CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   OpenFOAM on HPC (https://www.cfd-online.com/Forums/openfoam-solving/246808-openfoam-hpc.html)

philiplu December 26, 2022 00:32

OpenFOAM on HPC
 
Hello,

I'm new in HPC:
after "module load OF", running "blockMesh -help" gives
"cmd not found"

does someone, who's experiences in HPC, know what's wrong and why?


thank you in advance

Philip

Swirl December 26, 2022 10:48

It's possible, that you have to activate the module before being able to use it.
After
module load...
try
"foamInit" or
"source $FOAM_INIT"
these were the commands necessary on the HPC I used to work with.

philiplu December 27, 2022 00:01

hello, Swirl,
thank you very much for the reply.

btw. I want to check with you:

if possible, to compile/install other 3rd party package under the HPC-pre-installed OF-environment? if yes, how?

thank you in advance

Swirl December 27, 2022 08:32

Do you mean software like custom solvers or models? You should be able to compile those normally, like you would on any other PC, given the admins haven't blocked it. You could potentially also use the <nameOfApplication>.so file that was created from compiling the custom code on another machine and save it to either the $FOAM_USER_APPBIN (for solvers) or $FOAM_USER_LIBBIN (other types of custom software) directories on the HPC filesystem. This however only works, if the machine the program was compiled on has a similar CPU architecture than the HPC. Not sure if this is what you meant however.

philiplu December 28, 2022 03:05

hello, Swirl,
many thanks for the reply and you're right, the custom solver based on OF.

Actually now I can load, thus use OF in hpc, so i get 2 further options:

- one alternative is after loading OF in hpc, compile the custom solver directly
- or first install OF (again) in my hpc-account, then compile the custom solver based on the newly installed OF

not sure yet, which one works out

anyway, thank you for the hints

philiplu December 30, 2022 04:37

btw. I just start using hpc, i.e. no experiences yet, so
could someone give some ref., info, advices regarding the node-, core-,and memory-allocation based on a simulation task?

thank you in advance

bestucan January 1, 2023 08:52

what's the output of command "module load OF"? Is it succeed?
have you ever list all the modules available?

philiplu January 2, 2023 02:44

hello, bestucan,
many thanx for the reply.
yes, the "load module" worked out.... when I get time you'll tell you more.

now I'm busy with studying the "node, core allocation" ..... it would be great if you or someone could share me some hints

thank you again and wish you a great 2023

kk415 January 3, 2023 07:01

Hello All,


I am trying to run a case on HPC. The setup sometimes run and sometime shows an error as



Code:

[1672747228.750479] [cn098:39424:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.750536] [cn098:39424:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747231.037862] [cn100:48502:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/48501/fd/25 flags=0x0) failed: No such file or directory
[1672747231.037952] [cn100:48502:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc00000064000bd75: Shared memory error
[1672747228.750603] [cn098:39423:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.750657] [cn098:39423:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747228.750690] [cn098:39422:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.750734] [cn098:39422:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747231.037965] [cn100:48500:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/48501/fd/25 flags=0x0) failed: No such file or directory
[1672747231.038008] [cn100:48500:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc00000064000bd75: Shared memory error
[1672747228.750916] [cn098:39425:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.750965] [cn098:39425:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747228.750961] [cn098:39418:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.751017] [cn098:39418:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747228.751253] [cn098:39419:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.751301] [cn098:39419:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747231.038670] [cn100:48509:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/48501/fd/25 flags=0x0) failed: No such file or directory
[1672747228.751358] [cn098:39421:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.751412] [cn098:39421:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747228.751364] [cn098:39420:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/39378/fd/25 flags=0x0) failed: No such file or directory
[1672747228.751417] [cn098:39420:0]          mm_ep.c:159  UCX  ERROR  mm ep failed to connect to remote FIFO id 0xc0000006400099d2: Shared memory error
[1672747231.038743] [cn100:48498:0]        mm_posix.c:207  UCX  ERROR  open(file_name=/proc/48501/fd/25 flags=0x0) failed: No such file or directory


I have tested both shared memory option as yes and no in decomposePar dict. How can I solve this issue? Pardon me if I am missing any detail to give as I don't know what else is playing a role here. Any more information is required please let me know.


Regards
krishna


All times are GMT -4. The time now is 16:22.