Sign in
Microsoft.com
 
Microsoft
 
 
 

none
Session.CreateSession fails, even through everything seems setup fine... RRS feed

 > 

  • Question

  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi, 

    My first attempt to use SOA with HPC2012 R2 is not off to a great start... 8-( 

    The test code is very simple: 

    using System;
    using System.Collections.Generic;
    using System.Linq;
    using System.Text;
    using Microsoft.Hpc.Scheduler.Session ; 
    using Microsoft.Hpc.Scheduler.Properties ;
    using System.ServiceModel; 

    namespace ConsoleTest
    {
        class Program
        {
            static void Main(string[] args)
            {
                const string scheduler = "headnode hostname";
                const string service = "EchoService";

                SessionStartInfo info = new SessionStartInfo(scheduler, service);

                Session session = Session.CreateSession(info);
                Console.WriteLine("Connected to {0}", session.EndpointReference.ToString()); 
            }
        }
    }

    Compiles fine, throws this exception in CreateSession: 


    An unhandled exception of type 'Microsoft.Hpc.Scheduler.Session.SessionException' occurred in Microsoft.Hpc.Scheduler.Session.dll

    Additional information: Cannot find a broker node in the cluster that meets the session's requirements. Please check the documentation on how to configure a broker node. If you are using failover broker nodes and creating a DurableSession, please be sure one or more failover broker nodes are online

    The head node is up, it's configured as a wcf broker node, it's running in that role, I can connect to it and submit regular non-SOA jobs, etc, etc. What are those "session's requirements" which are not being met? How can I find out more about what it's complaining about? 

    Thanks

    Damian

    Friday, March 27, 2015 5:38 PM
    |

Answers

  • Question
    Sign in to vote
    1
    Sign in to vote

    Hi, 

    This is no longer an issue, after digging through some exceptions I found a clue to the problem... 

    The scheduler was replying to the client by saying "here, connect to THIS wcf broker node". Unfortunately the THIS was not a fully qualified domain name, just the machine name.. Since the client was in a different domain, dns lookup failed... 

    Would have been nice if the default exception message said that, haha... The above exception msg was useless... 

    As soon as I added that machine to my hosts file, it started to work fine... 

    Thanks

    Damian

    • Proposed as answer by Yongjun Tian (MSFT) Wednesday, April 8, 2015 1:28 AM
    • Marked as answer by DamianRK Wednesday, April 8, 2015 1:02 PM
    Tuesday, April 7, 2015 2:35 PM
    |

All replies

  • Question
    Sign in to vote
    0
    Sign in to vote

    Ohh, I wrapped that in a try/catch, and the exception message was basically the same thing as above, so didn't provide any more info/data... 

    using System;
    using System.Collections.Generic;
    using System.Linq;
    using System.Text;
    using Microsoft.Hpc.Scheduler.Session ; 
    using Microsoft.Hpc.Scheduler.Properties ;
    using System.ServiceModel; 

    namespace ConsoleTest
    {
        class Program
        {
            static void Main(string[] args)
            {
                const string scheduler = "host name";
                const string service = "EchoService";

                SessionStartInfo info = new SessionStartInfo(scheduler, service);

                try
                {
                    Session session = Session.CreateSession(info);
                    Console.WriteLine("Connected to {0}", session.EndpointReference.ToString()); 
                }
                catch (Exception e)
                {
                    string msg = e.Message;
                }
            }
        }
    }


    • Edited by DamianRK Friday, March 27, 2015 5:53 PM
    Friday, March 27, 2015 5:52 PM
    |
  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi, DamianRK,

    You have update "scheduler" to your real hpc cluster name, right?

    const string scheduler = "host name";

    And can you check the broker node (head node) state, and whether it is already online

    Tuesday, April 7, 2015 1:39 AM
    |
  • Question
    Sign in to vote
    1
    Sign in to vote

    Hi, 

    This is no longer an issue, after digging through some exceptions I found a clue to the problem... 

    The scheduler was replying to the client by saying "here, connect to THIS wcf broker node". Unfortunately the THIS was not a fully qualified domain name, just the machine name.. Since the client was in a different domain, dns lookup failed... 

    Would have been nice if the default exception message said that, haha... The above exception msg was useless... 

    As soon as I added that machine to my hosts file, it started to work fine... 

    Thanks

    Damian

    • Proposed as answer by Yongjun Tian (MSFT) Wednesday, April 8, 2015 1:28 AM
    • Marked as answer by DamianRK Wednesday, April 8, 2015 1:02 PM
    Tuesday, April 7, 2015 2:35 PM
    |
  • Question
    Sign in to vote
    0
    Sign in to vote

    Changed the WCF service to just run as local system account instead of specifying it in the service panel.

    Sunday, November 10, 2019 6:12 PM
    |