Cluster Profile Manager: SPMD job test (createCommunicatingJob) FAILED

28 visualizzazioni (ultimi 30 giorni)
Hi,
I am trying to validate my local (default) cluster profile and the validation fails at the 3rd stage titled "SPMD job test (createCommunicatingJob)." From the validation results, it seems that MPI is not allowed to access a socket. I even turned off Windows Firewall to no avail.
I use a Windows 7 machine with MATLAB 2013a and Parallel Computing Toolbox Version 6.2. Any help on this issue is highly appreciated.
The contents of the validation results are given below:
-------------------------------------------------------------------------
VALIDATION DETAILS Profile: local Scheduler Type: Local
Stage: Cluster connection test (parcluster) Status: Passed Description:Validation Passed Command Line Output:(none) Error Report:(none) Debug Log:(none)
Stage: Job test (createJob) Status: Passed Description:Validation Passed Command Line Output:(none) Error Report:(none) Debug Log:(none)
Stage: SPMD job test (createCommunicatingJob) Status: Failed Description:The job errored or did not reach state finished. Command Line Output:(none) Error Report:(none) Debug Log: LOG FILE OUTPUT: [0]MATLAB is running in headless mode. Figure windows will not be displayed. [1]MATLAB is running in headless mode. Figure windows will not be displayed. [0] [0]To get started, type one of these: helpwin, helpdesk, or demo. [0]For product information, visit www.mathworks.com. [0] [1] [1]To get started, type one of these: helpwin, helpdesk, or demo. [1]For product information, visit www.mathworks.com. [1] [1]startup.m just ran! [0]startup.m just ran! [1]2013-06-22 22:13:25 | About to evaluate task with DistcompEvaluateFileTask [0]2013-06-22 22:13:25 | About to evaluate task with DistcompEvaluateFileTask [1]» [0]» 2013-06-22 22:13:25 | Enter distcomp_evaluate_filetask_core [1]2013-06-22 22:13:25 | Enter distcomp_evaluate_filetask_core/iSetup [1]2013-06-22 22:13:25 | This process will exit on any fault. 2013-06-22 22:13:25 | Enter distcomp_evaluate_filetask_core [0]2013-06-22 22:13:25 | Enter distcomp_evaluate_filetask_core/iSetup [0]2013-06-22 22:13:25 | This process will exit on any fault. [1]2013-06-22 22:13:25 | This process will exit when its parent process dies. [1]2013-06-22 22:13:25 | About to initialize MPI. [0]2013-06-22 22:13:25 | This process will exit when its parent process dies. [0]2013-06-22 22:13:25 | About to initialize MPI. [1][01:11468]..ERROR:Error while connecting to host, An attempt was made to access a socket in a way forbidden by its access permissions. (10013) [1][01:11468]..ERROR:Error while connecting to host, An attempt was made to access a socket in a way forbidden by its access permissions. (10013) [0][01:4248]..ERROR:Error while connecting to host, An attempt was made to access a socket in a way forbidden by its access permissions. (10013) [0][01:4248]..ERROR:Error while connecting to host, An attempt was made to access a socket in a way forbidden by its access permissions. (10013) [0][01:4248]..ERROR:Connect on sock (host=ECE-055960.fit.edu fe80::4c8d:bfcc:9295:5ede%29 fe80::b476:eb01:b33f:9af2%29 172.31.4.101 192.168.2.|Î, port=52860) failed, exhaused all end points [0]SMPDU_Sock_post_connect failed. [0][0] PMI_ConnectToHost failed: unable to post a connect to ECE-055960.fit.edu fe80::4c8d:bfcc:9295:5ede%29 fe80::b476:eb01:b33f:9af2%29 172.31.4.101 192.168.2.|Î:52860, error: Undefined dynamic error code [0]uPMI_ConnectToHost returning PMI_FAIL [0][0] PMI_Init failed. [0]Fatal error in MPI_Init_thread: Other MPI error, error stack: [0]MPIR_Init_thread(394): Initialization failed [0]MPID_Init(103).......: channel initialization failed [0]MPID_Init(374).......: PMI_Init returned -1 [1]Fatal error in MPI_Init_thread: Other MPI error, error stack: [1]MPIR_Init_thread(394): Initialization failed [1]MPID_Init(103).......: channel initialization failed [1]MPID_Init(374).......: PMI_Init returned -1 [1][01:11468]..ERROR:Connect on sock (host=ECE-055960.fit.edu fe80::4c8d:bfcc:9295:5ede%29 fe80::b476:eb01:b33f:9af2%29 172.31.4.101 192.168.2.uÎ, port=52853) failed, exhaused all end points [1]SMPDU_Sock_post_connect failed. [1][1] PMI_ConnectToHost failed: unable to post a connect to ECE-055960.fit.edu fe80::4c8d:bfcc:9295:5ede%29 fe80::b476:eb01:b33f:9af2%29 172.31.4.101 192.168.2.uÎ:52853, error: Undefined dynamic error code [1]uPMI_ConnectToHost returning PMI_FAIL [1][1] PMI_Init failed.
Stage: Pool job test (createCommunicatingJob) Status: Skipped Description:Validation skipped due to previous failure. Command Line Output:(none) Error Report:(none) Debug Log:(none)
Stage: MATLAB pool test (matlabpool) Status: Skipped Description:Validation skipped due to previous failure. Command Line Output:(none) Error Report:(none) Debug Log:(none)

Risposte (1)

Rob Dupre
Rob Dupre il 16 Nov 2017
Probably too late, but when My machine fell over at this point, I found there was an issue with my startup.m. Try removing it and trying again

Categorie

Scopri di più su Cluster Configuration in Help Center e File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by