Manage Instance

In this document, you learn how to manage a compute instance from creation to deletion. Available Instance Types are bare metal and virtual machine (VM).

Launch an Instance

  1. Log in to Intel® Tiber™ AI Cloud.

  2. In the left side menu, click Catalog > Hardware.

  3. In the Hardware tab, select an instance from the options.

Configure an instance

Select options from pull-down menus to configure your instance.

  1. From Instance family, select your choice.

  2. From Instance Type, select your choice.

    Tip

    Optionally, select Compare instance types to compare instances’ specifications.

  3. From Machine image, select your choice.

  4. In Instance name, enter a name in lowercase. Optional: Use hyphens.

  5. Under Public Keys, select checkbox for SSH public keys added previously.

  6. Optional: Quick Connect. Select the radio button enable single-click access.

    Note

    Currently, Quick Connect is only available in selected regions.

    Selecting Quick Connect is performed once. You cannot add or remove this functionality after you launch an instance.

  7. Click on Launch to launch an instance.

  8. View the column State.

  9. Wait for for State to show Ready.

Continue at Connect to an Instance.

Instance States

State reflects the status of a comptue instance.

Provisioning

  1. After launching an instance, State shows Provisioning in the dashboard.

  2. Wait until State shows Ready.

Ready

  1. When State shows Ready, you may launch your instance.

  2. A connection is established with the host.

  3. Continue to Connect to an Instance.

Connect to an Instance

To proceed, your instance must show Ready in the dashboard.

  1. Under Instance Name, click on your instance name. Instance details are shown in the Details tab.

  2. Choose a method to connect.

    Tip

    You must select Quick Connect before launching an instance for this option to be available. See Configure an instance.

    1. Quick Connect via JupyterLab.

      1. Click the Connect button.

      2. Select Terminal to access instance.

      3. You’re all set.

    2. Connect via Terminal from your local machine.

      1. Select How to Connect via SSH.

      2. A new pop-up dialog appears:How to connect to your instance.

      3. Continue in next section.

Connect via local Terminal

  1. Follow the onscreen instructions in the dialog.

    1. Select your operating system (OS).

    2. Follow instructions.

    3. Copy the command shown to connect to your instance.

    4. Open a Terminal.

  2. In the Terminal, paste the command you copied and press enter.

  3. If prompted to add your public key, select Yes.

  4. After launching instance, run command to confirm Ubuntu 22.04 (or other).

    cat /etc/os-release

  5. Your Terminal output should appear similar to the one below.

Edit an Instance

  1. Navigate to Compute > Instances from main console.

  2. In the Instances tab, assure that your instance appears.

  3. With your instance, select Edit under Actions.

  4. In the page “Edit Instance”, modify settings as desired.

Example - Upload Public SSH Key

  1. Select Create Key.

  2. A dialog, “Upload a Public Key”, appears.

  3. Follow all instructions in the dialog.

  4. Finally, select Create Key.

Update OS and add packages

While in an SSH session, you can add Ubuntu 22.04 packages and update your OS.

  1. To update and upgrade your OS, enter one command at a time.

    sudo apt-get update -y
sudo apt-get upgrade -y

  2. Add net-tools or curl

    sudo apt-get install net-tools

    sudo apt-get install curl

Delete an Instance

  1. Navigate to Compute > Instances from main console.

  2. Under Actions, select the Delete button.

  3. At the dialog “Delete instance”, select Delete to confirm your choice.

  4. Select Cancel if you do not wish to delete your instance.

Instance Types

Instance Types

Type

Description

Current offerings

Sizes

Typical Use Case(s)

Bare Metal (BM)

Bare metal compute instances are intended for general use and deploying CPU, GPU, and AI-accelerated processors.

4th Generation Intel® Xeon® Scalable processors; Intel® Max Series GPU (PVC); Intel® Gaudi® 2 processor

256GB and 1TB

AI and core computing.

Virtual Machine (VM)

VM compute instances are intended for managing workloads in CPU and AI applications. They help support developers world-wide to test and experiment with on-demand workloads and applications. A VM requires a hypervisor, which consumes some of its computing power.

4th Generation Intel® Xeon® Scalable processors; Intel® Data Center GPU Max Series.

16GB (small), 32GB (medium), and 64GB (large).

Workload testing and application development using CPUs, GPUs, and memory in the Intel ecosystem.