Test Agent cannot connect to test controller in network isolated environment

Question

I have a lab environment configured with 6 machines. It is running in Network Isolation, has workflow capability enabled and test agents installed on all machines except the domain controller.

I have configured all of them, so the servers are using a service and running as network service. The clients are running interactive as a user on the local domain. When I start the the environment I get the following error on 3 of the 5 machines:

--------------------------------------------------
Environment message: Type=Warning; Message=TF267042: One or more machines are not ready to run tests. For more information, see the individual machine errors.;

Machine messages:

Machine name: Web_2003std_x86_SP2_LabAgent_TestAgent_IIS

Machine message: Type=Error; Message=TF267055: The machine is not ready to run tests because of the following error: Unable to connect to the controller on '<TestControllerFQDN>:6901'. Reason: No such host is known. Make sure the test controller is online and the firewall on the test controller is not blocking the connection..;

Machine name: DB_2003std_x86_SP2_Ora11_C4.X

Machine message: Type=Error; Message=TF267055: The machine is not ready to run tests because of the following error: Unable to connect to the controller on '<TestControllerFQDN>:6901'. Reason: No such host is known. Make sure the test controller is online and the firewall on the test controller is not blocking the connection..;

Machine name: XP_x86_UK_SP3_IE8_OF2007_LabTest

Machine message: Type=Error; Message=TF267055: The machine is not ready to run tests because of the following error: Unable to connect to the controller on '<TestControllerFQDN>:6901'. Reason: No such host is known. Make sure the test controller is online and the firewall on the test controller is not blocking the connection..;

--------------------------------------------------

Neither of these machines have Windows Firewall running.

The machines that are working are respectively a Win2008 server and a Win7 client.

Please help. This is blocking further progress!

Answer 1

Hi Thomas,

Would it be possible to check the following

1) Test controller service is up and running on <TestControllerFQDN>. Should be since 2 machines are able to reach it.

2) <TestControllerFQDN> is reachable from these machines, try ping.

3) The test agent service account is provisioned as a part of the TeamTestAgentService group on the test controller machine

 

Thanks

-Allen

Answer 2

Hi Allen,

I've now looked things over and

1 ) yes, the Test Controller service is running on the Testcontroller.

2) Yes, the testcontroller is reachable by ping from the test agent machines

3) There are not registered any accounts in the TestTestAgentService group on the test controller. As I understand it, this is not necessary, since the environment is provisioned through Lab Management with network isolation. Anyways, the Win7 and Win2008 machines both work with the same users that don't work with the other machines.

Regards,

Thomas

Answer 3

Some more info. When I try to repair the testing capability on my virtual environment, the error below is shown in the application event log on the controller. Please note that the users I use for the test agent service are all part of the local administrator group on the machines.

- <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">

- <System>

  <Provider Name="VSTTController" />

  <EventID Qualifiers="0">0</EventID>

  <Level>2</Level>

  <Task>0</Task>

  <Keywords>0x80000000000000</Keywords>

  <TimeCreated SystemTime="2011-01-23T14:38:14.000000000Z" />

  <EventRecordID>5199</EventRecordID>

  <Channel>Application</Channel>

  <Computer><TestControllerFQDN></Computer>

  <Security />

  </System>

- <EventData>

  <Data>(QTController.exe, PID 2152, Thread 42) User not in the Admin group.</Data>

  </EventData>

  </Event>

Answer 4

Hi again Thomas,

Could you see if the Hyper-V integration components are available on these VMs? You can do so via VMM Admin Console. This would block our communication channel to the VMs and cause such an error as well.

 

Thanks

-Allen Mathias

Answer 5

Hi Allen,

It says "Virtual machine additions properties: Detected", which I assume means this is a yes. Anyways, both network isolation and workflow are both enabled and working, and I have managed to get the build-deploy part going on the machines where I have issues.

Is there anything else I can try? This issue is becoming a great pain for us...

Regards

Thomas

Answer 6

Hi Thomas,

This error "No such host is known" is observed when the agent machine is not able to resolve the test controller machine name. Please ensure that the agent machine is able to resolve both the controller's FQDN and controller's netbios name.

>> Machine message: Type=Error; Message=TF267055: The machine is not ready to run tests because of the following error: Unable to connect to the controller on '<TestControllerFQDN>:6901'. Reason: No such host is known. Make sure the test controller is online and the firewall on the test controller is not blocking the connection..;

Regards

Aseem Bansal

Answer 7

I have tried doing a ping of the TestControllerFQDN on the machines in question and they resolve the testcontroller perfectly, both on the name and the IP address. It really doesn't make any sense that it gives the "No such host is known" error. I might add that I also veified that firewall on the testcontroller and port 6901 is indeed open to all machines.

Could it have anything to do with the errors I posted in the event log: "User not in the admin group"? Should be said though that the users on the test agent machines are indeed in the administrator group, so it might be somewhere else?

Answer 8

Hi Thomas,

This error "No such host is known" is observed when the agent machine is not able to resolve the test controller machine name. Please ensure that the agent machine is able to resolve both the controller's FQDN and controller's netbios name.

>> Machine message: Type=Error; Message=TF267055: The machine is not ready to run tests because of the following error: Unable to connect to the controller on '<TestControllerFQDN>:6901'. Reason: No such host is known. Make sure the test controller is online and the firewall on the test controller is not blocking the connection..;

Regards

Aseem Bansal

Bringing this one up again. The Netbios name cannot be resolved from this particular machine. Why this is happening in a seemingly random way is what I don't get. How do I make the controller available through Netbios again?

Answer 9

I solved this by adding an entry in my hosts file that pointed to the correct IP Address.

even though a dos command window on the desktop resolved the IP Address, for whatever reason, the agent could not.

Answer 10

Hi Thomas,

I solve this issue by adding a record in the private domain DNS so it can resolve the Test Contoller machine name, for more information I create a video describe how to do that, you can find it here

http://mohamedradwan.wordpress.com/2012/03/04/using-notwork-isolation-in-lab-management-and-solve-some-issues/

Updated

Good links for Lab Management

Troubleshooting errors in lab management

Network Isolation Troubleshooting Guide

---

We are volunteers, if the reply help you mark it as your answer. thanks!!

Blog: http://mohamedradwan.wordpress.com