Cloud Computing

Question 1

4 Types of Services

Answer 1

Compute - virtual computer running apps, programs

Storage - virtual hard drive storing files

Networking - virtual network defining internet connections of network isolations

Databases - virtual DB for storing data

Question 2

Benefits of Cloud

Answer 2

Cost Effective No upfront cost, PAYG

Global Launch anywhere in the world

Reliable Data back up, revoery, fault tolerance

Scalable Increase and decrease as necessary

Elastic Automate scaling during spikes and drops

Current Hardware managed and patched without interuption

Secure Secure by default or configured to granular level

Question 3

SaaS

Answer 3

Software managed by cloud provider. Don’t worry about anything else.

Gmail, Salesforce.

For customers who use it

Question 4

PaaS

Answer 4

We worry about deployment of our apps, but not about the underlying OS or infra

For developers

Question 5

IaaS

Answer 5

Basic building blocks of cloud IT.

Access to networking, computers, storage

Don’t worry about physical location, IT staff, data centers, hardware.

For Admins

Question 6

3 Deployment models

Answer 6

Public Cloud - All built on cloud provider via Virtual network

Private Cloud - On prem in own data center

Hybrid - Connecting on prem with public cloud

Question 7

Capex vs Opex

Answer 7

Capex - Upfront cost, deduct from tax bill over time, guess on what you need then grow into it

Opex - No upfront cost, lease what you need when you need it, less slack, deduct from tax bill in year.

Question 8

Availability

Answer 8

Your ability to ensure service remains available (highly available, or HA)

Ensure no single point of failure to ensure certain level of performance

Question 9

Availability Zones

Answer 9

Physical location made up of one of more data centers.

Having redundant servers in different data centers

Data centers sometimes referred to as AZ’s, but can be more than one.

If one goes down, the other can take over.

Question 10

Load balancer

Answer 10

Evenly distributes traffic to multiple servers in one or more datacenters

Distributes traffic across several Availability Zones to ensure high availability.

Once server is not functioning, traffic is stopped.

Question 11

Scalability

Answer 11

Ability to grow rapidly or unimpeded

Question 12

Vertical vs Horizontal scaling

Answer 12

Vertical - scaling up, get a bigger server

Horizontal - Add more servers the same size.

Question 13

Elasticity

Answer 13

Ability to shrink and grow to meet demand

Like scalability, but happens automatically

Done by scaling out (horizontal scaling) i.e. adding more VMs/servers

Question 14

Scale sets

Answer 14

Azure virtual machine scale sets let you create and manage a group of load balanced VMs.

The number of VM instances can automatically increase or decrease in response to demand or a defined schedule.

Question 15

Fault tolerance

Answer 15

Ability to prevent failure

Question 16

Disaster recovery

Answer 16

Ability to recover from a failure (Highly durable, or DR)

Question 17

Region

Answer 17

Grouping of several data centers (aka availability zones)

Each region contains around 3 availability zones (generally)

Question 18

Geography

Answer 18

Grouping of several regions that preserve data residency and compliance boundaries

Question 19

Paired Regions

Answer 19

Each region paired with another >300 miles away (US east and US west)

One of these is updated at a time for high availability

Question 20

Geo-Redundant storage (GRS)

Answer 20

Replicates data to a 2nd region automatically if primary region is unavailable.

Question 21

Recommended Regions

Answer 21

Provides broadest range of Azure services

Support availability zones

Other Regions may not support AZ’s

Question 22

Special Regions

Answer 22

Setup to meet compliance and legal reasons
e.g. US gov Iowa, Virginia, plus others in US.
China East and North.

Government employees

Question 23

Availability Zones:

Fault Domain

Answer 23

Fault domain is set of hardware components that share a single point of failure. E.g. if 4 servers are grouped and share power, cooling e.g. they are a single fault domain.

If the 4 servers each have individual power, cooling etc, they are more fault tolerant.

https://channel9.msdn.com/Blogs/windowsserver/Fault-Domain-Awareness-in-WS2016-Part-1-Overview

Question 24

Availability Zones:

Update Domain

Answer 24

Azure updates underlying software (takes everything offline)

Having an update domains means updates are staggered.

Question 25

Availability Zones:

Availability sets

Answer 25

Separate deployments in the same datacenter.

Logical groups of two or more VM’s that keep an application available during planned/unplanned maintenance.

Question 26

Public Cloud

Answer 26

All built on Cloud aka Cloud Native

+ Most cost effective
+ Easy to setup (no expert knowledge)
+ Security by default

• Limited customization
• Might not meet stringent security/compliance

Question 27

Private Cloud

Answer 27

All built on customer’s data centers aka On Prem

+ Customisable
+ Security: Can be as secure as you make it.

• Expensive
• Expertise required to setup
• You are responsible for security

Question 28

Hybrid Cloud

Answer 28

Connect On Prem and Cloud via ExpressRoute

+ Best of both worlds
+ Security: Can be as secure as you make it.

• Expertise required to setup
• You have to secure connection to cloud

Question 29

Azure Zones

Answer 29

Geographical grouping of Azure regions

Billing based on data transfers between zones.