Sign in
Microsoft.com
 
Microsoft
 
 
 

none
Bug in the job report ? RRS feed

 > 

  • Question

  • Question
    Sign in to vote
    0
    Sign in to vote
    Hi,
    during my tests on a cluster with 8 compute nodes, I found a mismach in the nodes allocation.

    Precisely, I ran 3 jobs simultaneously with 3 tasks each one:
    (requested resources: Auto-Auto Nodes)

    task 1.1 takes Node1 ,  task 1.2 takes Node2 ,  task 1.3 takes Node3
    task 2.1 takes Node4 ,  task 2.2 takes Node5 ,  task 2.3 takes Node6

    for the job3, the View Job dialog says that it takes the nodes 4,7,8
    but the Task Properties of the job says that it takes the nodes 7,8,8.

    (I don't know if the task has run in the Node4 or in the Node7)

    If you need more informations or screenshots, I can provide them privately.

    Bring me up to date :)

    Thanks,
    Marco


    Sunday, November 9, 2008 7:43 PM
    |

Answers

  • Question
    Sign in to vote
    0
    Sign in to vote
    Marco,
    I think this is what is happening:
    1. Job is allocated 3 nodes (4, 7, and 8)
    2. Task 1 is started on Node 7
    3. Task 2 is started on Node 8
    4. Task 2 finishes very quickly, since all they do is a dir
    5. Task 3 is started on Node 8 since it is already availale

    I think if you used longer running tasks, you'd see more of the expected behavior.

    -Josh
    Tuesday, December 2, 2008 2:52 AM
    |
    Moderator

All replies

  • Question
    Sign in to vote
    0
    Sign in to vote

    Hello Marco,
    It would be helpful to see the data you are talking about.  You can e-mail me screenshots if you'd like (jbarnard at microsoft dot com).  Easier might be to just run a job view and task view (possibly the with /details flag) from the command line and post the output of that up here.

    Thanks,
    Josh

    -Josh
    Friday, November 14, 2008 1:15 AM
    |
    Moderator
  • Question
    Sign in to vote
    0
    Sign in to vote
    Hello Josh,
    there are any news about this issue?

    Marco
    Friday, November 28, 2008 11:21 PM
    |
  • Question
    Sign in to vote
    0
    Sign in to vote
    Marco,
    I think this is what is happening:
    1. Job is allocated 3 nodes (4, 7, and 8)
    2. Task 1 is started on Node 7
    3. Task 2 is started on Node 8
    4. Task 2 finishes very quickly, since all they do is a dir
    5. Task 3 is started on Node 8 since it is already availale

    I think if you used longer running tasks, you'd see more of the expected behavior.

    -Josh
    Tuesday, December 2, 2008 2:52 AM
    |
    Moderator
  • Question
    Sign in to vote
    0
    Sign in to vote
    P.S.  Sorry for the delayed response . . . your mail was flagged as Spam for me :(
    -Josh
    Tuesday, December 2, 2008 2:53 AM
    |
    Moderator