Reference¶
This section covers error messages, reservation events, support routing, the glossary, HPC vocabulary, group management, policy, and software.
| Page | Covers |
|---|---|
| Error Messages | Error messages and failure symptoms with their causes and fixes, for signing in, command-line launches, kubectl get pods statuses, the reservation app, and code and notebooks. |
| Reservation Events | The 15 Kubernetes events the reservation system writes on a GPU pod, what each means, and what to do; and the waiting states that produce no event. |
| Getting Help | Support tiers and contacts, the contents of a support ticket, response targets, 1:1 Consultation, and administrative requests. |
| Glossary | Definitions of the terms used across the documentation, in alphabetical order. |
| Coming from HPC | HPC and Slurm terms and their DSMLP equivalents, differences from Slurm behavior, the Slurm compatibility wrappers and their limitations, and unsupported HPC features. |
| Managing a Group | Privilege tiers, workspace manager actions, budgets and limits set by administrators, and the usage reports. |
| Policy | Data classification, appropriate use, hosting of services, use of shared resources, self-supporting programs, and scheduled maintenance. |
| Software | R and RStudio, MATLAB and GNU Octave, Stata, licensed software, complex and experimental capabilities, and software installed without a ticket. |