topic Re:Cluster Environment: MPI_Comm_spawn() is hanged in Intel® MPI Library

Cluster Environment: MPI_Comm_spawn() is hanged

SarahVN — Tue, 19 Jul 2022 06:30:05 GMT

Hi ,

I would like to develop new application based on Intel MPI and run in cluster environment.
I launch my app as controller in host1 with command: -host host1 -n 1 [executable path]
In my app, I use MPI_Comm_spawn() method to create 2 workers in host2 with MPI_Info (MPI_Info include -host(host2), -wdir(WorkDir)).
When I run my app, it create controller on host1, but it create ONLY ONE worker on host2 and it is hanged at MPI_Comm_spawn() method.
1. Why is it hanged?
2. How should I config to run in Cluster?

Thank
Sarah.H

Re:Cluster Environment: MPI_Comm_spawn() is hanged

ShivaniK_Intel — Wed, 20 Jul 2022 11:12:30 GMT

Hi,

Thanks for reaching out to us.

Could you please let us know the MPI version, OS details, cluster details, and interconnect you have been using?

Could you also please provide us with the sample reproducer code with steps to reproduce the issue?

Thanks & Regards

Shivani

Re:Cluster Environment: MPI_Comm_spawn() is hanged

ShivaniK_Intel — Thu, 28 Jul 2022 05:56:05 GMT

Hi,

As we didn't hear back from you, Could you please provide the details that have been asked in my previous post so that we can investigate more on your issue?

Thanks & Regards

Shivani

Re:Cluster Environment: MPI_Comm_spawn() is hanged

ShivaniK_Intel — Fri, 05 Aug 2022 06:37:20 GMT

Hi,

I have not heard back from you. This thread will no longer be monitored by Intel. If you need further assistance please raise a new question.

Thanks & Regards

Shivani