, IPAddress& address, ConnectSocketState state, IAsyncResult asyncResult, Int32 timeout, Exception& exception)     — End of inner exception stack trace —     at System.Net.HttpWebRequest.GetRequestStream(TransportContext& context)     at System.Net.HttpWebRequest.GetRequestStream()     at System.ServiceModel.Channels.HttpOutput.WebRequestHttpOutput. GetOutputStream()     — End of inner exception stack trace —    Server stack trace:      at System.ServiceModel.Channels.HttpOutput.WebRequestHttpOutput. GetOutputStream()     at System.ServiceModel.Channels.HttpOutput.Send(TimeSpan timeout)     at System.ServiceModel.Channels.HttpChannelFactory.HttpRequestChannel. HttpChannelRequest.SendRequest(Message message, TimeSpan timeout)     at System.ServiceModel.Channels.RequestChannel.Request(Message message, TimeSpan timeout)     at System.ServiceModel.Channels.SecurityChannelFactory`1. SecurityRequestChannel.Request(Message message, TimeSpan timeout)     at System.ServiceModel.Dispatcher.RequestChannelBinder.Request(Message message, TimeSpan timeout)     at System.ServiceModel.Channels.ServiceChannel.Call(String action, Boolean oneway, ProxyOperationRuntime operation, Object[] ins, Object[] outs, TimeSpan timeout)     at System.ServiceModel.Channels.ServiceChannelProxy.InvokeService(IMethodCallMessage methodCall, ProxyOperationRuntime operation)     at System.ServiceModel.Channels.ServiceChannelProxy.Invoke(IMessage message)    Exception rethrown at [0]:      at System.Runtime.Remoting.Proxies.RealProxy.HandleReturnMessage(IMessage reqMsg, IMessage retMsg)     at System.Runtime.Remoting.Proxies.RealProxy.PrivateInvoke(MessageData& msgData, Int32 type)     at Microsoft.Office.Server.UserProfiles.IProfilePropertyService. GetProfileProperties()     at Microsoft.Office.Server.UserProfiles.ProfilePropertyServiceClient. <>c__DisplayClass1.<GetProfileProperties>b__0(IProfilePropertyService channel)     at Microsoft.Office.Server.UserProfiles.MossClientBase`1.ExecuteOnChannel(String operationName, CodeBlock codeBlock)

and

SharePoint Web Services Round Robin Service Load Balancer Event: EndpointFailure Process Name: w3wp Process ID: 6428 AppDomain Name: /LM/W3SVC/728495272/ROOT-1-129866768378254013 AppDomain ID: 2 Service Application Uri: urn:schemas-microsoft-com:sharepoint:service:82f426f16fb14278811f7cdce8c9912c#authority= urn:uuid:b217762e94f442b6b1c5d89a864f2a11&authority= https://servername:Port/Topology/topology.svc Active Endpoints: 1 Failed Endpoints:1 Affected Endpoint: https://servername:Port/serviceappGUID/ProfileService.svc  

Which tells me that when I try to access a site running on another farm (one which consumes the user profile service from the “services farm”) SharePoint tries to connect to the server this service runs on which is not accessible and for that gives you an error.

For me I draw the conclusion that it would be recommended to loadbalance the servers hosting your userprofile service application within your service farm to avoid that this server becomes the single point of failure.