Merge pull request #10 from Sydney-Informatics-Hub/dgx-workshop-updates

edit the guide in preparation for the workshop

Author: jx-fan

_quarto.yml

@@ -18,28 +18,33 @@ website:
         contents:
         - text: "Introduction to the SIH GPU Cluster"
           href: notebooks/dgx.html
+        - text: "Accessing the SIH GPU Cluster"
+          href: notebooks/access.html
       - section: "Getting Started with Run:ai"
         contents:
         - text: "Login"
           href: notebooks/login.html
         - text: "Navigating the User Interface"
           href: notebooks/user_interface.html
-      - section: "How-to Guides"
+      - section: "Run:ai Features"
         contents:
-        - text: "How to Use the Dashboards"
+        - text: "Dashboards"
           href: notebooks/dashboards.html
-        - text: "How to Manage Projects"
+        - text: "Projects"
           href: notebooks/projects.html
-        - text: "How to Configure Environments"
+        - text: "Environments"
           href: notebooks/environments.html
+        - text: "Data Sources"
+          href: notebooks/data_sources.html
+
+      - section: "Tutorials"
+        contents:
+        - text: "Creating a basic Jupyter Lab workload"
+          href: notebooks/jupyter_tutorial.html
         - text: "How to transfer data to/from the SIH GPU Cluster"
           href: notebooks/data_transfer.html
         - text: "How to run a terminal using the Command Line Interface (CLI)"
           href: notebooks/CLI.html
-      - section: "Tutorials"
-        contents:
-        - text: "Creating a Jupyter Lab Workload"
-          href: notebooks/jupyter_tutorial.html
   cookie-consent: 
     type: express
     style: simple

fig/dashboards_overview_clusterload.png

@@ -9,11 +9,12 @@ output:
 :::{.callout-note}
 The SIH GPU Cluster is managed by the [Sydney Informatics Hub](https://www.sydney.edu.au/research/facilities/sydney-informatics-hub.html) at the University of Sydney. Contact <sih.info@sydney.edu.au> for more questions.
 :::
-
+This Onboarding Guide is structured as follows:
 
 |  |  |
 | --- | :--- |
 | **Section** | **Content** |
-| Getting Started with the SIH GPU Cluster | Cluster fundamentals |
-| Getting Started with Run:ai | Run:ai fundamentals |
-| How-to Guides | Provide step-by-step instructions |
+| Getting Started with the SIH GPU Cluster | Overview and basic information about the GPU Cluster |
+| Getting Started with Run:ai | Introduction to Run:ai concepts and fundamentals |
+| Run:ai Features | Step-by-step instructions for using various Run:ai features. More details can be found in the [official Run:ai user guide (version 2.18)](https://docs.run.ai/v2.18/Researcher/overview-researcher/) |
+| Tutorials | Practical examples and demonstrations of different use cases |

fig/environment_archetechture.png fig/environment_architecture.pngfig/environment_archetechture.png renamed to fig/environment_architecture.png

@@ -0,0 +1,10 @@
+# Accessing the SIH GPU Cluster
+
+For early acccess to the SIH GPU Cluster, please directly contact the [Sydney Informatics Hub](https://www.sydney.edu.au/research/facilities/sydney-informatics-hub.html) with the following information:
+
+- **DashR Research Project Shortcode**: Please log in to the [DashR portal](https://dashr.sydney.edu.au/projects), click on the project you intend to work on, and locate the *Project Short Code*. This is essential for linking your compute usage to the correct research allocation.
+- **Billing Code**: This should be the internal billing code associated with your project or department. If you are unsure, please consult your supervisor or finance administrator.
+
+:::{.callout-note}
+**Work in Progress**: Long term project provisioning and access to the SIH GPU Cluster will be streamlined and directly managed on the university's [Researcher Dashboard (DashR)](https://dashr.sydney.edu.au/).
+:::

fig/environment_predefined.png

@@ -1,29 +1,35 @@
-# How to Use the Dashboards
+# Run:ai Dashboards
 ## Overview
 This dashboard view provides holistic infrastructure information useful for both researchers and system administrators in managing and planning the resources.
 
 ### Indicators
 This section presents high-level statistics of the GPU computing resources
-![](../fig/dashboards_overview_indicators.png)
+
+![The "Indicators" panel under "Overview"](../fig/dashboards_overview_indicators.png)
 
 ### Cluster Load
 Real-time monitoring of the cluster status in terms of GPU and CPU utilisation.
-![](../fig/dashboards_overview_clusterload.png)
+
+![Monitoring system-wide workload](../fig/dashboards_overview_clusterload.png)
 
 ### Queueing
-Inspecting all queueing jobs. Possible reasons why jobs are queueing include:
+Inspecting queueing jobs. Possible reasons why your jobs are queueing include:
 
-- The number of GPUs requested to be allocated to the job has exceeded the remaining GPUs in the project.
+- The number of GPUs requested to be allocated to the job has exceeded the remaining GPU quota in the project.
+- The GPU cluster is currently at full capacity and therefore has no available resources to schedule the job.
 - The job is waiting for other jobs to finish before it can be scheduled.
-![](../fig/dashboards_overview_queueing.png)
+
+![Queueing jobs](../fig/dashboards_overview_queueing.png)
 
 ### Idle GPUs
 Displaying the number of idle GPUs currently allocated to running workloads.
-![](../fig/dashboards_overview_idleGPUs.png)
+
+![Idle GPUs](../fig/dashboards_overview_idleGPUs.png)
 
 ### Running Workloads
 Summary of the list of running workloads.
-![](../fig/dashboards_overview_workloads.png)
+
+![Running workloads](../fig/dashboards_overview_workloads.png)
 
 ## Analytics
 This dashboard provides more detailed breakdowns of the SIH GPU running status. Key
@@ -33,4 +39,5 @@ statistics that are reported at separate levels:
 - Project
 - Workloads
 - Nodes
-![](../fig/dashboards_analytics.png)
+
+![System analytics](../fig/dashboards_analytics.png)

fig/project_example.png

@@ -0,0 +1 @@
+# Data Sources