Miniconda from Centos to Rocky, some issues

Hi,
We work on a HPC cluster and the sys admins have recently update the OS from CentOS to Linux Rocky. For our project we have a common miniconda install with about 50 environments. Before the OS upgrade I have created yml files from all of our environments so that we could reinstall them after the upgrade, when needed.

We have now found that most of our environments seem to work after the OS change, without reinstallation. No idea why that is. There are however some issues with our conda install. For instance the conda info command does not work properly, which prevents us now from creating new environments, or upgrading them.

Because of these issues I wanted to test if installing a new miniconda would work, and if I could install the environments from the yml files. I removed my links from the install we have in our project and I set-up an miniconda installation for myself. That works as it should.

I have then used some of the yml files to install a few environments. One of those envs did not install, and I got an conda error message related to this issue: https://github.com/conda/conda/issues/9356
I then did a conda clean -i and that removed the issue. However, the installation seems to stall at the solving environments step for the program nanoplot=1.40.2.

What would you advice on how to deal with this situation?

Can you include the error message below that you get when the conda environment fails to solve?

Also if you could include a list of packages you installed into this conda environment someone can attempt to recreate the problem.

1 Like

Okay, the only way to resolve this issue was to remove the old miniconda installation and set-up a new miniconda environment. That also meant I needed to reinstall all conda environments.

That worked for of the environments using the yml files I made before without a problem. With old environments it did not work properly, but we often had a newer version already present.