OCI:Create account: Difference between revisions

From DISI
Jump to navigation Jump to search
(asfd)
Line 62: Line 62:
allow service compute_management to read app-catalog-listing in tenancy</nowiki>
allow service compute_management to read app-catalog-listing in tenancy</nowiki>


== Create Stack ==


* Go to marketplace, search for "hpc". Find a stack called simply "HPC Cluster". Click on that, accept the agreement, select the compartment you just created in the drop-down menu, and click the big "Launch Stack" button.


* Before we get into this- use ssh-keygen to create an ssh keypair, which you will need.


* Click next to proceed past the first step, "Stack Information". In the second step, "Configure Variables", paste the ssh key you just generated into the ssh key field. Keep all default options, except for the following:


* Headnode options -> Availabity domain -> select one of your choosing
* Compute node options -> Availability domain -> select one
* Compute node options -> Use cluster network -> disable




[[Category:OCI]]
[[Category:OCI]]

Revision as of 20:13, 22 March 2023

create a free account

https://www.oracle.com/cloud/

"Try OCI for free"

"Start for free"

Once you create the account, get to the "dashboard"

You want US East (Ashburn) https://cloud.oracle.com/?region=us-ashburn-1

create compartment

create a compartment slurm autoscaling in the create dynamic group

hamburger -> identity and security -> compartments ->

so you are now here: https://cloud.oracle.com/identity/compartments?region=us-ashburn-1

click on "create compartment", select your root compartment as the parent compartment (this is the default)

name: slurm-autoscaling (does not matter, any name will work) description: molecular docking work (does not matter)

When finished creating, click on the compartment in the menu to view details. Copy the compartment OCID for later use.

create dynamic group

hamburger -> identity and security -> Dynamic Groups -> Create dynamic group

Add the following policy to the Dynamic Group:

Any {instance.compartment.id = '<your compartment OCID>'}
  • you get the OCID from Identity>Compartments>slurm-autoscaling>OCID copy
  • then you go to Identity > Domains > Default > Dynamic Groups > Create Dynamic Group

policies

now go to Identity -> Policies select root compartment create policy

name: slurm-autoscaling-policy. (does not matter) description: molecular docking policy (actual value does not matter)

Add the following as policies, changing names where needed. We are using the names "SlurmDynamicGroup" and "SlurmAutoScaling" (use the manual editor):

Allow dynamic-group SlurmDynamicGroup to read app-catalog-listing in tenancy
Allow dynamic-group SlurmDynamicGroup to use tag-namespace in tenancy
Allow dynamic-group SlurmDynamicGroup to manage compute-management-family in compartment SlurmAutoScaling
Allow dynamic-group SlurmDynamicGroup to manage instance-family in compartment SlurmAutoScaling
Allow dynamic-group SlurmDynamicGroup to use virtual-network-family in compartment SlurmAutoScaling
Allow dynamic-group SlurmDynamicGroup to use volumes in compartment SlurmAutoScaling
Allow any-user to manage all-resources in compartment SlurmAutoScaling
allow service compute_management to use tag-namespace in tenancy
allow service compute_management to manage compute-management-family in tenancy
allow service compute_management to read app-catalog-listing in tenancy

Create Stack

  • Go to marketplace, search for "hpc". Find a stack called simply "HPC Cluster". Click on that, accept the agreement, select the compartment you just created in the drop-down menu, and click the big "Launch Stack" button.
  • Before we get into this- use ssh-keygen to create an ssh keypair, which you will need.
  • Click next to proceed past the first step, "Stack Information". In the second step, "Configure Variables", paste the ssh key you just generated into the ssh key field. Keep all default options, except for the following:
  • Headnode options -> Availabity domain -> select one of your choosing
  • Compute node options -> Availability domain -> select one
  • Compute node options -> Use cluster network -> disable