Skip to content

Latest commit

 

History

History
163 lines (106 loc) · 3.49 KB

Azure_VM_Setup_DP.md

File metadata and controls

163 lines (106 loc) · 3.49 KB

Microsoft Azure Virtual Machine Setup for Data Processing

Updated 2020-11-12

Note: These are example instructions and various specifics will need to be modified.

Create a VM From the Web Interface

Go to https://portal.azure.com and log in

Click on "Virtual machines"

Click on "Add" or "Create a Virtual Machine"

Basics

  • Subscription: Free Trial

  • Resource group: ubuntu_trial (Create new if needed)

  • Virtual machine name: UbuntuTrial01

  • Region: East

  • Availability options: No infrastructure redundancy required

  • Image: Ubuntu Server 18.04 LTS

  • Size: Standard D2s v3 (2 vcpus, 8GiB memory)

RSA Alternative:

  • Username: user
  • SSH public key: (contents of .ssh/id_rsa.pub)

Password Alternative:

  • Authentication type: Password

  • Username: user

  • Password: pass

  • Public inbound ports: Allow selected ports

  • Selected inbound ports: HTTP, SSH

Disks

  • OS disk type: Premium SSD

  • Advanced: Use managed disks -- Yes; Use ephemeral OS disk -- No

  • Data disks: Create and attach a new disk

  • Name: UbuntuTrial01_DataDisk_0

  • Source type: None

  • Size: 1023 GiB (Premium SSD)

Click "OK" button

  • LUN: 0; NAME; SIZE; DIST TYPE HOST CACHING: None

Networking

  • Virtual network: (new) ubuntu_trial-vnet

  • Subnet: (new) default (10.0.0.0/24)

  • Public IP: (new) UbunutuTrial01-ip

  • NIC network security group: Basic

  • Public inbound ports: Allow selected ports

  • Select inboundn ports: HTTP, HTTPS, SSH

  • Accelerated networking: Off

  • Load balancing: No

Management

  • Enable basic plan for free

  • Boot diagnostics: On

  • OS guest diagnostics: Off

  • Diagnosstics storage account: ubuntutrialdiag

  • Identify: Off

  • Azure Active Directory: Off

  • Auto-shutdown: Off

  • Backup: Off

Advanced

Nothing

Tags

Nothing

Review and create

Click "Create"

Validation passed -- 0.0960 USD/hr

... Your deployment is underway ...

Various things appear:

  • UbuntuTrial01 -- virtualMachines
  • ubuntutrial01295 -- networkInterfaces
  • UbuntuTrial01_DataDist_0 -- disks
  • ubuntutrialdiag -- storageAccounts
  • UbuntuTrial01-ip -- publicIpAddresses
  • UbuntuTrial01-nsg -- networkSecurityGroups
  • ubuntu_trial-vnet -- virtualNetworks

When everything is finished: Click "Go to resource"

Setup of the VM is now finished. From here on, provisioning takes place inside the instance.

Log in to the VM from a terminal

ssh <ip address>

Mount the Data Disk

In this step we mount the "Data disk" we created and attached to the VM.

General instructions are available at: https://docs.microsoft.com/en-us/azure/virtual-machines/linux/attach-disk-portal#connect-to-the-linux-vm-to-mount-the-new-disk

Use dmesg | grep SCSI to find the data disk -- sdc.

Skip the partitioning step

Write the file system

sudo mkfs.xfs /dev/sdc

From Bill Broomall:

xfs has a number of advantages, but two are especially convenient: it's much quicker to format; and it's very easy to grow (using xfs_growfs). When you make the filesystem directly on the volume, you don't have to resize the partition. In Azure, you have to detach the volume first (or shut down the VM), but once you resize the volume, on reboot just do 'xfs_growfs /var/www/html.

Write a file system with:

Mount the file system

sudo mkdir -p /var/www/html
sudo mount /dev/sdc /var/www/html

Find the UUID of the new drive with:

sudo -i blkid | grep "/dev/sdc"

Edit /etc/fstab to have this line:

UUID=<uuid-from-blkid-cmd> /var/www/html    xfs    defaults    0 0