Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Use more than 1 MPI process in CI. #2

Open
matz-e opened this issue May 10, 2024 · 0 comments
Open

Use more than 1 MPI process in CI. #2

matz-e opened this issue May 10, 2024 · 0 comments

Comments

@matz-e
Copy link
Member

matz-e commented May 10, 2024

Following up with HighFive more than one rank in MPI should work. During testing, I got a bunch of oversubscription error messages for two and more ranks:

#9 27.57 There are not enough slots available in the system to satisfy the 4
#9 27.57 slots that were requested by the application:
#9 27.57 
#9 27.57   /build/src/parquet2hdf5
#9 27.57 
#9 27.57 Either request fewer slots for your application, or make more slots
#9 27.57 available for use.
#9 27.57 
#9 27.57 A "slot" is the Open MPI term for an allocatable unit where we can
#9 27.57 launch a process.  The number of slots available are defined by the
#9 27.57 environment in which Open MPI processes are run:
#9 27.57 
#9 27.57   1. Hostfile, via "slots=N" clauses (N defaults to number of
#9 27.57      processor cores if not provided)
#9 27.57   2. The --host command line parameter, via a ":N" suffix on the
#9 27.57      hostname (N defaults to 1 if not provided)
#9 27.57   3. Resource manager (e.g., SLURM, PBS/Torque, LSF, etc.)
#9 27.57   4. If none of a hostfile, the --host command line parameter, or an
#9 27.57      RM is present, Open MPI defaults to the number of processor cores
#9 27.57 
#9 27.57 In all the above cases, if you want Open MPI to default to the number
#9 27.57 of hardware threads instead of the number of processor cores, use the
#9 27.57 --use-hwthread-cpus option.
#9 27.57 
#9 27.57 Alternatively, you can use the --oversubscribe option to ignore the
#9 27.57 number of available slots when deciding the number of processes to
#9 27.57 launch.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant