Mark Grondona

> In RFC 33 we did define an [access policy](https://flux-framework.readthedocs.io/projects/flux-rfc/en/latest/spec_33.html#access-policy), so if we didn't want to load mf_priority in a subinstance, we could potentially generate a list of allowed users...

For reference, here is a snippet of how Slurm accounts for reservations: > Jobs executed within a reservation are accounted for using the appropriate user and bank account. If resources...

Good question @trws. And if we need to use a bank/account to control access to a DAT job, then we would need some way to create the access control list...

An alternate solution for DATs was proposed in today's meeting. The proposal AIUI was to offer an initially empty queue for use by DATs, then have the reservation assign resources...

That may need more thought. Should that also during an unintended restart, a restart due to an upgrade?

Canonical jobspec supports this concept as described in the RFC 14 [Resources](https://flux-framework.readthedocs.io/projects/flux-rfc/en/latest/spec_14.html#resources) section (see the description of the `count:` `min`, `max`, `operator` and `operand` keys). However, Jobspec V1 requires `count`...

I think Fluxion does support some portion of canonical Jobspec beyond V1. I'd check in the testsuite for hints, e.g. `t/t3009-resource-minmax.t` and `t/t3031-resource-minmax2.t` look promising.

There's an example jobspec using min/max referenced by those tests here https://github.com/flux-framework/flux-sched/blob/fe872c8dc056934e4073b5fb2932335bb69ca73a/t/data/resource/jobspecs/min_max/test002.yaml#L1-L18

> Perform a satisfiability check rather than a match "can we represent this or not" (yes/no) Flux schedulers may offer a `sched.feasibility` RPC for doing a "satisfiability" check. Fluxion supports...

I'm wondering if sdexec will need assistance from the IMP here, since the processes running within the cgroup are owned by other users that the system instance user. Probably needs...