mpi/parallelization problem: solver does not give any feedback in parallel

Bruno_Kwothe · September 11, 2019, 10:04

Hey there, people!

I am not completely sure my problem falls under this thread, but I do have a problem when using mpirun.

Until about a week ago, I could use mpirun on my laptop with no problems, and had done dozens of analysis in the past month. But now something happened (and to be honest I do not know yet if it is a problem with my Ubuntu or with OpenFOAM): when I give the command mpirun -np 8 interFoam -parallel > log.interFoam &, it simply does not compute. The log.interFoam gets created, but remains eternally empty.

The same analyses run completely fine in serial.

I've been having other problems with Ubuntu: it now refuses to shut down or restart when told to do so, and crashes in the process. I also cannot access system settings, because Ubuntu crashes when I try to do it. I am still not sure these problems are related to the one I have with parallel analyses, but I did notice them appearing on my laptop around the same time. I tried reinstalling Ubuntu, but the problems didn't seem to disappear.

I really hope someone can shed some light on this matter for me! Thanks in advance!

Kind regards,
Bruno Almeida

[Moderator note: Moved from mpirun problems: exited on signal 11 (segmentation fault) ]

wyldckat · October 9, 2019, 19:00

Quick'ish answer: I don't know if you have solved this issue yet or not, but there are two main possibilities:

Either Ubuntu has some new software update that is incompatible with your laptop.
Or your laptop is suffering from extended overheating sessions, resulting in hardware getting damaged.

To check if it's the second alternative, follow these instructions to check for issues: https://openfoamwiki.net/index.php/H...and_CPU_are_OK

If stressapptest tells you that there are issues in communicating between RAM and CPU, then your laptop is damaged. It could also just be a problem with due to overheating, so if there is dust clogging the ventilation of the laptop, then it's only going to get worse.

You should also check if the CPU is overheating by installing "lm-sensors":

Code:

sudo apt install lm-sensors

After it's installed, run:

Code:

sensors

and it will tell you about the current temperature and fan status, if there is any available.

If CPU temperature goes consistently above 70-80ºC, then that is really not good for the CPU's health and life span.

Beyond all of this, perhaps the machine is unable to shutdown properly if you recently launched a simulation with 8 subdomains... have you checked the System Monitor (sometimes also known as Task Manager) to see how much CPU and RAM is being used and by which applications?

Bruno_Kwothe · October 10, 2019, 13:18

Thank you for all your tips!

I still have the problem, but I am using a cluster to run my analyses, so I am not with my hands fully tied. I can still prepare my cases in my laptop without using parallelization, after all.

I tried using stressapptest, but I guess it did not find anything unusual in my hardware (or maybe I am misinterpreting its results). I got messages of :

"Stats: Found 0 hardware incidents
Stats: Completed: 444878.00M in 20.03s 22207.93MB/s, with 0 hardware incidents, 0 errors
Stats: Memory Copy: 444878.00M at 22240.49MB/s
Stats: File Copy: 0.00M at 0.00MB/s
Stats: Net Copy: 0.00M at 0.00MB/s
Stats: Data Check: 0.00M at 0.00MB/s
Stats: Invert Data: 0.00M at 0.00MB/s
Stats: Disk: 0.00M at 0.00MB/s

Status: PASS - please verify no corrected errors"

One thing I find quite strange is that when I change to my Windows (I am using dual-boot), it behaves completely as expected. I can turn-off or restart my laptop with no major incidents. And I believe another software of mine (Solidworks Flow Simulation) is able to parallelize its analyses.

I do believe, however, that the problems in my Ubuntu tend to get worse over time. It becomes more unstable. For example, I can no longer access my System Settings without the laptop crashing and recently even my Wi-fi connection was permanently lost. Things go back to normal once I reinstall Ubuntu, but that only lasts while I keep my laptop turned on. As soon as I turn it off, it stops accepting parallelization and accessing System Settings, and throughout the weeks it will keep getting more and more limited (until I get tired of the situation and decide to reinstall Ubuntu again).

I will see with a computer technician if my laptop's ventilation is clogged with dust or malfunctioning for some other reason.

Once again, thank you for your tips and your patience!

September 11, 2019, 10:04	mpi/parallelization problem: solver does not give any feedback in parallel	#1
Bruno_Kwothe New Member Bruno de Almeida Regina Join Date: Jul 2019 Location: Brazil Posts: 6 Rep Power: 6	Hey there, people! I am not completely sure my problem falls under this thread, but I do have a problem when using mpirun. Until about a week ago, I could use mpirun on my laptop with no problems, and had done dozens of analysis in the past month. But now something happened (and to be honest I do not know yet if it is a problem with my Ubuntu or with OpenFOAM): when I give the command mpirun -np 8 interFoam -parallel > log.interFoam &, it simply does not compute. The log.interFoam gets created, but remains eternally empty. The same analyses run completely fine in serial. I've been having other problems with Ubuntu: it now refuses to shut down or restart when told to do so, and crashes in the process. I also cannot access system settings, because Ubuntu crashes when I try to do it. I am still not sure these problems are related to the one I have with parallel analyses, but I did notice them appearing on my laptop around the same time. I tried reinstalling Ubuntu, but the problems didn't seem to disappear. I really hope someone can shed some light on this matter for me! Thanks in advance! Kind regards, Bruno Almeida [Moderator note: Moved from mpirun problems: exited on signal 11 (segmentation fault) ] Last edited by wyldckat; October 9, 2019 at 18:50. Reason: see "Moderator note:"

October 9, 2019, 19:00		#2
wyldckat Retired Super Moderator Bruno Santos Join Date: Mar 2009 Location: Lisbon, Portugal Posts: 10,975 Blog Entries: 45 Rep Power: 128	Quick'ish answer: I don't know if you have solved this issue yet or not, but there are two main possibilities: Either Ubuntu has some new software update that is incompatible with your laptop. Or your laptop is suffering from extended overheating sessions, resulting in hardware getting damaged. To check if it's the second alternative, follow these instructions to check for issues: https://openfoamwiki.net/index.php/H...and_CPU_are_OK If stressapptest tells you that there are issues in communicating between RAM and CPU, then your laptop is damaged. It could also just be a problem with due to overheating, so if there is dust clogging the ventilation of the laptop, then it's only going to get worse. You should also check if the CPU is overheating by installing "lm-sensors": Code: sudo apt install lm-sensors After it's installed, run: Code: sensors and it will tell you about the current temperature and fan status, if there is any available. If CPU temperature goes consistently above 70-80ºC, then that is really not good for the CPU's health and life span. Beyond all of this, perhaps the machine is unable to shutdown properly if you recently launched a simulation with 8 subdomains... have you checked the System Monitor (sometimes also known as Task Manager) to see how much CPU and RAM is being used and by which applications? __________________ OpenFOAM: FAQ \| Getting started Forum: How to get help, to post code/output and forum guide Read this before sending me PM

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
different results between serial solver and parallel solver	wlt_1985	FLUENT	11	October 12, 2018 08:23
Problem with running customized solver parallel on cluster	shinri1217	OpenFOAM Running, Solving & CFD	0	June 27, 2018 13:26
multiphase solver - parallel processing - GAMG	thibault_pringuey	OpenFOAM Programming & Development	2	August 27, 2013 22:03
PROBLEM IN SOLVER RUNNING PARALLEL	ÃÃ±igo	CFX	3	February 4, 2007 05:08
MPP Parallel solver problem with 3.2!	tomosada	Siemens	4	August 12, 2004 13:49

October 10, 2019, 13:18		#3
Bruno_Kwothe New Member Bruno de Almeida Regina Join Date: Jul 2019 Location: Brazil Posts: 6 Rep Power: 6	Thank you for all your tips! I still have the problem, but I am using a cluster to run my analyses, so I am not with my hands fully tied. I can still prepare my cases in my laptop without using parallelization, after all. I tried using stressapptest, but I guess it did not find anything unusual in my hardware (or maybe I am misinterpreting its results). I got messages of : "Stats: Found 0 hardware incidents Stats: Completed: 444878.00M in 20.03s 22207.93MB/s, with 0 hardware incidents, 0 errors Stats: Memory Copy: 444878.00M at 22240.49MB/s Stats: File Copy: 0.00M at 0.00MB/s Stats: Net Copy: 0.00M at 0.00MB/s Stats: Data Check: 0.00M at 0.00MB/s Stats: Invert Data: 0.00M at 0.00MB/s Stats: Disk: 0.00M at 0.00MB/s Status: PASS - please verify no corrected errors" One thing I find quite strange is that when I change to my Windows (I am using dual-boot), it behaves completely as expected. I can turn-off or restart my laptop with no major incidents. And I believe another software of mine (Solidworks Flow Simulation) is able to parallelize its analyses. I do believe, however, that the problems in my Ubuntu tend to get worse over time. It becomes more unstable. For example, I can no longer access my System Settings without the laptop crashing and recently even my Wi-fi connection was permanently lost. Things go back to normal once I reinstall Ubuntu, but that only lasts while I keep my laptop turned on. As soon as I turn it off, it stops accepting parallelization and accessing System Settings, and throughout the weeks it will keep getting more and more limited (until I get tired of the situation and decide to reinstall Ubuntu again). I will see with a computer technician if my laptop's ventilation is clogged with dust or malfunctioning for some other reason. Once again, thank you for your tips and your patience!