run h5py import test using mpirun in sanity check for h5py 2.10.0 easyconfigs using */2020a toolchain

[boegel]

(created using eb --new-pr)

[boegel]

Test report by @boegel
FAILED
Build succeeded for 1 out of 2 (2 easyconfigs in this PR)
node3144.skitty.os - Linux centos linux 7.7.1908, Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz, Python 3.6.8
See https://gist.github.com/d21a27540c0a3a0384f29cc1f80e13a3 for a full test report.

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 2 out of 2 (2 easyconfigs in this PR)
node2439.golett.os - Linux centos linux 7.7.1908, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz, Python 2.7.5
See https://gist.github.com/bd14b1ed6188aea61128ccbe9619d4c2 for a full test report.

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 2 out of 2 (2 easyconfigs in this PR)
generoso - Linux centos linux 7.6.1810, Intel(R) Xeon(R) CPU E5-2660 v4 @ 2.00GHz, Python 3.6.8
See https://gist.github.com/fdad50061622668a89dd185a7fa26998 for a full test report.

[lexming]

@boegel I got feedback from Intel on the issue related to this PR. They acknowledge it and it seems that it will be fixed in a future release. Therefore, I prefer very much to solve it with your fix than with mine in easybuilders/easybuild-easyblocks#2001, as it is something that likely won't be needed in the future.
I would just reduce even more the application of this fix to h5py-2.10.0-intel-2020a-Python-3.8.2.eb as OpenMPI is not affected by this issue.

[boegel]

@lexming This works fine for me now, as long as #10499 is used as well for UCX...

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 2 out of 2 (2 easyconfigs in this PR)
node3402.kirlia.os - Linux centos linux 7.7.1908, x86_64, Intel(R) Xeon(R) Gold 6240 CPU @ 2.60GHz (cascadelake), Python 2.7.5
See https://gist.github.com/1a2c1824888ac592872d4842e1e30d03 for a full test report.

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 2 out of 2 (2 easyconfigs in this PR)
generoso - Linux centos linux 7.6.1810, x86_64, Intel(R) Xeon(R) CPU E5-2660 v4 @ 2.00GHz, Python 3.6.8
See https://gist.github.com/4574fadf4e4a745c19117a24cdb0cea9 for a full test report.

[lexming]

Test report by @lexming
SUCCESS
Build succeeded for 2 out of 2 (2 easyconfigs in this PR)
node154.hydra.os - Linux centos linux 7.7.1908, x86_64, Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz, Python 2.7.5
See https://gist.github.com/4e787bab83953f9a3c0061091bdb8009 for a full test report.

[lexming]

LGTM

[lexming]

Going in, thanks @boegel !

[boegel]

Note to self: sometimes the import h5py test fails with Segmentation fault or Floating point exception when running in a Slurm job, which seems to be fixable by using export I_MPI_HYDRA_TOPOLIB=ipl first (see also #10314).
It doesn't always occur though, it seems to be related to the number of available cores/amount of available memory in the job?!