Christian Meesters

Hi, I, too, am interested in evaluating and deploying kanki. As I am working as a CompScientist for a HPC cluster, I am mainly interested in the Linux-Client (both for...

Hi, I hoped to at least spark a bit of a discussion, as the mentioned issues still persists for multiuser systems. Whether the work with `pgzip` is appreciated is admittedly...

same for me - just the CI ...

@johanneskoester you deleted the slurm setup from the github-workflow? is this a mistake, perhaps?

> It seems as if the job never really runs but is still reported as completed. This is in line with our dummy srun from the step before, which also...

Actually, the debug information and the error should be printed - I do not see any of it in the logs, ideas?

> Do we need to modify our srun statement for the calc rule? Yes, I noticed that - I will modify the srun test temporarily to figure out, which mpi...

setting tasks ought to be redundant: this information should be propagated from the slurmd to the slurmstepd (read from sbatch to srun)

OK, we can now safely conclude: The CI-slurm is not build with MPI-support. But why is there absolutely no output? Trying one thing ...

``` tests/test_slurm.py::test_slurm_mpi PASSED [ 33%] ``` Oh yeah! ``` tests/test_slurm.py::test_slurm_group_job FAILED [ 66%] ``` Oh no! Wait: ``` > assert os.path.exists(original_snakefile) E AssertionError ``` what does that mean? Why is...