Running pyMAPDL on Jupyter

[ay8tb]

I'm trying to make use of my institution's HPC system so I can work on pyANSYS on my Macbook as well. I was able to create a container to install pyMAPDL, but I'm having trouble launching it on the Jupyter notebook since it needs the location of the ANSYS222.exe file. All I have from the institutions is the license file and server in the following format:

export ANSYSLMD_LICENSE_FILE=XXXXXX
export ANSYSLI_SERVERS=XXXX

How would I go about instructing launch_mapdl() that this is my license file and server so I can launch the software?

[germa89]

Hi @ay8tb. You need to create a container for MAPDL. Not PyMAPDL. Check: https://mapdl.docs.pyansys.com/version/stable/getting_started/make_container.html
Although the instructions might be a bit outdated. Fill free to ping me here in this question, if you have any issue.

Once you have a container with MAPDL installed, and with entrypoint pointing to Ansys executable, we can go over the next steps.

Additional reference: https://mapdl.docs.pyansys.com/version/stable/getting_started/macos.html

[ay8tb]

Hi @germa89, thanks for getting back to me. I'm trying to avoid installing Ubuntu on a separate machine since I don't have a separate machine available, which is why I'm trying to use the HPC system that I can access virtually. Would that still work?

[germa89]

Mmhh... Maybe you can do "docker-in-docker", where you launch a docker container with ubuntu, install docker in it, and use it for create another container with MAPDL. I have successfully attempted this, but with a Ubuntu OS as host, not with MacOS.

Let me know if it works.

[ay8tb]

Hi @germa89,
Just to follow up, I've managed to install Docker on my Mac and have downloaded ANSYS2022_R2 for Linux from the software gateway at the university. How would I install ANSYS on Docker and use that to launch mapdl?