Cloud Computing

Service Level Agreement

sadiakhan4422@gmail.com
 Cloud Computing
 What is Service Level Agreement
 SLA Contents
 Web service SLA
 Difference between Web service SLA & Cloud SLA
 Types of SLA
 Non-negotiable SLA
 Negotiable SLA
 Negotiation via trusted agent
 Negotiation via multiple trusted agent
 Service Level Objectives (SLOs)
 Service Level Management
 Cloud Computing
 Considerations for SLA
 SLA Requirements
 key Performance Indicators (KPIs)
 Industry-defined KPIs
 Metrics for Monitoring and Auditing
 SLA Requirements w.r.t Cloud Delivery Models
 Example Cloud SLAs
 What is Service Level Agreement

 the Service Level Agreement (SLA) - Formal contract

between a service provider (SP) and a service consumer.
 SLA: foundation of consumer’ trust on the provider.
 Purpose: to define a formal bases for performance and
availability which the SP guarantees to deliver.
 The objective of an SLA is to reduce area of potential
conflicts and their resolution before these issues get
materialized.
 Different vendors offer different SLA structures,
 service offerings,
 performance levels,
 negotiation opportunities.
4
 What is Service Level Agreement
 SLAs can also be used to select vendors
on the basis of
 data protection

 continuity

 cost.

 SLA contains Service Level Objectives (SLOs)

 It consists of a set of measurable attributes called SLA
parameters (95% performance) which are established by
some objectively measurable conditions, termed as Service
Level Objectives (SLOs).
 SLA & SLOs: bases of selection of Cloud Service Provider
 SLOs may vary consumer to consumer 5
 SLA Contents
 a set of services which the provider will deliver like IaaS,
PaaS, SaaS, XaaS, SaaS, SecaaS etc
 a complete, specific definition of each service
 the responsibilities of the provider and the consumer
 a set of metrics to measure whether the provider is offering
the services as guaranteed (QoS)
 prohibiting (leaving out) clauses
 an auditing mechanism to monitor the services ; violations
occur or not
 the remedies (payback/ penalty or incentives) available to
the consumer and the provider if the terms are not satisfied.
 how the SLAs will change over time 6
Difference between Web service SLA
and Cloud SLA

7
 Types of SLA
 The present cloud market place features the two types of
SLAs:
 Off-the-shelf SLA or non-negotiable SLA/Direct negotiation:
 Not favorable for mission-critical data or applications.

 Provider creates an unique template and defines all SLA

criteria such as period of contract, billing, and response
time, availability, etc.
 Followed by the present day state-of-the-art clouds
(public cloud).
 Negotiable SLA:
 Negotiation via trusted agent:

 Negotiation via multiple trusted agent:

8
 Types of SLA
 Negotiable SLA: Terms are expensive and has two
categories:

 Negotiation via trusted agent: A trusted external agent

selects the cloud provider and defines the critical
parameters for SLA.

 Negotiation via multiple trusted agent: Efficient when the

cloud consumer requires more than one type of cloud
services.

9
 Service Level Objectives (SLOs)
 SLOs are specific measurable characteristics of the SLA
parameters and represent the level of quality-of-service
which is expected to be provided.
 Include multiple QoS parameters viz. Availability, Service
availability, Penalties ,Response time, Throughput, or
quality.

 Some examples of SLOs are as follows:

 Availability of a service X is 99.9%.

 Response time of a database query Q is between 3 to 5

seconds
 Throughput of a server S at peak load time is 0.875
10
 SMAPPING BETWEEN SLA PARAMETER
AND KPIS

 Each high-level SLA parameter is a function of one or more

key performance indicators (KPIs), which are low-level
resource metrics that are composed, aggregated, or
converted to form high-level SLA parameters.


11
Service Level Management

12
 Considerations for SLA
 As consumers are deciding what terms they need in an
SLA, there are a number of factors they should consider.

 Business Level Objectives: Consumers should know why

they are using cloud computing before they decide how to
use cloud computing.

 Responsibilities of the Provider and Consumer: The

balance of responsibilities between providers and
consumers will vary according to the type of service.

 Business Continuity and Disaster Recovery: Consumers

should ensure their cloud providers have adequate
protection in case of a disaster.
13
 Considerations for SLA
 System Redundancy: Many cloud providers deliver their
services via massively redundant systems. Those systems
are designed so that even if hard drives or network
connections or servers fail, consumers will not experience
any outages.
 Maintenance: maintenance can affect any type of cloud
offering and that it applies to hardware as well as software.

 Location of Data: If a cloud service provider promises to

enforce data location regulations, the consumer must be
able to audit the provider to prove that regulations are being
followed. (data location in pakistan)

 Jurisdiction: Consumers should understand the laws that 14

apply to any cloud providers they consider.
 Considerations for SLA
 Seizure of Data: if law enforcement targets the data and
applications associated with a particular consumer, the
multi-tenant nature of cloud computing makes it likely that
other consumers will be affected. Therefore, Consumers
should consider using a third party to keep back up their
data and applications.

 Failure of the Provider: Consumers should consider the

financial health of their provider and make emergency
plans if the provider were to shut its doors.

 In addition, the provider's policies of handling the

consumer's data and applications if the consumer's account
is delinquent (criminal) or in dispute should be made clear.15
SLA Requirements

16
SLA Requirements

17
key Performance Indicators (KPIs)

18
Industry-defined KPIs

19
 Metrics for Monitoring and Auditing
 The service qualities which the SP guarantees to offer
through the SLA are measured by some metrics based on
which monitoring and auditing of SLAs may be done. These
metrics are known as SLA parameters.
 Throughput: How quickly the service respond.
 Availability: Represented as a % of up time for a service in a
given observed time.
 Reliability: How often the service is available.
 Load Balancing: when elasticity is kicks in (new VMs are
booted or terminated, for example )
 Durability: How likely the data is to be lost.
 Elasticity: The ability for a given resource to grow infinitely,
with limits(the maximum amount of storage or bandwidth, for
example) clearly stated. 20
 Linearity: How a system performs as the load increases.
 Metrics for Monitoring and Auditing
 Agility: How quickly the provider responds as the customer’s
resource load scales up and down
 Automation: what % of requests to the provider are handled
without any human interaction.
 Customer service response times: How quickly the provider
responds to a service request. This refers to the human
interactions required when something goes wrong with the
on-demand, self-service aspects of the cloud.
 Service-level violation rate: Expressed as the mean rate of
SLA violation due to violations of agreed warranty levels
 Transaction time: Time that has elapsed from when a
service is evoked till the completion of the transaction,
including the delays.
 Resolution time: Time period between detection of service21
problem and its resolution.
SLA Requirements w.r.t Cloud Delivery Models

22
Example Cloud SLAs

23
Example Cloud SLAs

24
Example Cloud SLAs

25
SLA Problems

26
SLA Problems

27
 SLA: Example
 Consider a scenario where a company “X” wants to use a cloud
service from a provider “P”. The SLA guarantees are as follows:
 Availability guarantee: 99.95% time over the service period
 Service period: 30 days
 Maximum service hour per day: 12 hours
 Cost: $ 50 per day
 Service credits are awarded to customer if availability
guarantees are not satisfied. Monthly connectivity uptime
service levels are percentage
Monthly uptime given as: Service credit
<99.95% 10%
<99% 25%

 However, in reality it was found that over the service, the cloud
service suffered five outages of durations: 5 hours 30mins,
30mins, 1 hour 30mins, 15mins and 2 hours 25mins, each on
different days, due to which normal service guarantees were
violated. If SLA negotiations are honored, compute the effective
cost payable towards the buying the cloud service. 28
SLA Problems

29
SLA Problems

30
 SLA: Example
 Consider a scenario where a company “X” wants to use a cloud
service from a provider “P”. The SLA guarantees are as follows:
 Availability guarantee: 99.95% time over the service period
 Service period: 30 days
 Maximum service hour per day: 12 hours
 Cost: $ 50 per day
 Service credits are awarded to customer if availability
guarantees are not satisfied. Monthly connectivity uptime
service levels are percentage
Monthly uptime given as: Service credit
<99.95% 10%
<99% 25%

 However, in reality it was found that over the service, the cloud
service suffered five outages of durations: 5 hours 30mins,
30mins, 1 hour 30mins, 15mins and 2 hours 25mins, each on
different days, due to which normal service guarantees were
violated. If SLA negotiations are honored, compute the effective
cost payable towards the buying the cloud service. 31