James Corbett
James Corbett
Older versions of Flux work properly though. See below. ``` [corbett8@lassen27:~]$ lrun -n 2 flux start echo foobar foobar foobar [corbett8@lassen27:~]$ lrun -n 2 /usr/global/tools/flux/blueos_3_ppc64le_ib/flux-c0.29.0-s0.18.0/bin/flux start echo foobar foobar [corbett8@lassen27:~]$...
Dong and I met with @davidbloss and @dawson6 on Monday to discuss how best to support ATS. @dongahn wanted to have an issue to discuss next steps and issues that...
At the CORAL2 Quarterly, HPE described their “low noise mode” they have where they dedicate some CPUs to OS operations in order to improve application performance. It's similar to the...
A few users, including @davidbloss, are interested in using the python bindings for Flux from their own Python installations, rather than the Python that Flux was built with. From a...
Rabbit Load Balancing - Ensure that allocations are spread as broadly across the rabbits as possible (no rabbit "hotspots") - Ensure that MDTs from different jobs are distributed as much...
Fluxion has a limitation when it comes to pool resources. From the `resource-query` docs: > Please note that the granularity of exclusive allocation/reservation is the whole resource pool vertex, not...
Need to be able to make a reservation for resources at a fixed time in the future. Useful for many things, one example of which is a planned downtime of...
For the OST part of Lustre FS's on Rabbits, Fluxion needs to satisfy the requested aggregate amount of storage slices, but we want to be able to do so without...
Some changes to flux-sched break things over in flux-coral2, see e.g. https://github.com/flux-framework/flux-coral2/pull/23 . Would be nice if there was a way to be alerted that this was going to happen...
Following up on a coffee hour discussion with @grondo and @milroy , when users request persistent rabbit allocations which are independent of compute node jobs, users will need to be...