Odeslat dotazOdeslat dotaz
 

Odpovědětmpiexec -tracefile option does not work on cluster?

  • 30. června 2009 19:22tosa.yasunari Uživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaile
     
    I thought that it is a good idea to try "mpiexec -tracefile mytrace.etl MPIApp.exe".    I get the error:
    Aborting: failed to start tracing on CNODE00
    Error(183) Cannot create a file when that file already exists.

    I have no problem running "mpiexec MPIApp.exe".  

    Can you tell me how I can get the trace file? 

    Thank you.

Odpovědi

Všechny reakce

  • 3. července 2009 7:12Johannes_de Uživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaile
     
    Hello,

    there might be an issue with the access permissions.
    What working directory did you set? Have you write /create permissions there?
    Otherwise the trace file might be written to one default location for all users and you might not have sufficient rights to overwrite the trace files of other users.

    Johannes
    JH
  • 6. července 2009 17:01LioMSFTUživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaileUživatelské medaile
     Odpovědět
    Hi,

    One posability is that the tracing session is still on from a failed (Ctrl-C) run or a cancelled job; hence the new run can not create that session again.

    Run "logman stop msmpi -ets" and then retry tracing.

    .Erez