Installation problem for MFiX 24.1.1 and 24.2 on OSX

Hello,

I have been using MFiX 24.1.1 for close to 45 days, since 24.2 has been released, I have uninstalled it and tried to install 24.2, but was getting an error. So I have shifted back to 24.1.1, but now even that isn’t opening after installation.
Please find the attachment of the image of the terminal.

Also this is the message Python prints out.
Python 3.11 quit unexpectedly.pdf (385.1 KB)

Please let me know if there’s any other information that I can provide which can help resolve this problem.
Thank you so much.

An update: I was able to solve the issue of installation; I have installed mfix-24.2.
However, I am getting an error that I haven’t encountered before. Please let me know if that’s something related to the installation process.

[H-MacBook-Pro.local:07170] shmem: mmap: an error occurred while determining whether or not /var/folders/rj/b7mxhr792v57bq380b_kjd6w0000gp/T//ompi.H-MacBook-Pro.502/jf.0/3431268352/sm_segment.H-MacBook-Pro.502.cc850000.0 could be created.

While the above error occurred, the simulation is still running, and has not stopped.
Looking forward to hearing from you soon.
Thank you.

Hi @Sumanth_Penmetcha
We don’t have a lot of users on the Mac platform but it should work.
Please help me out by giving some more info

  1. What version of OSX are you running?
  2. Does this message appear when installing MFiX or when running it? Does it happen when you start up the GUI or when you start the simulation?
  3. Is the message printed more than once?
  4. Does it affect the running of either MFiX or the simulation?

Thanks,

– Charles

I have tried mfix-24.2 on MacOS “Ventura” (13.4.1) and I’m seeing the message you posted in your first post - “One of the two will be used. Which one is undefined” followed by a segmentation fault. You said you solved this, what did you do? In the conda package list (conda list) I see

qt-main                   5.15.8              hcd44e0d_22    conda-forge
qt6-main                  6.7.2                hbe5e5a2_3    conda-forge

This is not right, MFiX uses qt5 and that qt6 package should not be in there. I don’t know where this is coming from, this only happens on OSX. We may need to rebuild the OSX packages for 24.2.

In addition to the questions I asked above, please also include the output of conda info and conda list (with the mfix-24.2 environment activated)

Thank you!

OK, one more update. I found that I could get 24.2 to run on OSX by doing

rm $CONDA_PREFIX/lib/libQt6*

Is that what you did? These Qt6 files should not have been installed in the first place, but until we resolve that issue, deleting them seems to clear up the first of your problems…

I have not been able to reproduce the second problem, the shmem: mmap: warning. Do you have a reliable way to reproduce this? I suspect it is a harmless warning but I’d like to investigate further.

Hello,

Thanks for your responses.

To answer few of your questions:
1)I’m using Sonoma 14.4.1
2)The message of shmem: mmap: occurs after I start the simulation to run for mfix-24.2; and the one in the image I have attached of terminal was when I am trying to open mfix-24.1.1(but later I have uninstalled 24.1.1 and installed 24.2)
3)Once I am in 24.2, as I mentioned, the shmem: mmap: problem arises and also the attached image-


also occurs as an error, wherein the same file runs successfully in an Ubuntu OS.

To answer your other questions, yes I have removed qt6-main 6.7.2 for MFiX 24.2 to work and I have used the same code you provided, which is, rm $CONDA_PREFIX/lib/libQt6*

So I hope my answers help in improving and understanding of the problem, I’ll keep you posted if I encounter any other difficulties or errors from my end.
Thank you for your patience and time.

After dealing with the Qt6 issues (which will be fixed in the next release) I am able to start MFiX on my “Ventura” (13.4.1) Mac without errors. I ran the mixer_3d and cyclone_3d tutorials.

I’ll try updating to “Sonoma” and see if that makes any difference.

Can you please send me your case files so I can try with your case?

Hello,

Apologies for the late response, the problem was solved in the Ubuntu version, and I am currently working on that OS. However, here are the files that I have that I used for the above case
run16_x_1_p_0_desen_62_desenwall_47.mfx (11.0 KB)
particle_input.dat (80.7 KB)

Also I have a similar problem with the same files, that I have posted here Do not know the reason for high run-time of the MFiX-SQP case
please let me know if you could help, any suggestion or advise would be greatly appreciated.
Looking forward to hearing from you soon.

Sincerely,
Sumanth Penmetcha