Microservices in cloud-based infrastructure

the way we see itInfrastructure Services

Introduction
Everything comes in cycles. Back in 2005 we
discussed, defined and developed service-
orientated architecture (SOA) based solutions1
.
Today, when asked what the best-in-class
software blueprint is, we tend to refer to
microservices2
.
Behind both terms sits the same, simple concept
– to develop and design independent, “utility-
based” services that can drive and provide
flexibility, security and agility3
.
We are moving from the 2nd
to the 3rd
platform (the
1st
refers to the mainframe era (monolithic), the
2nd
refers to client-server based computing and
the 3rd
is the web native landscape where users
have access to apps and data from everywhere
at any time with any device). More and more
organisations are moving into a digital landscape
where applications, data and automation are
king.
Introduced 10 years ago, microservices are one
way to design and build applications. Both a
monolithic and/or SOA based approach is and
will still be a valid option. However, the application
of microservices based design is on the increase,
in particular for organizations that make use of
3rd
platform based capabilities. Microservices
demand a new way of constructing Infrastructure
capabilities – Microservices-aware Infrastructure
has to follow a Lego®
-Brick approach. The
resulting infrastructure platform will have to be
fully software based. This implies that control of
the data center is fully automated by software so
that hardware configuration, storage provisioning
and network configuration are managed through
software. This is in contrast to traditional data
centers, where the infrastructure is typically
defined by hardware and devices.
To achieve full digital maturity, and to accelerate
the digital journey, an organisation’s infrastructure
landscape has to conform to a clear set of design
patterns – for example, ease of use, self-service,
agility and flexibility.
However, it can be tricky to navigate around all
the terms involved, so to provide some guidance
on microservices, this document was created
with two main intentions: To provide a detailed
view of the impact microservices is having on
infrastructure; and to examine how today’s
infrastructure should be designed to support
microservices.
Gunnar Menzel
VP, Chief Architect, Infrastructure Services
1 See Capgemini’s SOA (Service Orientated Architecture) point of view papers as well as whitepapers [1,2] were we
outlined the notion of “everything as a service” in 2005.
2 The term “microservice” was discussed at a workshop of software architects near Venice in May, 2011
3 https://en.wikipedia.org/wiki/Microservices
Microservices
demand a new
way of constructing
infrastructure
capabilities –
microservices-aware
infrastructure has to
follow a Lego®
-Brick
approach”
2

Service (Oriented) Architecture
A design pattern mostly related to an enterprise
architecture related approach, which aligns
business, information, information systems and
technical infrastructure
Business Services
A design approach for business operating
models, which combines business processes
and business events, and which has defined
value contracts
Application Services
A design approach to deliver application
functions supporting Business Services,
implemented through a variety of technology
solutions and standards (this is the area
microservices is mostly related to)
Web Services
A special application service implemented
using Web services standards for mass
access, specifically to receive and return
XML documents, for example within a defined
contract; pervasive standards make the
difference
Information Services
A design approach to deliver information to
application services, implemented through a
variety of technology solutions and standards
Infrastructure Services
Specialized or shared infrastructure services,
which support application, Web and information
services
The main intention of a service based approach is to develop or create fine grained
capabilities that can easily be used in different context settings.
Service Terminology
Unravelled
There are many architectural buzzwords, such as service-orientated architecture (SOA),
event-driven architecture, microservices and software defined data centre (SDDC). In
the absence of standard definitions, some common views are that services are business
or technological capabilities, or physical products or an architecture approach. In fact,
a service has many abstraction levels as the term “service” can be used in different
contextual settings:
Business
Applications
Data
Infrastructure
Business Service
Application Service
Data Service
Infrastructure Service
Figure 1 | The different types of services
3

Microservices
Microservices relates to a design pattern that refers to a number of independent
application services delivering one single business capability in an independent, loosely
connected and self-contained fashion.
Following the SOA style (see below), microservices are more autonomous, highly
independent and much smaller or finer grained than their SOA based counterparts.
Microservices architecture can be viewed as a marriage between component-oriented
architecture and service-oriented architecture. Software as a suite is composed of
many small business components with very specific business domain responsibility.
Their interface to the outside world is through an application program interface (API) of
clearly defined services.
One area where microservices departs radically from SOA is in the ownership model for
the services: with microservices, only a single team or person will develop and change
the code for a given service – Netflix is one organisation that is using this approach.
The reason microservices has taken off is wholly down to it being a realisation of many
aspects of Eric Evans’ Domain Driven Design [1] – in that they can be created and
maintained by small, independent teams.
Loosely coupled services can be updated independently of each other; a group of small
services that has to be updated together are not microservices because they are not
loosely coupled. This also applies to sharing a database across services, where updating
a particular service means changing the entire schema (or where a change to a shared
schema results in a change to >1 microservice – an “antipattern”).
<2000
Monolithic Approach / 1st
Platform
A software design pattern that includes all
functional and non-functional features into
one “box.”
ONE SIZE FITS ALL
2000–2010
Service-oriented / 2nd
Platform
ESB
SOA is mainly about “exposing”
discrete components of an application
as web services.
FINE-GRAINED COMPONENTS
>2010
Microservices / 3rd
Ptatform
Independent application services
delivering one single business capability
in an independent, loosely connected
and self-contained fashion.
SUPERFINE
Figure 2 | The three main software patterns
4

Microservices
Best Practice
As outlined before, microservices and SOA have a lot in common. A service oriented
approach is mainly about “exposing” discrete components of an application, as web
services and SOA based applications are comprised of more loosely coupled
components that can be used in different contexts. SOA was a key pattern to move
from the 1st to the 2nd platform: Microservices are key to moving organizations onto
the 3rd platform.
Developing and providing a detailed description of a way to define microservices-
based applications would go far beyond the scope of this document. However, to
understand what characteristics an infrastructure landscape has to comply with in
order to support microservices-based applications, it is important to define
microservices-based design principles. The key design criteria for microservices are
related to the service descriptions.
Following these design principles leads to a microservices-based architectural style;
to develop and implement microservices-based applications, other aspects must be
covered. What is important from an infrastructure perspective is that microservices
based applications demand a “utility-based” approach where infrastructure is “invisible”
to the requesting microservice. Utility-based infrastructure service delivery refers to the
packaging of infrastructure resources, such as computation, storage and data transport,
similarly to metered services like traditional public utilities (such as electricity, water,
natural gas and the telephone network). In other words: a utility infrastructure service is a
metered service supplied through shared infrastructure, in which services are separated,
supporting the above defined design principles.
“Good” microservices:
1. Have fine-grained capabilities that are stateless4
and are “stupid”.
2. Have well-defined interfaces, which hide how the service is executed (i.e. I care
about the interface, not how the service is executed “under the hood”).
3. Use a “very loosely-coupled” approach (one service can be changed without
impacting another).
4. Are “disposable”, and have “no ESB5
– “dumb pipes and smart services”.
5. Are autonomous, explicitly versioned and highly independent.
6. Are stupid – i.e. they “do one thing, and do it well”.
7. Are fully cost and value defined.
4 This is becoming less the case in certain architectures. There was a talk at Strange Loop conference in
2015 around stateful microservices. Statelessness does therefore not determine whether something is a
“microservice”. It is, however, something to be avoided unless absolutely necessary.
5 ESB = Enterprise Service Bus
A utility infrastructure service
is a metered service supplied
through shared infrastructure, in
which services are separated,
supporting the design principles
as defined previously.”
5

Microservices
and Infrastructure
For microservices to work, the underlying infrastructure capability has to support the
microservices design pattern and must focus on all relevant non-functional characteristics
(availability, stability, reliability, performance and security). Microservices will require
“infrastructure” out of the box, which can be constructed in a “Lego®
-based” approach [2].
Infrastructure near capabilities like compute, storage, network, as well as infrastructure
related capabilities like load balancing, replication, etc., must have a “plug and play”
approach and the implementation should be invisible to the consuming microservices.
One key enabler of invisibility is clearly cloud; using a cloud approach to infrastructure
will accelerate the adoption of microservices based applications.
Infrastructure, or more precisely, technical infrastructure, is the combination of all technical
components needed to store, manipulate and transmit data used or consumed by
information systems (applications) in a particular context; with certain performance
characteristics and set against a set of so-called non-functional requirements.
<2000
Monolithic Approach / 1st
Ptatform
ONE SIZE FITS ALL
Storage Storage
Physical
CPU CPU CPU CPU
2000–2010
Service-oriented / 2nd
Platform
ESB
FINE-GRAINED COMPONENTS
Compute Network
Storage
Web Farm
Compute Network
Storage
Database Farm
Physical
>2010
Microservices / 3rd
Ptatform
SUPERFINE
CPU
Network
pool
Discovery Billing Orchestration
CPU Disks Switch
On-premise Off-premise
Storage
pool
Compute
pool
Figure 3 | Software pattern and the related infrastructure blueprint
6

Microservices and Infrastructure
Design Principles
To create “invisibility” for microservices,
infrastructure services have to comply with
particular characteristics.
Looking closer, it means that a microservices
developer must have not only the ability
to create, use, move, expand, contract,
delete and compute, but also have network
and storage capability. This includes all
customer facing infrastructure services.
This is where concepts like convergence,
hyper convergence and fully convergent
infrastructure (or better software bases)
deployed on or off-premise using private,
public or hybrid cloud will be the template
for many organizations to make use of
microservices.
A microservices based infrastructure
platform should be fully software based.
This means that control of the data
center is fully automated by software,
so that hardware configuration, storage
provisioning and network configuration
are managed through software. This is in
contrast to traditional data centers where
the infrastructure is typically defined by
hardware and devices where changes
might reqiure modification executed by
manual (hands-on) activities.
Modular
Service Orientated
Automated Provisioning
Fit for Purpose
Scalable
Sustainable
Microservices-Based
Figure 4 | Infrastructure characteristics to support microservices
Infrastructure
Based on modular and Lego - based building
blocks to reduce complexity and support
multivendor strategy. This also supports the
separation of the concern concept, which is
aimed at avoiding duplicate functionalities.
Preventing complex interfaces and therefore
leading to cost effective operations. An enabler
for invisible infrastructure services to act as an
integrator for best-of-breed infrastructure
services.
Service provisioning and delivery must be
automated to drive cost efficiency in operation.
It supports business users provisioning and
unprovisioning invisible infrastructure services.
Invisible infrastructure services must provide
services and solutions meeting the articulated
requirements, indirectly supporting the overall
mission, but without squandering valuable
resources.
Scalable, seamless (or preferably “step-less” or
“linear”) scalability allows for solutions to grow
or shrink to any required size. This allows
flexible growth, predictable costs, low effort
growth in solutions and fit-for-purpose service
delivery.
Providing and operating sustainable infrastruc-
ture involves the re-thinking of asset purchas-
ing, use and disposition, and goes beyond
Green IT, to full lifecycle analysis.
Globally Operated
Support for all SLAs and KPIs
Secure
Based on Open Standards
Cloud-Data
Centre Interconnection
Private Cloud,
Hybrid Cloud, Public Cloud
Microservices based applications running a 24/7
global organization, and microservices will
deployed in a global context. There is a need to
ensure that the services operate globally, around
the clock.
Services will have to be consumed/used accord-
ing to platinum (99.999%), gold (99.99%), silver
(99.9%) and bronze (99%) service levels. The
infrastructure platform has to cater for all
non-functional requirements.
Any internal and/or external related activity has to
be fully secure – meaning that only authenticated
and authorized services, people or systems are
allowed access.
Invisible infrastructure services are based on
"Open Standards" to facilitate interoperability and
data exchange and to avoid vendor or technolo-
gy lock-in. These are developed (or approved)
and maintained by consensus.
Network connectivity that “follows” all infrastruc-
ture services is critical to deliver all bandwidth
and latency related capabilities.
A cloud-based approach is key for a microservic-
es. Cloud is becoming a ubiquitous delivery
option for all kinds of IT – including microservices
based infrastructure services.
®
7

The Lego®
-Brick Approach
Microservices requires infrastructure
services that are able to comply with
the key service related requirements
as outlined before. This means that
a person (or a microservice) who requires
infrastructure services, should be able
to construct and consume infrastructure
capabilities following the Lego®
-brick
principle by using a “simple” shopping list
and delivering utility-type services.
LEGO POWER
Compute
Storage
Network
Infrastructure
HYBRID
Compute
Storage
Network
Infrastructure
HYBRID
Compute
Storage
Network
Infrastructure
HYBRID
End-to-End Business Process
Input
Functional
KPI, SLA KPI, SLA KPI, SLA
OutputBUSINESS
SERVICE A
APPLICATION A
BUSINESS
SERVICE C
Functional Functional
Output/Input Output/Input
Application/Infrastructure Orchestration
APPLICATION CAPPLICATION B
BUSINESS
SERVICE B
CPU RAM IOPS STORAGE BANDWIDTH
Figure 5 | Utility based Lego® power
8

Infrastructure services should have standard building blocks (such as server, storage
and network), as well as additional components that are constructed from these building
blocks (such as web server, pre-production environment, etc.). Each standard building
block and the additional components should be able to cater for different SLAs and
KPIs without modification.
For instance:
• For a Web server:
–– A simple web server compliant with Bronze SLA or
–– A web server compliant with Platinum SLA
• A data storage solution that:
–– Is optimized for standard database storage or
–– Can provide maximum capacity or
–– Is highly available and highly performing
• A full application environment to:
–– Develop the application further or
–– Perform full-load and performance testing or
–– Run-critical training for end users
ComputeStorageNetwork
Small
01 02
03 04
05 06
Medium
Use Server Building Blocks based
on performance characteristics
Large
Very
Large Platinum Gold Silver Bronze
Platinum Gold Silver Bronze
Platinum Gold Silver Bronze
Tier 4 Tier 3
Use Storage Building Blocks based
Tier 2 Tier 1
Level 4 Level 3
Platinum = 99.999%, Gold = 99.99%, Silver = 99.9%, Bronze = 99%
Use Network Building Blocks based
Level 2 level 1
Small Medium
Web Server
Large
Very
Large
Dev Test
Environments
UAT
Perf
Test
Figure 6 | Infrastructure Lego® based approach
9

Key components of this “shopping list Lego®
-based approach” are the service level
characteristics. Each client environment is different. However, each client will have a
certain set of service levels – from Platinum (99.999%) to Bronze (99%), and of course
a whole set of different non-functional requirements that will shape the underlying
infrastructure. Examples are:
Scalability: Ensure the solution supports current and projected business volumes
Reliability
centering:
Ensure the solution provides an appropriate level of robustness in support
of business processes
Manageability: Ensure the solution can be managed and maintained efficiently and effectively
Availability: Ensure the solution provides the required levels of service
Both the service level, from Platinum to Bronze, and the non-functional requirements can
be translated into very specific infrastructure related requirements, from which standard
Lego®
-based infrastructure components can be defined.
Each pre-assembled infra Lego®
block will be supplied with standard or pre-defined
“behaviors” and each infra Lego®
block will be able to cater for certain availability, stability,
performance and security related requirements – so called service characteristics. As
mentioned above, there are four main categories: platinum, gold, silver and bronze,
which are defined as illustrated below.
Note that this grouping and the detailed business service characteristics are only
examples. Each client environment is different and for some a gold business service
(or IT Application) might have a 6-hour recovery time objective and a 45-minute recovery
point objective.
In the microservices world, infrastructure services have to be consumed in a repeatable,
automatic and simple fashion. Non-functional requirements are a key consideration
Mission Critical
“5 nines”
Downtime pa
5.26 min
Cost
+ + + + +
• 99.999% availability
• 2h RTO, 5min RPO
• 7 days 24 hours
• Two DC implementation
• Hot live standby
• Data restore <=2h
• 1h response for P1 calls
• Highly secure
PLATINUM
Highly Critical
“4 nines”
Downtime pa
52.6 min
Cost
+ + + +
• 4h RTO, 30min RPO
• 7 days 24 hours
• One data center
• Cold standby in
second data center
• Highly secure
GOLD
Critical
“3 nines”
Downtime pa
8.76 h
Cost
+ + +
• 12h RTO, 4h RPO
• std backup
• Extended weekend
ofﬁce hours
• DR – Tape
• Medium secure
SILVER
Non-Critical
“2 nines”
Downtime pa
3.65 days
Cost
+ +
• 99% availability
• 72h RTO, no RPO
• std backup
• Extended
ofﬁce hours
• Low security
BRONZE
Figure 7 | Possible6
service level characteristics
6 Each client environment is different and therefore the detail provided here are only examples
when constructing infrastructure services as these will act as functional infrastructure
requirements to be created, used, moved, expanded, contracted and deleted as needed.
10

Microservices Infrastructure –
The Lego®
Blueprints
As per the Lego®
principle, each construction using infrastructure Lego®
bricks (ILB) will
follow a different blueprint to cater for the different set of non-functional requirements
and behaviors needed. As with real Lego®
bricks, there are endless options to build a
plane, a house or castle – whatever is needed. The only restrictions are the physical
dimensions and characteristics.
Using the Lego®
principle in an infrastructure context, the same principles apply – one
set of ILBs can be used to create one pre-production environment and the same can
be used to create a highly available production environment. The key is to use standard
ILBs and construct the target based on a clear blueprint.
Figure 8 | Lego®’s scalabilty, one size ﬁts all
Using the same sized
bricks, you can build
any model, on any
scale”
11

The Resulting
Conceptual Model
A microservices based infrastructure has to fit into an enterprise-wide architecture model7
that covers all related infrastructure, as outlined in this document.
In particular, aspects, such as segmentation, orchestration, automation, pooling and
security in a cloud context are key conceptual capabilities an organisation should
consider when embarking on using microservices. Such a conceptual model can help
with performing precise product selection as part of transforming from a SOA to a
microservices based approach.
CONTROL LAYER
SoftwareDeﬁnedManagement&Policy
SoftwareDeﬁnedSecurity
ABSTRACTION LAYER
PHYSICAL LAYER
Segmentation Orchestration Automation
Data Center
Resource Pool
Storage
Resource Pool
Virtual
Network Policies
Compute
Resource Pool
Data Center Storage AllocationCompute Network
HypervisionApplicationFacilities ServiceVM
Container
Engine
Figure 9 | Infrastructure Lego® based approach
7 The same applies to SOA; with some key differences
12

Microservices Infrastructure –
Possible Implementation Steps
No client environment will move to use or provide microservices in one step8
– and to
ensure that existing infrastructure services are able to support and provide microservices
based capabilities, the following four steps should be considered:
Shift towards a harmonised,
cost-efﬁcient infrastructure
environment through technology
and process standardization
Standardised and
CONSOLIDATED
Improved efﬁciency and
infrastructure service quality
through process and tool
optimization and automation,
enabling issue prevention and
anticipation of needs
Time
Maturity
Infrastructure services can
be constructed and consumed
in a semi-orchestrated and utility
based approach
Infrastructure services are
fully orchestrated, support
autonomous capabilities and
operate as “Everything”
as a service
Integrated and
AUTOMATED
Service
ORIENTED
Service
ORCHESTRATED
1
2
3
4
Figure 10 | Microservices implementation steps
8 Unless the implementation is “green field” of course
13

Summary
Appendix:
References
Further Reading
Microservices are not always the best option: Monolithic and SOA based solutions still
have (and will continue to have) a place in IT going forward. However, microservices is
the state-of-the-art software pattern, and it is impacting the way we design, build and
operate infrastructure services.
The key to success from an infrastructure perspective is applying a Lego®
-based principle
to create infrastructure services that follow a “plug, play and orchestrate” approach and
where the implementation is invisible to the consuming microservices.
However, the path to a microservices-based infrastructure landscape is not straight
– service complexity, implementation cost, the existing landscape as well as the ever-
changing new environment must be overcome. Successful organizations that are making
use of microservices-based infrastructure follow the four-step implementation plan:
consolidation, automation, orientation and orchestration.
[1] Eric Evans, 2003, Domain-driven Design: Tackling Complexity in the Heart of Software
http://www.amazon.com/Domain-Driven-Design-Tackling-Complexity-Software/
dp/0321125215
[2] Virtual Lego, Technovision 2016, http://www.capgemini.com/blog/cto-blog/2015/11/
technovision-2016-virtual-lego
From Big to Small, Capgemini 2005, http://www.nlondon.bcs.org/pres/am2jan05.pdf
Service-Oriented Enterprise: How To Make Your Business Fast, Flexible and Responsive,
Capgemini 2005, http://www.cio.co.uk/cmsdata/whitepapers/3651/soe_flexibleresponsive.
pdf
The Second Machine age – How DevOps radically shifts organisations to accelerate to
survive: https://www.linkedin.com/pulse/second-machine-age-how-devops-radically-
shift-survive-gunnar-menzel?trk=mp-reader-card
Martin Fowler, http://martinfowler.com/articles/microservices.html
Tom Huston https://smartbear.com/learn/api-design/what-are-microservices/
Andrew Harmel Law http://capgemini.github.io/architecture/microservices-reality-check/
14

For more details contact:
Gunnar Menzel
Infrastructure Services
Vice President, Chief Architect Officer
40 Holborn Viaduct, London EC1N 2PB
Gunnar.Menzel@capgemini.com
About Capgemini
With more than 180,000 people in over 40 countries, Capgemini is one of
the world’s foremost providers of consulting, technology and outsourcing
services. The Group reported 2015 global revenues of EUR 11.9 billion.
Together with its clients, Capgemini creates and delivers business and
technology solutions that fit their needs and drive the results they want.
A deeply multicultural organization, Capgemini has developed its own
way of working, the Collaborative Business ExperienceTM
, and draws on
Rightshore®
, its worldwide delivery model.
Learn more about us at
www.capgemini.com
The information contained in this document is proprietary. ©2016 Capgemini. All rights reserved.
Rightshore®
is a trademark belonging to Capgemini.
EDGE 1602.2016.04

Microservices in cloud-based infrastructure

More Related Content

What's hot (18)

Similar to Microservices in cloud-based infrastructure (20)

Recently uploaded (20)

Microservices in cloud-based infrastructure