Mark Grondona
Mark Grondona
Looks like numbers like `.5` or `5.` are somewhat unexpectedly encoded as strings with `-o, --setopt`. Not a big deal, but probably will cause some confusion in the near term...
As noted in #5518, libsubprocess has a set of fixed buffers (though configurable in size), with a default of 4M. When the buffer is full, `flux_subprocess_write()` returns a partial write,...
It would be useful to be able to launch a subinstance in a system instance which is also capable of running multiuser jobs. This feature in combination with user-based access...
@jameshcorbett reported that a flux-coral2 CI test is failing with most recent flux-core+flux-sched with the following log: ``` ok 1 - job-manager: load dws-jobtap plugin PASS: t1000-dws-dependencies.t 1 - job-manager:...
A user was asking for the equivalent of ``` sacct --format="MaxRSS" -j ``` in Flux. AFAIK, we don't have an issue open on basic resource usage accounting for Flux, though...
Got this verbose set of messages after an allocation expiration of a `flux alloc` job on 36 nodes. Hm, maybe this is only so bad because the job happened to...
A group of users have run into some usability issues with the way Flux allocates cores and GPUs. Ideally they'd like greedy or perhaps exclusive allocation of cores-per-task when only...
In #5443 a couple missing checks in `configure` were discovered: - The python ply package - libsodium I'm not sure about libsodium. It seems to only be required by `flux-keygen`,...
As mentioned in a meeting yesterday, during preparation for scale testing it was noted that `flux resource list` response times were excessively slow when Fluxion was the loaded scheduler module....
RFC 27 now defines a [sched.expiration](https://flux-framework.readthedocs.io/projects/flux-rfc/en/latest/spec_27.html#expiration) RPC which schedulers may implement to support adjustment to the duration/expiration of a running job. This RPC should be implemented in Fluxion so that...