Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

Rethink watcloud cli

Aperta
  • Qualcosa non va?
#10 3 commenti 0 reazioni 0 assegnatari Vedi su GitHub

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
5/5
Tempo stimato
Più di una settimana
Idoneità per principianti
25/100
Tipo di issue
Refactoring
Chiarezza
Abbastanza chiara
Stato di attività
Attiva
Stack tecnologico
docker, go
Ambito
cli, devops, infrastructure

Direzione di ricerca

Start by inventorying the watcloud CLI command implementations and the installation configuration; the issue names status, quota, docker, config, slurm capacity, du, and srun-related work but no files or tests. Separate the proposed removals, fixes, and additions into independently scoped tasks, then identify the existing command entry points and validation coverage. Done is not a single change: each selected command must match the behavior and reporting requirements described in the issue.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

I'd like to propose a round of changes to the watcloud CLI, for three reasons:

  1. The command interface is confusing.
  2. Some commands aren't useful, some print more than anyone needs, some don't work well, and some don't work at all. Most aren't nice to look at.
  3. No user has used it.

On that last point — the "Maintenance Nodes" section of watcloud status has never worked, and nobody has filed a bug. The commands also appear nowhere in /docs/compute-cluster and nowhere in the login MOTD.

What I think the tool is for. Right now it's a collection of readouts. A cluster user has a small number of real questions — what can I get right now, how much of what I have am I using, why did my job fail — and the tool should answer those well and skip everything else. Most of what follows falls out of that.

Fix

1. watcloud status — remove it. It re-renders status.watonomous.ca in a terminal, worse. Its maintenance section has never worked. status.watonomous.ca is hosted on GitHub Pages.

2. watcloud quota memory measures something other than the limit. The limit is enforced on the kernel's count of your memory, which differs from the per-process sum the command shows, and the two can diverge in either direction. It should show the kernel's number.

3. watcloud quota cpu reports the whole node, not your share. It prints one row per core for every core on the machine — 80 of them on the larger compute nodes — and the utilisation shown is everyone's work combined, including other people's jobs. It should show only the CPUs allocated to your job, and how much of that allocation you're actually using.

4. watcloud quota disk invents a temp disk number outside a job. Home disk is right everywhere. /tmp only has an allocation inside a SLURM job, so on a login node or a direct SSH to a compute node there is nothing to report and the row invents a red 100% of zero bytes. Show the real figure where there is one, and leave the row out where there isn't.

5. watcloud docker status — remove it. On a healthy login node it shows a red "stopped" for a daemon that isn't meant to run there, and it reports Slurm's state under a command called docker. Red should mean something is wrong.

6. The size argument to watcloud docker start — remove it. The value is silently discarded, and it couldn't work anyway, because scratch disk is fixed when the job is created. A documented option that does nothing is worse than no option.

7. watcloud config — remove it. A whole command surface for a single setting. Not worth it until there's more than one.

8. The partitions and job-queue blocks of watcloud slurm capacity — remove them. Partitions is two near-identical rows; that's one line of information. The job queue repeats jobs already listed directly above it. Neither answers "what can I get right now."

9. watcloud slurm capacity shows two of the four resources you can request. Scratch disk and GPU sharing are both invisible, and those are the two people most often get wrong. It can also report 25 GiB of GPU memory free on a node with 576 MiB left.

10. Long names break the watcloud slurm capacity table. Usernames and job names overflow their columns and push everything after them out of line. Truncate them.

11. Install it only where it's used. It currently lands on all 74 managed hosts, including storage nodes nobody logs into. It belongs on the machines users actually touch.

Add

12. watcloud du — an instant home-directory explorer. ncdu-like disk usage explorer for the home directory, using getfattr -n ceph.dir.rbytes <dir> instead of recursively adding up disk usage

13. A command that composes the srun line. The most common unrecoverable mistake is asking for too little scratch disk — you find out when a container image fails to download, and by then the job has to be thrown away and resubmitted. A command that builds the request makes that decision at the only moment it can be made.

14. Show users their GPU usage against what they asked for. We've been recording per-job GPU memory and utilisation since 2024 and nobody has ever looked at it. One verified example: a job that requested 4 GiB of GPU memory used 21.5 GiB. A line at job exit saying so is feedback that exists nowhere today.

Smaller things

No way to ask which version is installed. No machine-readable output. Every command reports success even when it failed. No shell completion. The README's Features section is empty.

Lingua principale
Go
Stelle
0
Fork
1
Merge medio
1g 2h
PR unite (30g)
4

Preparare l'ambiente

Questo progetto non fornisce container di sviluppo, Dockerfile né guida per i contributori, quindi l'ambiente è a tuo carico: parti dal suo README e consulta la nostra guida al primo contributo per i passaggi generali.

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di WATonomous/watcloud-cli

Tutte le issue di WATonomous/watcloud-cli

Issue simili

Altre issue su Go

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.