Slurm
Running, waiting and finished jobs in three sections that collapse and resize like the ones in Files: live squeue, and recent sacct history with exit codes and memory, where jobs that failed, timed out or ran out of memory are marked in red. A fourth section at the bottom shows what the cluster has free and in use (CPUs, memory, GPUs and the jobs waiting, for the whole cluster and per partition). Click a job for its log viewer, which follows stdout/stderr; a job that hasn't finished opens on its details instead. The This project / All jobs switch in the header chooses between the project's jobs (the default) and all your jobs. A job counts as the project's if its Slurm comment is the project's tag, codeome:<project>-<id>, which Codeome adds with sbatch --comment when you submit through it, or if it ran inside one of the project's saved HPC folders (see Project settings). Jobs started outside Codeome, such as ehio or Drakkar batches, appear under This project only when their working folder is inside a saved folder. With no folders saved, the folder open in the HPC Files tab is used, and while that is ~ every job is shown. Also cancel, submit (sbatch) from the editor, and a job-script template.