Unit 5 – Automation and Orche

stration (Arranging or Assigning

),
Automation focuses on making tasks easier and faster
for humans to perform.

Introduction

Previously we have discussed about

Motivation for cloud computing,
Physical infrastructure, and
Key virtualizations that cloud systems use.

Automation
-Why cloud systems need automated support
mechanisms.
-Aspects of automation
-Why so many automation systems have arisen.
Groups That Use Automation

Cloud automation mechanisms have been developed

for three categories of users:
Individual customers
Large cloud customers
Cloud providers

1. Individual customers.
- Individual subscribers often use SaaS apps -
document editing system
- Providers offer access through a web browser or
a dedicated app
- The interfaces presented to individuals hide
details, allowing a user to focus on using the services.
- Providers offer ways for individuals to try more
advanced services.
For example, to encourage individuals to create web
sites, providers offer
1. free trials with step-by-step instructions and
2. a point-and-click web interface.
The interface must be backed by underlying
automated systems that handle many works on behalf
of the individual
 2. Large cloud customers.
Organization that moves to a public cloud needs
tools to control and manage computing.
Types of automated tools for large cloud customers:
1. One type, available from the provider or a third
party, allows a customer to download and run the
tools to deploy and manage apps.
Examples:
i. Kubernetes, which automates deployment and
operation of a service built with containers
ii. Hadoop, that automates MapReduce
computations.

2 . The other type consists of tools offered by a

provider that allow large customers to configure,
deploy, and manage apps and services without
downloading software.
3. Cloud providers
Cloud providers designed most sophisticated and
advanced automation tools, and use them to manage
cloud data centers
Next section - consider the scope of the problems that
providers face and the ways automation can help.

Cloud Providers builds

- tools to configure, monitor, and manage
the underlying cloud infrastructure,
- tools that handle requests from cloud
customers automatically.
9.3 The Need For Automation In A Data Center

Tasks of building and operating a large data center:

1. Installing a raised floor,
2. Air conditioning facilities,
3. Thousands of racks,
4. Equipment must be loaded into each rack.
5. Each piece of equipment must be connected to
both a power source and to the data center network.
6. Each piece of equipment must be assigned a
network address, and
7. The switches must be configured to forward
packets along a shor- test path from any source to any
destination,
8. Each server must be configured to run
appropriate software. For example, a server that will
host VMs must be configured to run a type 1
hypervisor.

Operating a data center is much more complex

than operating IT facilities for a single
organization: Four aspects of a cloud data center
stand out.
1. Extreme scale
2. Diverse services
3. Constant change
4. Human error
1. Extreme scale:
Cloud provider must accommodate thousands of tenants:
Individuals or Enterprise customers.
Each enterprise customer may deploy dozens of VMs
and hundreds of containerized services.
The total number of services can become extremely large.

2. Diverse services:
Cloud data center provider allows each customer to choose
software and services to run.
Cloud data center may run software for thousands of
services at the same time.

3. Constant change:
Data center provider must handle dynamically changing
requirements.
At one extreme, provider will get ample time to configure
facilities for the new tenant.
At the other extreme, a tenant can request a new VM or
deploy a container at any time, and the data provider must
accommodate the request quickly.

4. Human error. Data center operators report that:

Many problems in a data center can be traced to human

error.
 9.4 An Example Deployment – To understand
data center operations

Example steps a provider may take when deploying

a VM

1. Choose a server on which to run the VM

2. Configure the hypervisor on the server to run the
VM Assign an IP address to the VM
3. Configure the network to forward packets for the
VM, which may involve configuring the tenant’s
virtual network (e.g., VXLAN)
4. Choose a remote disk server and allocate storage
for the VM, according to the tenant’s specification
5. Configure the hypervisor to send requests from
the VM to the storage server

Performing each step manually will require a

significant amount of time.
Automation is needed to perform operations quickly
and to accommodate the scale of a data center.
9.5 What Can Be Automated?

Involve the management of data and computer software, almost

all operational tasks in a data center can be automated.
Example, Data centers t h a t use a “lights-out” approach - >
Automated systems run the data center.

1. Creation and deployment of new virtual resources

2. Workload monitoring and accounting

3. Optimizations

4. Safety and recovery

5. Software update and upgrade

6. Administration of security policies

 9.6 Levels Of Automation

An example model with five levels of automation. Successively

higher numbers denote increased automation.

5 Automated remediation of problems

4 Automated identification of root causes
3 Automated analysis of trends and prediction
2 Automated monitoring and measurement
1 Automated preparation and configuration
0 No automation (manual operation)
Level 1: Automated preparation and configuration. Level 1 refers
to the automa- tion of tasks that are performed before installation
occurs.
Level 2: Automated monitoring and measurement. Level 2
automation refers to monitoring a data center and making
measurements available to human operators.
For example, in addition to servers,
network links and switchs, and
the disks used in storage facilities,
power and cooling equipment.
Performance, and
load on each server,
Traffic on each link in the network, and
Performance of storage systems.
Mechanisms to alert a data center operator For example, an alert
might be raised if a SAN storage system crashes or the load on a net
work link exceeds 95% of the link capacity.
 Level 3: Automated analysis of trends and prediction.
Can collect measurements over a long period and use Analytic
software to analyze changes and trends.
Example, Average load on a given network link increases at a
rate of 3% each month.
Analytic software monitor thousands of links without missing
longterm trends.
Key advantage of level 3 analysis lies in the
ability to predict needs,
allowing the data center owner to plan ahead rather than waiting
for a crisis to occur.
Level 4: Automated identification of root causes.
Uses data gathered from monitoring along with knowledge of
both the data center infrastructure and layers of virtualization
Employs Artificial Intelligence (AI) technologies that can
understand relationships and reason about cause and effect.
Level 5: Automated remediation of problems
Remedial actions may be straightforward (e.g., restarting a VM
that has crashed or rebooting a storage server) or
complicated (e.g., running a program to reconfigure network
forwarding for a particular destination or a set of destinations).
Automated system may be incapable of solving all problems in a
data center.
Until robots become available that can replace defective
equipment, a human is required to provide a permanent solution.
Note:
Levels 3 and above may use machine learning (ML) software. The top
levels may use additional forms of Artificial Intelligence (AI).
AIops (Artificial Intelligence operations) - an automation system
that uses AI and can operate a data center without human
intervention.
9.8 A Plethora (abundance) Of Automation Tools

Management complexity - Operating a data center is an

extremely complex task
Automating each small task independently - It is easiest
to automate each small part of the task independently

Many data center automation tools exist because

each tool handles one small task. Small tools work
well for tasks that involve details because using a tool
can reduce errors.

.
9.9 Automation Of Manual Data Center Practices
To understand the tools, one must know how humans
operated data centers.
1 A tenant signs a contract for a new VM and a
new work order (i.e., a “ticket”) is created
2 A human from the group that handles VMs
configures a new VM and passes the ticket on
3 A human from the group that handles
networking configures the network and
passes the ticket on
4 A human from the group that handles storage
configures a SAN server and passes the ticket
on
5 The tenant is notified that the VM is ready
10. Zero Touch Provisioning And Infrastructure As Code
Zero Touch Provisioning (ZTP) and Infrastructure as Code
(IaC) to refer to a process where a data center operator creates a
specification and uses software to read the specification and
configure underlying systems.

Two approaches have been used: push and pull.

The push version follows the traditional pattern of installing a
configuration: a tool reads a specification and performs the
commands needed to configure the underlying system.
The pull version requires an entity to initiate configuration.
For example, when a new software system starts, the system can
be designed to pull configuration information from a server.

11. Declarative, Imperative, And Intent-Based

Specifications

The specifications used with automated tools can take many

forms, Two aspects have become important:
Declarative vs. imperative
Intent-based vs. detailed
 Declarative vs. imperative.

1. An imperative specification states an action to be performed.

For example, an imperative specification to assign an IP
address to a VM might have the form:
Assign IP address 192.168.1.17 to the VM’s Ethernet
interface

2. A declarative specification focuses on the result rather

than the steps used to achieve it.
For example, a declarative specification for IP address
assignment might have the form:

Main IP address: 192.168.1.17

Intent-based vs. detailed.

3. Intent-based specification allows a human to state the

desired outcome without giving details about how to achieve the
desired outcome or specific values to be used.
For example, an intent-based specification for IP address
assignment might state:

Each VM receives a unique IP address on the tenant’s IP

subnet

without specifying the IP address to be used on each VM.

Intent-based specifications offer generality and flexibility.
allow many possible implementations to be used.
allow configuration tools the freedom to assign any values that
produce the desired outcome and thereby encourage innovation.