Intel® MPI Library
Get help with building, analyzing, optimizing, and scaling high-performance computing (HPC) applications.
2241 Discussions

MPI stops in a stochastic manner but at the same line in the code

Anders_S_1
New Contributor III
1,877 Views

Hi,

MPI stops at the same line in my ifort code but in a stochastic manner. Sometimes after short time, sometimes after a longer time. An error message is attached. How can I debug MPI?

Best regards

Anders_S

0 Kudos
7 Replies
RabiyaSK_Intel
Employee
1,840 Views

Hi,


Thanks for posting in Intel Communities.


Could you please provide the following details to reproduce your issue at our end:

1. Intel HPC toolkit and Intel MPI Library versions

2. Sample reproducer along with steps to reproduce

3. CPU, OS and hardware details


Thanks & Regards,

Shaik Rabiya


0 Kudos
Anders_S_1
New Contributor III
1,818 Views

Hi Shaik,

Thanks for replying!

I get this error on two desktops (AMD Ryzen 3990X 64-core and Intel Core i9-7980XE 18-core)

I have installed the latest versions of Base Tool kit and HPC Tool kit, I use Visual Studio for

developing the code and I use ifort under Windows. Project type is QuickWin.

As this happens in a rather large code, supplying a sample reproducer is not trivial.

Is it possible to use a MPI debugger? I supplied a part of the error message (I use 55 threads).

Is there any place where a more detailed error message is written from MPI?

Best regards

Anders_S

0 Kudos
RabiyaSK_Intel
Employee
1,806 Views

Hi,

 

>>>I get this error on two desktops (AMD Ryzen 3990X 64-core 

We regret to inform you that we can only offer direct support for Intel hardware platforms that the Intel® oneAPI product supports. Intel provides instructions on how to compile oneAPI code for both CPU and a wide range of GPU accelerators. Please go through the following link:

https://intel.github.io/llvm-docs/GetStartedGuide.html

 

>>> Intel Core i9-7980XE 18-core

But we will triage for this desktop.

 

>>>I have installed the latest versions of Base Tool kit and HPC Tool kit, I use Visual Studio

Could you please specify the version and sub-version of Visual Studio along with Intel oneAPI toolkits version as it might help the users with similar issue.

 

>>>As this happens in a rather large code, supplying a sample reproducer is not trivial.

For Community support we would require you to submit a minimum reproduction sample code specific to your issue that provides us the most relevant background information for triage. In case you require privacy, and unable to share the issue / sample with us publicly and If you are a licensed oneAPI product customer and/or member of Intel’s oneAPI Academic Program please submit a ticket for Priority support so that your application can be handled with the required data protection and privacy regulations

 

>>>Is there any place where a more detailed error message is written from MPI?

You can try Intel HPC Toolkit's Intel Analyzer and Collector(ITAC) for tracing your application and for a detailed error diagnosis. Please check the link provided for using ITAC:

https://www.intel.com/content/www/us/en/docs/trace-analyzer-collector/user-guide-reference/2023-1/overview.html

 

You can also try either exporting I_MPI_DEBUG variable with sufficient value with respect to your application before running the generated executable. Please go through this link for more information:

https://www.intel.com/content/www/us/en/docs/mpi-library/developer-guide-windows/2021-10/displaying-mpi-debug-information.html

 

Thanks & Regards,

Shaik Rabiya

0 Kudos
RabiyaSK_Intel
Employee
1,726 Views

Hi,


We haven't heard from you. Could you please respond to my previous post?


Thanks & Regards,

Shaik Rabiya


0 Kudos
Anders_S_1
New Contributor III
1,711 Views

Hi Shaik,

As I told you, I am waiting for a response from one of your colleagues on

a problem to install the latest versions of oneAPI on the i9 cpu.

Best regards

Anders_S

0 Kudos
RabiyaSK_Intel
Employee
1,655 Views

Hi,


>>> I am waiting for a response from one of your colleagues on a problem to install the latest versions of oneAPI on the i9 cpu.

Thanks for the update. Could you please let us know whether your issue regarding the installation is resolved? If not could we close this thread at our end? If your installation issue is resolved, you could raise a new query. We will definitely assist with the query.


Thanks & Regards,

Shaik Rabiya



0 Kudos
RabiyaSK_Intel
Employee
1,586 Views

Hi,


We haven't heard back from you. If you need any additional information, you can post a new question as this thread will no longer be monitored by Intel.


Thanks & Regards,

Shaik Rabiya


0 Kudos
Reply