Differences

This shows you the differences between two versions of the page.

--- hpc:how_to_use_openondemand [2023/11/10 16:42] – Adrien Albert
+++ hpc:how_to_use_openondemand [2024/06/27 09:21] (current) – [How To connect] Yann Sagon
@@ Line 31: / Line 31: @@
 ====== How To connect ======
-First each cluster will have his own openondemand instance. To allow all users (unige, extern and outsider users) to connect to ondemand, you must to authenticate through switch.
+First each cluster will have his own OpenOnDemand instance. To allow all users (unige, extern and outsider users) to connect to OpenOnDemand, you must to authenticate through switch.
-You MUST use the same bind email to the unige account. It means:
+You MUST use the same email bound to your HPC account. It means:
   * Unige email (@unige.ch) for unige members (Student/collaborator/external)
   * Email used for registration for Outsider members
-If not, you will not be able to connect through OnDemand web Interface
+If not, you will not be able to connect through OpenOnDemand web Interface
-To access to Baobab OnDemand web app: (https://ondemand.baobab.hpc.unige.ch/) \\
+Access to Baobab OpenOnDemand web instance: https://openondemand.baobab.hpc.unige.ch \\
-** :!: Yggdrasil Ondemand: Not implemented yet**
+Access to Bamboo OpenOnDemand web instance: https://openondemand.bamboo.hpc.unige.ch (need to be connected through UNIGE VPN)\\
+** :!: Yggdrasil OpenOndemand: Not implemented yet**
 ====== Tabs ======
 ===== Dashboard =====
@@ Line 104: / Line 107: @@
 > :!: Some adjustment are required to make your own app work.
 ===== Logout =====
 To Logout you need to close you browser and not only the current tab.
+===== Launching applications =====
+This documentation has been taken from https://docs.hpc.qmul.ac.uk/ondemand/
+After choosing from the available applications, you can requested the necessary resource for the session, and then click Launch.
+{{:hpc:hpc:resources.png?nolink&600|}}
+Typically, when requesting resources, you will be given the option to choose number of cores, maximum running time and memory per core. If multiple versions of the application are available, this may be offered too.
+For many use-cases, including non-computational visualisation, and many Jupyter/R Studio sessions, a maximum of 1 core will be used by the application.
+There is also the option to request an email to notify you when the job starts, in the case of sessions which require longer queueing time.
+For graphical applications, once the session is established, you have the option to alter the image quality and compression level. For most cases, using the defaults should be fine, but for precise graphical work, or taking screenshots, you may want to increase the image quality.
+{{:hpc:quality.png?nolink&600|}}
+<WRAP center round info 60%>
+Note also the **View only** link. OnDemand sessions are secure and confined only to you. However, you have the option to share a **read-only view** of your session for purposes of monitoring progress of a job with a collaborator or to demonstrate work to a colleague.
+</WRAP>
+==== Session persistence ====
+A session remains active for the time period you have requested it for. If you have finished your work, then we recommend that you close the application cleanly, followed by clicking Delete next to your session on My Interactive Sessions page. This will release resources back to the cluster.
+If you are running a graphical application which uses a desktop environment, any computation will continue even after the browser tab is closed. Clicking Launch again from the My Interactive Sessions screen will reconnect to the session.
+Closing a browser tab while running a web service session such as Jupyter  will interrupt computation, in the same way as it would if you did the same while running Jupyter locally on your own PC. Jupyter and both offer ways to save and checkpoint your work if you want to resume later.
+{{:hpc:interactive_app.png?nolink&700|}}