Title
stringlengths 18
136
| Content
stringlengths 293
255k
| Category
stringclasses 1
value | Role
stringclasses 1
value | Whitepaper
stringclasses 1
value |
|---|---|---|---|---|
10_Considerations_for_a_Cloud_Procurement
|
Archived10 Considerations for a Cloud Procurement March 2017 This version has been archived For the most recent version of this paper see: https://docsawsamazoncom/whitepapers/latest/considerationsfor cloudprocurement/considerationsforcloudprocurementhtmlArchived© 2017 Amazon Web Services Inc or its affiliates All rights reserved Notices This document is provided for informational purposes only It represents AWS’s current product offerings and practices as of the date of issue of this document which are subject to change without notice Customers are responsible for making their own independent assessment of the information in this document and any use of AWS’s products or services each of which is provided “as is” without warranty of any kind whether express or implied This document does not create any warranties representations contractual commitments conditions or assurances from AWS its affiliates suppliers or licensors The responsibilities and liabilities of AWS to its customers are controlled by AWS agreements and this document is not part of nor does it modify any agreement between AWS and its customers ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 1 Contents Purpose 2 Ten Procurement Considerations 2 1 Understand Why Cloud Computing is Different 2 2 Plan Early To Extract the Full Benefit of the Cloud 3 3 Avoid Overly Prescriptive Requirements 3 4 Separate Cloud Infrastructure (Unmanaged Services) from Managed Services 4 5 Incorporate a Utility Pricing Model 4 6 Leverage ThirdParty Accreditations for Security Privacy and Auditing 5 7 Understand That Security is a Shared Responsibility 6 8 Design and Implement Cloud Data Governance 6 9 Specify Commercial Item Terms 6 10 Define Cloud Evaluation Criteria 7 Conclusion 7 ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 2 Purpose Amazon Web Services (AWS) offers scalable costefficient cloud services that public sector customers can use to meet mandates reduce costs drive efficiencies and accelerate innovation The procurement of an infrastructure as a service (IaaS) cloud is unlike traditional technology purchasing Traditional public sector procurement and contracting approaches that are designed to purchase products such as hardware and related software can be inconsistent with cloud services (like IaaS) A failure to modernize contracting and procurement approaches can reduce the pool of competitors and inhibit customer ability to adopt and leverage cloud technology Ten Procurement Considerations Cloud procurement presents an opportunity to reevaluate existing procurement strategies so you can create a flexible acquisition process that enables your public sector organization to extract the full benefits of the cloud The following procurement considerations are key components that can form the basis of a broader public sector cloud procurement strategy 1 Understand Why Cloud Computing is Different Hyperscale Cloud Service Providers (CSPs) offer commercial cloud services at massive scale and in the same way to all customers Customers tap into standardized commercial services on demand They pay only for what they use The standardized commercial delivery model of cloud computing is fundamentally different from the traditional model for onpremises IT purchases (which has a high degree of customization and might not be a commercial item) Understanding this difference can help you structure a more effective procurement model IaaS cloud services eliminate the customer ’s need to own physical assets There is an ongoing shift away from physical asset ownership toward ondemand utilitystyle infrastructure services Public sector entities should understand how standardized utilitystyle services are budgeted for procured and used and then build a cloud procurement strategy that is ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 3 intentionally different from traditional IT —designed to harness the benefits of the cloud delivery model 2 Plan Early To Extract the Full Benefit of the Cloud A key element of a successful cloud strategy is the involvement of all key stakeholders (procurement legal budget/finance security IT and business leadership) at an early stage This involvement ensures that the stakeholders can understand how cloud adoption will influence existing practices It provides an opportunity to reset expectations for budgeting for IT risk management security controls and compliance Promoting a culture of innovation and educating staff on the benefits of the cloud and how to use cloud technology helps those with institutional knowledge understand the cloud It also helps to accelerate buyin during the cloud adoption journey 3 Avoid Overly Prescriptive Requirements Public sector stakeholders involved in cloud procurements should ask the right questions in order to solicit the best solutions I n a cloud model physical assets are not purchased so traditional data center procurement requirements are no longer relevant Continuing to recycle data center questions will inevitably lead to data center solutions which might result in CSPs being unable to bid or worse lead to poorly designed contracts that hinder public sector customers from leveraging the capabilities and benefits of the cloud Successful cloud procurement strategies focus on applicationlevel performancebased requirements that prioritize workloads and outcomes rather than dictating the underlying methods infrastructure or hardware used to achieve performance requirements Customers can leverage a CSP’s established best practices for data center operations because the CSP has the depth of expertise and experience in offering secure hyperscale Iaa S cloud services It is not necessary to dictate customized specifications for equipment operations and procedures (eg racks server types and distances between data centers) By leveraging commercial cloud industry standards and best practices (including industryrecognized accreditations and certifications) customers avoid placing unnecessary restrictions on the services they can use and ensure access to innovative and costeffective cloud solutions ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 4 4 Separate Cloud Infrastructure (Unmanaged Services) from Managed Services There is a difference between procuring cloud infrastructure (IaaS) and procuring labor to utilize cloud infrastructure or managed services such as Software as a Service (SaaS) cloud Successful cloud procurements separate cloud infrastructure from “hands on keyboard” services and labor or other managed services purchases Cloud infrastructure and services such as labor for planning developing executing and maintaining cloud migrations and workloads can be provided by CSP partners (or other third parties) as one comprehensive solution However cloud infrastructure should be regarded as a separate “service” with distinct roles and responsibilities service level agreements (SLAs) and terms and conditions 5 Incorporate a Utility Pricing Model To realize the benefits of cloud computing you need to think beyond the commonly accepted approach of fixedprice contracting To contract for the cloud in a manner that accounts for fluctuating demand you need a contract that lets you pay for services as they are consumed CSP pricing should be: Offered using a pay asyougo utility model where at the end of each month customers simply pay for their usage Allowed the flexibility to fluctuate based on market pricing so that customers can take advantage of the dynamic and competitive nature of cloud pricing Allowing CSPs to offer pay asyougo pricing or flexible payper use pricing gives customers the opportunity to evaluate what the cost of the usage will be instead of having to guess their future needs and over procure CSPs should provide publicly available up todate pricing and tools that allow customers to evaluate their pricing such as the AWS Simple Monthly Calculator: http://awsamazoncom/calculator Additionally CSPs should provide customers with the tools to generate detailed and customizable billing reports t o meet business and compliance needs ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 5 CSPs should also provide features that enable customers to analyze cloud usage and spending so that customers can build in alerts to notify them when they approach their usage thresholds and projected/budgeted spend Such alerts enable organizations to determine whether to reduce usage to avoid overages or prepare additional funding to cover costs that exceed their projected budget 6 Leverage ThirdParty Accreditations for Security Privacy and Auditing Leveraging industry best practices regarding security privacy and auditing provides assurance that effective physical and logical security controls are in place This prevents overly burdensome processes and duplicative approval workflows that are often unjustified by real risk and compliance needs There are many security frameworks best practices audit standards and standardized controls that cloud solicitations can cite such as the following: Federal Risk and Authorization Management Program (FedRAMP) Service Organization Controls (SOC) 1/American Institute of Certified Public Accountants (AICPA): AT 801 (formerly Statement on Standards for Attestation Engagements [SSAE] No 16)/International Standard on Assurance Engagements (ISAE) 3402 (formerly Statement on Auditing Standards [SAS] No 70) SOC 2 SOC 3 Payment Card Industry Data Security Standard (PCI DSS) International Organization for Standardization (ISO) 27001 ISO 27017 ISO 27108 ISO 9001 Department of Defense (DoD) Security Requirements Guide (SRG) Federal Information Security Management Act (FISMA) International Traffic in Arms Regulations (ITAR) Family Educational Rights and Privacy Act (FERPA) Information Security Registered Assessors Program (IRAP) (Australia) ITGrundschutz (Germany) Federal Information Processing Standard (FIPS) 1402 ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 6 7 Understand That Security is a Shared Responsibility As cloud computing customers are building systems on a cloud infrastructure the security and compliance responsibilities are shared between service providers and cloud consumers In an IaaS model customers control both how they architect and secure their applications and the data they put on the infrastructure CSPs are responsible for providing services through a highly secure and controlled infrastructure and for providing a wide array of additional security features The respective responsibilities of the CSP and the customer depend on the cloud deployment model that is used either IaaS SaaS or Platform as a Service (PaaS)Customers should clearly understand their security responsibilities in each cloud model 8 Design and Implement Cloud Data Governance Organizations should retain full control and ownership over their data and have the ability to choose the geographic locations in which to store their data with CSP identity and access controls available to restrict access to customer infrastructure and data Customers should clearly understand their responsibilities regarding how they store manage protect and encrypt their data A major benefit of cloud computing as compared to traditional IT infrastructure is that customers have the flexibility to avoid traditional vendor lock in Cloud customers are not buying physical assets and CSPs provide the ability to move up and down the IT stack as needed with greater portability and interoperability than the old IT paradigm Public sector entities should require that CSPs: 1) provide access to cloud portability tools and services that enable customers to move data on and off their cloud infrastructure as needed and 2) have no required minimum commitments or required longterm contracts 9 Specify Commercial Item Terms Cloud computing should be purchased as a commercial item and organizations should consider which terms and conditions are appropriate (and not appropriate) in this context A commercial item is recognized as an item that is of a type that has been sold leased licensed or otherwise offered for sale to the general public and generally performs the same for all users/customers both comme rcial and government IaaS CSP terms and conditions are designed to reflect how a cloud services model functions (ie physical assets are not being ArchivedAmazon Web Services – 10 Considerations for a Cloud Procurement Page 7 purchased and CSPs operate at massive scale to offer standardized commercial services) It is critical that a CSP’s terms and conditions are incorporated and utilized to the fullest extent 10 Define Cloud Evaluation Criteria Cloud evaluation criteria should focus on system performance requirements Select the appropriate CSP from an established resource pool to take advantage of the cloud’s elasticity cost efficiencies and rapid scalability This approach ensures that you get the best cloud services to meet your needs the best value in these services and the ability to take advantage of marketdriven innovation The National Institute of Standards and Technology (NIST) definitions of cloud benefits are an excellent starting point to use for determining cloud evaluation criteria: http://nvlpubsnistgov/nistpubs/Legacy/SP/nistspecialpublication800 146pdf Conclusion Thousands of public sector customers use AWS to quickly launch services using an efficient cloudcentric procurement process Keeping these ten steps in mind will help organizations deliver even greater citizen student and mission focused outcomes
|
General
|
consultant
|
Best Practices
|
5_Ways_the_Cloud_Can_Drive_Economic_Development
|
Archived5 Ways the Cloud Can Drive Economic Development August 2018 This paper has been archived For the latest technical content see the AWS Whitepapers & Guides page: https://awsamazoncom/whitepapersArchived © 201 8 Amazon Web Services Inc or its affiliates All rights reserved Notices This document is provided for informational purposes only It represents Amazon Web Services’s (“AWS”) current product offerings and practices as of the date of issue of this document which are subject to change without notice Customers are responsible for making their own independent assessment of the information in this docu ment and any use of AWS’s products or services each of which is provided “as is” without warranty of any kind whether express or implied This document does not create any warranties representations contractual commitments conditions or assurances fro m AWS its affiliates suppliers or licensors The responsibilities and liabilities of AWS to its customers are controlled by AWS agreements and this document is not part of nor does it modify any agreement between AWS and its customers Archived Contents Introduction 1 Sharing More Data and Information 1 Increasing Productivity 3 Preparing Citizens for the Workforce & Building Skills 5 Driving Local Development 6 Allocating Resources More Effectively 8 Key Takeaway 9 Contributors 9 Archived Abstract Government agencies often look to promote new technology for cost savings and efficiency but it does not stop there The second and third tier effects of technology can be long lasting for citizens businesses and economies When public institutions adop t the cloud they experience an internal transformation Inside an organization cloud usage drives greater accessibility of data and information sharing increases worker productivity and improves resource allocation The external benefit of the cloud is recognized through a government ’s ability to put reclaim ed time and resource s toward serving citizens This includes provision ing public services such as occupational skills training quicker and more effective service delivery a pathway to a more productive workforce and ultimately a boost to local development This whitepaper examines the enterprise level benefits of the cloud as well as the residual impact on economic development The US Economic Development Administration defines economic development as “[creating] the conditions for economic growth and improved quality of life by expanding the capacity of individuals firms and communities to maximize the use of their talents and skills to support innovation lower transaction costs and responsibly produce and trade valuable goods and services” We explore this concept through the lens of the cloud ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 1 Introduction Technology empowers governments to improve how and when they reach citizens It improves the quality and accessibility of public service s ultimately creat ing a more productive environment where citizens can thrive Leveraging the cloud is one way governments can accelerate this shift with benefits occurring first inside the institution Sharing More Data and Information One enterprise level benefit of the cloud is its emphasis on data and information sharing The cloud ’s data sharing tools encourage staff to store information in a central location adding visib ility inside the workplace A more collaborative environment can lead to increased communication and idea sharing among agencies and teams that might otherwise op erate in siloes This is true for federal regional and local governments as well as for businesses and entrepreneurs The result is n ear real time access to critical information across an array of industries Examples include data on job creation by location and level retention statistics payroll by industry classification – or North American Industry Classification System code s in the US – in addition to information on health services trade and commerce weather patterns and more Data and IoT solutions can help address development challenges Nexleaf Analytics is one organization harnessing the power of data to tackle global development issues From climate change to public health and food insecurity its mission is to preserve hu man life and protect the planet through sensor technologies and data analytics and by advocating for data driven solutions The organizatio n developed Internet ofThings (IoT) platforms ColdTrace and StoveTrace to help governments ensure the potency of life saving vaccines at the ‘last mile’ and to facilitate the adoption of cleaner cookstoves respectively ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 2 “Data is at the core of creating sustainable change By getting meaningful real time data flowing from the bottom up people have the tools and insights they need to take responsive actions” according to Mar tin Lukac Nexleaf’s CTO and cofounder Nexleaf’s solution powered by A mazon Web Services Inc (AWS) aggregates crucial data that can lead to responsive interventions By collaborating with governments and NGOs in 10 countries across Asia and Africa the organization ensures its solutions adhere to local country laws and preferences and identifies the right tools and analytics to benefit constituents Engaging people on the ground empowers a data driven approach to improving the effic iency of their systems advocating for better resources and tap ping into potential avenues for economic and social development Data drives c ommunity collaboration and innovation The cloud encourages partnerships and collaboration within communities It can lead local governments to facilitate relationships with small and medium sized enterprises (SMEs) which according to an Organisation for Economic Co operati on and Development (OECD ) report “account for over 95% of firms and 60% 70% of employment and generate a large share of new jobs in OECD economies” In Boston Massachusetts the Mayor's Office of New Urban Mechanics took an innovative approach to proble msolving through crowdsourcing Teaming with a technology firm the government sought creative ideas from across Boston to help improve Street Bump its app to collect roadside maintenance and plan long term investments for the city The use of big data and community engagement helped the agency find a creative solution to a public issue Street Bump’s website now reports that te ns of thousand s of bumps have been detected through the app The public private partnership brought automation and speed to an otherwise manual city improvement process and also gave local startups a platform to voice and implement innovative ideas that otherwise may n ot have been discovered Newport Wales is another example of a city optimizing public data in this case to assess environmental conditions It began using IoT sensors to collect ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 3 data such as pollution levels augmenting earlier process es of collecting air samples in glass vials across 85 different location s Together with Pinacl Solutions and Davra Networks Newport is working toward a solution for improving air quality flood control and waste management gleaning timely insights from sensor data via solutions hosted on AWS The effort aimed to boost citizens’ safety and quality of life as part of a vision to improve Newport’s economy The Humanitarian OpenStreetMap Team (HOT) is yet another global organization applying the pri nciples of open source and open data shar ing to humanitarian response and economic development Known for its ability to rapidly coordinate volunteers to map sites impacted by disaster HOT relies on a collaboration with Digi talGlobe Inc for critical satellite imagery data accessible through its Open Data Program and imagery license If not for this partnership HOT would not exist as it is today according to HOT’s Director of Technology Cristiano Giovando Additionally through the AWS Public Datasets Program anyone can analyze data and build comple mentary services using a broad range of compute and data analytics tools The cloud combines fragmented data from a variety of sources improving users’ access and enabling more time for analysis This can facilitate innovation and the possibility of new discover ies Increasing Productivity Consistent r eliability and a lack of physical infrastructure can d rive productivity gains inside and out of a cloud using organization Workforce productivity can improve up to 50% following a large scale AWS migration according to AWS migration experts In addition AWS’s more than 90 solutions offers organizations faster access to services they would otherwise have to build and maintain themselves Government organizations around the world including a road and traffic agency in Belgium and Italy’s public finance regulator have realized increased productivity from the cloud – both for the benefit of their operation s and their citizens ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 4 Productivity gains help institutions better deliver on their mission The Agentschap Wegen & Verkeer (AWV) deploy s new maintenance capabilities up to eight times faster thanks to the automation of services and databases through the AWS Cloud according to Bert Weyne planning & coordination lead at AWV The agency manages 6970 kilometers of roads and 7668 kilometers of cycle lanes in Belgium with its team of 250 road i nspectors having a direct impact on citizen safety In the event of a pothole for example the team uses an app to log information about the issue and prioritize repairs “When we wer e running on in house servers our road inspectors complained about the app’s reliability At times they were unable to access the app and would have to use paper and p en instead It was embarrassing ” says Weyne In addition to bett er performance Weyne’s team has used the cloud to reduce costs speed development and cut infrastructure management time He adds “… by using managed services we’ve slashed system admin time by 67 percent which has improved our agility We can now dev elop and test features three times faster” The cloud has also enabled Italy’s auditing and oversight authority for public accounts and bu dgets to operate more effectively as a remote team Prior to working with AWS Corte dei conti (Cdc) felt constrained by physical IT infrastructure “We wanted to change the way our 3000 plus employees worked enabling them to access applications from anywh ere on any device But we had to ensure that this flexibility for staff didn’ t jeopardize the safety of data ” said C dc’s IT officer Leandro Gelasi This was attainable through a hybrid architecture migration approach and through collaboration with AWS Advanced Consulting Partner XPeppers Srl “As a result [employees are] much more productive Decisions get made faster and the whole system works better It’s a brilliant result fo r our entire organization” said Gelasi As Gelasi and his team prove their ability to fulfill duties securely from any location it may lend an opportunity to employ more workers in small towns and rural locations ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 5 Preparing Citizens for the Workforce & Building Skills Skilldevelopment and education programs offer meaningful contributions to economic development In line with the United Nation s’ 2030 Sustainable Development Goals which includes training and skill building for youth cl oud technology provisions the scaling of educational content and innovative teaching formats to reach learners wherever they are Quality inclusive and relevant education is a key factor in breaking cycles of poverty and reduc ing gender inequalities worldwide By expanding learning beyond the confines of a physical classroom technology helps increase access to courses and level s the playing field for learners of diverse geographical and socio economic backgrounds For schools and educators the cloud offers not only cost savings and agility but also the opportunity to develop breakthroughs in educational models and student engagement Reaching diverse job seekers where ver they are Digital Divide Data (DDD) is a nonprofit social enterprise that uses AWS to support regional workforce development Its goal is to create sustainable tech jobs for youth through Impact Sourcing a model that provides economically marginalized youth with training and jobs in next generation technologies such as cloud computing machine learning cyber security and data analytics In col laboration with Intel AWS worked with DDD to launch the first ofits kind AWS Cloud Academy in Kenya to train certify and employ underserved youth in cloud computing as a stepping stone to more advanced IT careers The program's first cohort included 30 hi gh school graduates from Kibera Nairobi with the second cohort compris ed of 70% women The social enterprise plan s to train five cohorts annually graduating 150200 clo ud engineers p er year – all of whom have the option to work for DDD as cloud computing engineers or to pursue cloud opportunities in the growing local tech sector In terms of workforce benefits DDD and AWS graduates earn five times more than their peers While i nformal workers in Kenya earn an average of $116 USD ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 6 per month AWS graduates earn an average of $575 USD per month The combination of training and w ork experience propels DDD graduates to earn higher income gain economic security and ultimately create better futures for themselves and their families In the US the Loui siana Department of Public Safety and Corrections manages nine state correctional facilities that house 19000 adult prisoners The state run agency offers educational and vocational programs with the goal of helping inmates earn degrees gain job training secure employment and avoid re incarceration The agency sought to implement a new IT environment that would support a better and more reliable online learning solution It also needed effective system security to prevent inmates from accessing the inte rnet amid concerns about victims’ safety and other criminal activity After opting for Amazon WorkSpaces – a managed secure desktop computing service on AWS – the agency along with partner ATLO Software succeeded in launching educational training labs at four Louisiana correctional facilities With the addition of an Amazon Virtual Private Cloud they were operating on a secure network Thanks to onsite labs inmates now have better access to vocational training have the opportunity to earn college credits or degrees and can potentially participate in the labor market Driving Local Development Retaining Local Talent Retaining local talent can be a challenge for cities Moreover a concentration of intellectual capital and innovative businesses and startups can be a strong indicator of economic development Cloud technology can help give new businesses a boost in their forecasting demand generation and innovation when bringing their products or services to market AWS accelerate s this process through AWS Activate a program designed to provide startups with resourc es and credits to get started with the cloud; through access to tools like Amazon LightSail which provides technology like virtual private servers to enterprises of all sizes for the cost of a cup of ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 7 coffee ; and by encouraging public private partnerships and small business linkages namely through the strength of the AWS Partner Network (APN) Additionally AWS CloudStart formed to encourage the growth of SMEs and economic development organizations by providing resources to educate train and help these entities embrace the cost effectiveness of the AWS Cloud “As small businesses leverage a broader portfolio of digital solutions they can see an increase in agility while simultaneously lowering costs and reducing time to innovation” according to Zandile Keebine found er of participating organization GirlCode a nonprofit that aims to empower girls through technology In the US Kansas City Missouri is one example of a city that is successfully using smart technology to attract talent to an emerging business center Along the two mile corridor of the Kansas City Streetcar a $15 million public private partnership supports the deployment of 328 Wi Fi access points and 178 smart streetlights that can detect traffic patterns and open parking spaces It has also funded 25 video kiosks pavement sensors video cameras and other devices all connected by the city’s nearly ubiquitous fiber optic data network The successful use of smart city technology has been a key component in bringi ng people back to Kansas City’s core “Ten years ago we had fewer than 5000 people living downtown” said Bob Bennett Kansans City’s chief innovation officer “We have seen a 520 percent growth in the number of residents in downtown and a 400 percent gr owth in development investment I believe our smart city project has played a prominent role in getting people excited about living here” Entrepreneur ship and p ublicprivate partnerships Cloud technology provides governments with the means to educate and train citizens boosting workforce participation and eligibility Driving local entrepreneurship is an important outgrowth of this investment “A vibrant entrepreneurial sector is essential to small firm development” according to the OECD It adds that regions with “pockets of high entrepreneurial activity” and public private partnerships can lead to more job opportunities and innovation ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 8 A municipality in Sweden is feeling the effects of a strategic partnership aimed at helping small bu sinesses adapt and thrive Consultant CAG Malardalen in Västerås Sweden uses the cloud to help constituents make more data driven decisions deploy resources more efficiently and help shape the economic conditions essential for attract ing new economic activity “[We are] striving to bring the region the latest in cloud technology Our ambition is to always deliver the most relevant IT solutions to our customers Through working with AWS CloudStart our customers benefit from the foundational knowledge we have gathered and we are already seeing a lot of new possibilities for us as a service provider across Sweden” says Tomas Täuber CEO of CAG Malardalen Allocating Resources More Effectively Cloud technology allows governments to rethink critical processes It builds new efficiencies across procurement security compliance and data protection Additionally the cost effectiveness of the cloud enables agencies to redirect resources toward advancing their mission freeing up capacity to create more innovative public services Increased access to new and better citizen services ushers in a higher standard of living offering the potential to draw new inhabitants to a city or region The cloud can act as a catalyst for this type of development driving organizations tow ard increased operational efficiencies and enabling a greater focus on the mission In the Middle East the Kingdom of Bahrain underwent a shift in how it procures resources in its plan to digitize its economy Using the cloud to efficiently deliver ser vices to constituents The Kingdom of Bahrain Information & eGovernment Authority (iGA) is accountable for moving all of its government services online It is responsible for information and communications technolog y (ICT) governance and procurement for the entire Bahraini government The iGA launched a cloud first policy to support its economic development plans ArchivedAmazon Web Services Inc – 5 Ways t he Cloud Can Drive Economi c Development Page 9 Bahrain’s adoption of a cloud first policy boosted efficiency across the public sector and trimmed IT e xpenditures by up to 90% in 2017 according to the Economic Development Board annual report “Through adopting a cloud first policy we have helped reduce the government procurement process for new technology from months to less than two weeks” said Mohammed Ali Al Qaed CEO of Bahrain iGA With cloud based technology as the focus for public ICT procurement the Bahraini government can exercise minimal upfront investment by paying only for the services it needs With tools for cost alloc ation and service provisioning the AWS Cloud offers built in resource discipline enabling governments to shift their focus toward advancing development goals Key Takeaway Technology driven innovation is one way public institutions can drive economic development With the right technology governments nonprof its economic development organizations and other entities can improve their internal operations become more productive and ultimately focus more acutely on serving citizens This can create co nditions in which citizens enjoy improved quality of life and where businesses flourish As organizations increasingly embrace cloud based solutions long lasting effects can be realized in the form of community wide collaboration partnerships with local businesses and increased innovation This can help these institutions wield greater influence on economic development Contributors The following individuals and organizations contributed to this document: • Carina Veksler Public Sector Solutions AWS Public Sector • Randi Larson Public Sector Content AWS Public Sector • John Brennan International Expansion AWS Public Sector • Mike Grella Economic Development AWS Public Policy
|
General
|
consultant
|
Best Practices
|
A_Platform_for_Computing_at_the_Mobile_Edge_Joint_Solution_with_HPE_Saguna_and_AWS
|
ArchivedA Platform for Computing at the Mobile Edge: Joint Solution with HPE Saguna and AWS February 2018 This paper has been archived For the latest technical content see the AWS Whitepapers & Guides page: https://awsamazoncom/whitepapersArchived© 2018 Amazon Web Services Inc or its affiliates All rights reserved Notices This document is provided for informational purposes only It represents AWS’s current product offerings and practices as of the date of issue of this document which are subject to change without notice Customers are responsible for making their own independent assessment of the information in this document and any use of AWS’s products or services each of which is provided “as is” without warranty of any kind whether express or implied This document does not create any warranties representations contractual commitments conditions or assurances from AWS its a ffiliates suppliers or licensors The responsibilities and liabilities of AWS to its customers are controlled by AWS agreements and this document is not part of nor does it modify any agreement between AWS and its customers ArchivedContents Introduction 1 The Business Case for Multi Access Edge Computing 1 MEC Addresses the Need for Localized Cloud Services 2 MEC Leverages the Capabilities Inherent in Mobile Networks 2 MEC Provides a S tandards Based Solution that Enables an Ecosystem of Edge Applications 2 Mobile Edge Solution Overview 4 Example Reference Architectures for Edge Applications 6 Smart City Surveillance 7 AR/VR Edge Applications 10 Connected Vehicle (V2X) 13 Conclusion 15 Contributors 15 Appendix 15 Infrastructure Layer 16 Application Enablement Layer 22 ArchivedAbstract This whitepaper is written for communication service providers with network infrastructure as well as for application developers and technology suppliers who are exploring applications that can benefit from edge computing In this paper we esta blish the value of a standards based computing platform at the mobile network edge describe use cases that are well suited for this platform and present a reference architecture base d on the solutions offered by AWS Saguna and HPE A subset of use cases are reviewed in detail that illustrat e how the reference architecture can be adapted as a platform to serve use case specific requirementsArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 1 Introduction Imagine a world where cars can alert drivers about dangerous road conditions to help them take action to avoid collision and where devices can help fleets of cars drive autonomously and predict traffic patterns Consider a new Industrial Revolution where Internet of Things (IoT) devices or sensors report data collected in real time from large and small machines allowing for intelligent automation and orchestration in industries such as manufacturing agriculture healthcare and logistics Envision city and public services that provide intelligent parking congestion management pollution detection and mitigation emergency response and security While this is happening internet users access bandwidth of 10 times the current maximums and latencies at 1/100 th of current averages using a seamless combination of mobile WiFi and fixed access Fifth generation mobile network (5G) applications are enabling these scenarios by providing 10 ti mes the current bandwidth maximum and 1 This new generation of applications is fueling technological developments and creating new business opportunities for mobile operators One such technological and business development which is key to enabling many new generation of applications is “edge computing ” Edge computing addresses the latency requirements of specialized 5G applications helps manage the potentially exorbitant access cost and network load due to fast growing data demand and support s data localization where necessary By providing a cloud enabled platform for edge computing mobile operators are well positioned to take a leading role in the 5G ecosystem while opening up completely new business cases and revenue streams This whitepaper present s a solution that allow s you to leverage the infrastructure of your existing mobile networks and establis h a platform to enable new revenue generating applications and 5G use case s The Business Case for Multi Access Edge Computing Multi Access Edge Computing (MEC) is a cloud based IT service environment at the edge infrastructure of networks that serves multiple channels of telecommunications access for example mobile wide area networks Wi Fi or LTE based local area networks and wireline ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 2 In this section we discuss the many benefits of a MEC platform that sits at the edge of the cellular mobile network MEC Addresses t he Need for L ocalized Cloud Services Agility scalability ela sticity and cost efficiencies of c loud computing have made it the platform of choice for application development and delivery IoT applications need local cloud services that operat e close to connected devices to improve the economics of telemetry data processing to minimize latency for time critical applications and to ensure that sensitive information is protected locally MEC L everages the Capabilities Inherent in Mobile Network s Mobile networks have expanded to the point where they offer coverage in most countries around the world These networks combine wireless access broadband capacity and security MEC Provides a Standards Based Solution that Enables an Ecos ystem of Edge Applications MEC transforms mobile communication networks into distributed cloud computing platforms that operate at the mobile access network Strategically located in proximity to end users and connected devices MEC enables mobile operators to open their networks to new differentiated services while providing application developers and content providers access to Edge Cloud benefits The ETSI MEC Industry Specification Group (ISG) has defined the first set of standardized APIs and services for MEC The standard is supported by a wide range of industry participants including leading mobile operators and industry vendors Both HP E and Saguna are active members in the ETSI ISG In the following sections we outline the key benefits provided by MEC ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 3 Extremely Low Latency Traditional internet based cloud environments have physical limitations that prohibit you from hosting applications that require extremely low latency Alternatively MEC provides a lowlatency cloud computing env ironment for edge app lications by operating close to end users and connected IoT devices Broadband Delivery Video content is typically delivered using TCP streams When network latency is compounded by congestion users experience annoying delays due to the drop in bitrate The MEC environment provides low latency and minimal jitter which creates a broadband highway for streaming at high bitrates Economical and Scalable In massive IoT uses cases many devices such as sensors or cameras send vast amounts of data upstream which current backhaul networks1 cannot support MEC provides a cloud computing environment at the network edge where IoT data can be aggregated and processed locally thus significantly reducing upstream data MEC infrastructure can scale as you grow by e xpanding local capacity or by deploying additional edge clouds in new locations Privacy and Security By deploying the MEC Edge Cloud locally you can ensure that your private data stays on premise s However unlike server based on premise s installations MEC is a fully automated edge cloud environment with centralized management Role of MEC in 5G MEC enable s ultra low latency use cases specified as part of the 5G network goals MEC also enables fast delivery of data and the connection of billions of devic es while allowing for cost economization related to transporting enormous volumes of data from user devices and IoT over the backhaul network It is important to note that MEC is currently deployed in 4G networks By deploying this standard based technolo gy in existing networks communication service providers can benefit from MEC today while creating an evolutionary path to their next generation 5G network ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 4 Mobile Edge Solution Overview Saguna has developed a MEC virtualized radio access network (vRAN) solution that runs on Hewlett Packard Enterprise (HPE) edge infrastructure This solution lets application developers create mobile edge applications using AWS services while allowing mobi le operators to effectively deploy MEC and operate edge applications within their mobile network Figure 1: End toend MEC solution architecture The proposed m obile edge solution consists of three main layer s as illustrated in Figure 1: • Edge Infrastructure Layer – Based on the powerful x86 compute platform this layer provides compute storage and networking resources at edge locations It supports a wide range of deployment options from RAN base d station sites to backhaul aggregation sit es and regional branch offices • MEC Layer – This layer lets you place an application within a mobile access network and provides a number of services including mobile traffic breakout and steering registration and certification services for application s deployed at the e dge and radio network information services It also provides optional integration point s with mobile core network services such as charging and lawful intercept ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 5 • Application Enablement Layer – This layer provides tools and frameworks to build deploy and maintain edge assisted application s This layer allows you to place certain application modules locally at the edge (eg latency critical or bandwidth hungry components) while keeping other application functions in the cloud The flexible design inherent in the MEC solution architecture allows you to scale the edge component to fit the needs of concrete use cases You can deploy t he edge component at the deepest edge of mobile network (eg colocated with eNodeB equipment at a RAN site) which lets you to deploy lowlatency and bandwidth demanding application components in close proximity to end devices You can also deploy an edge component at any traffic aggregation point between a base station and mobile core which allows you to serve traffic from multiple base stations The proposed m obile edge platform provides a variety of tools to build deploy and manage edge assisted applications such as: • Development libraries and frameworks spanning edge tocloud including function asaservice at the edge and cloud AI frameworks for creating and training models in the cloud seamless deployment and inference at the edge and communication brokerage between edge application services and cloud These development libraries and frameworks expose well defined A PIs and have been widely adopt ed in the developer community shortening the learning curve and accelerating time tomarket for edge assisted applications and use cases • Tools to automate deployment and life cycle management of edge application component s throughout massively distributed edge infrastructure • Infrastructure services such as virtual infrastructure services at the edge traffic steering policies at the edge DNS services radio awareness services integration of edge platform into overall netwo rk function virtualization ( NFV ) framework of mobile operator • Diverse compute resources fitted to the particular needs of edge application such as CPU GPU for acceleration of graphics intensive or AI workloads FPGA accelerators cryptographic and data compression accelerators etc ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 6 This unique combination of functionalities lets you quickly develop edge applications de ploy and manage edge infrastructure and applications at scale and lets you achieve a fast time tomarket with edge enabled use cases Example Reference Architectures for Edge Applications A mobile edge platform enables new app lication behaviors By adding the ability to run certain components and application logic at the mobile network edge in close proximity to the user devices/c lient s the mobile edge platform allows you to reengineer the functional split between c lient and application server s and enables a new generation of application experiences The following list provide s examples of possible mobile edge computing applications in industrial automotiv e public and consumer domains : • Industrial o Next generation augmented reality ( AR) wearables (eg s mart glasses) o IoT for a utomation predictive maintenance o Asset tracking • Automotive o Driverless cars o Connected vehicle tovehicle or vehicle toinfrastructure (V2X ) • Smart Cities o Surveillance cameras o Smart parking o Emergency response managemen t • Consumer Enhanced Mobile Broadband o Next generation Augmented Reality/Virtual Reality ( AR/VR) and video analytics o Social media highbandwidth media sharing ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 7 o Live event streaming o Gaming In the following sections we provide examples of how the mobile edge solution can be implemented for smart city surveillance AR/VR edge applications and Connected V2X Smart City Surveillance Cities can take advantage of IoT technologies to increase the safety security and overall quality of life for residents and keep operational costs down For example video recognition technology enables real time situational analysis (also called “video as a sensor”) which allow s you to detect a variety of objects from video feed (eg people vehicles personal items ) recognize the overall situation (eg a traffic jam fight trespassing and abandoned objects) and classify recognized objects (eg faces license plates) The mobile edge solution enables new abilities in building robust and cost efficient smart city surveill ance systems: • Efficient video processing at the edge – Computer vision systems in general require high quality video input (especially for extracting advanced attributes) and hardware acceleration of inference models The mobile edge solution lets you host a computing environment at the network edge This lets you offload backhaul networks and cloud connectivity from bandwidth hungry high resolution video feeds and allows lowlatency actions based on recognition results (eg opening gates for recognized vehicles or people controlling traffic with adaptive traffic lights) The mobile edge platform provides industry standard GPU resources to accelerate video recognition and any other artificial intelligence ( AI) models deployed at the edge • Flexible access network – End toend smart city surveillance system s might leverage different means to generate video input such as existing fixed surveillance cameras mobile wearable cameras (eg for law enforcement services or first responders) and drone mounted mobile surveillance The diversity of endpoints generating video input requires a high degree of flexibility from access network – leveraging fixed video networks and mobile cellular network s with native mobility support for wearable or unmanned aerial vehicle (UAV )mounted ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 8 cameras Additionally automated drone mounted system s require low latency access to control the flight of the drone which might require endtoend latencies of millisecond scale The mobile edge platform provides a means to use robust lowlatency cellular access with native mobility support for the latter cases and incorporate s existing fixed video networks • Flexible video recognition models – Robust video recognition AI model s usually require extensive training on sample sets of objec ts and events as well as periodic tuning (or development of models for extracting some new attributes) These compute intensive tasks use highly scalable lower cost compute cloud resources However seamless deployment of the trained models to the edge f or execution and managing the life cycle of the deployed models is a complex operational task The mobile edge platform provides seamless development and operational experience starting from creation training and tuning an AI model in the cloud to depl oying it at edge locations and managing the lifecycle of the deployed models The following diagram shows an example architecture of a smart city surveillance edge application: Figure 2: Edge assisted smart city surveillance application A smart city surveillance solution has three main domains: • Field domain – D iverse ecosystem of video producing devices eg body worn cameras from first responder units drones fixed video ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 9 surveillance systems and wireless fixed cameras Video feeds are ingested int o the mobile edge platform via cellular connectivity and use existing video networks • Edge sites – L ocated in close proximity to the video generating devices and host latency sensitive services ( eg UAV flight control local alerts processing) bandwidth hungry compute intensive applications (edge inference) and gateway functionalities for video infrastructure control (camera management) Video services extract target attributes from the video streams and share metadata with local alerting services and cloud services Video services at the edge can also produce low resolution video proxy or sampling video s for transferring only the video s of interest to the cloud • Cloud domain – H osts centralized non latency critical functions such as device and service management functions AAA and policies command and control center functions as well as compute intensive non latency critical tasks of AI model training You can augment a MEC smart city surveillance application with machine learning (ML) and inference models via: • Model training (for surveillance patterns of interest eg facial recognition person counts dwell time analysis heat maps activity detection) using deep learning AMIs on the AWS Cloud • Deployment of trained models to the MEC platform’s application container using AWS Greengrass and Amazon Sage Maker • Application of inference logic (eg alerts or alarms based on select pattern detection) using AWS Greengrass ML inference Figure 3: Detailed view of solution for smart city surveillance application ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 10 This design approach based on the mobile edge platform is a costefficient way of building and operating a s mart city surveillance system with edge processing for bandwidth hungry and laten cysensitive services AR/VR Edge Applications AR/VR is one of the use cases that benefits most from a mobile edge platform AR/VR edge applications can benefit from the m obile edge platform in the following ways: • Next generation AR wearables Current immersive AR experiences require heavy processing on the client side (eg calculating head and eye position and motion information from tracking sensors rendering of high quality 3 D graphics for the AR experience and running video recognition models) The requirement to run heavy computation s on AR device s (eg head mounted display s smart glasses smartphone s) has influenced the characteristics of the se devices —cost size weight battery life and overall aesthetic appeal Figure 4 : Nextgenerat ion AR devices You can avoid b ulkiness cost weight ergonomic and aesthetic limitations on the devices by offloading the heaviest computational tasks from the device s to a remote server or cloud However a truly immersive AR experience requires keeping coherence between AR content and the surrounding physical world with an end toend latency below 10 ms which is unachievable by offload ing to a traditional centralized cloud The m obile edge platform provides compute power at the network edge which allows you to offload latency critical functions from the AR device to the ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 11 network and enables the next generation of lightweight compact devices with long er battery life and native mobility • Mission critical operations AR experiences have been valuable in workforce enablement applications with remote collaboration applications AR assisted maintenance in the industrial space etc In many cases those AR experience s have become an important part of mission critical operations for example ARassisted mainte nance of equipment in hazardous conditions (eg oil extraction sites refineries and mines ) and in ARassisted healthcare Those use cases require high reliability from the AR application even when global connectivity from the c lient to the server side is degrad ed or broken The m obile edge platform provides the capability to re engineer an AR application in a way that the solution can operate offline with critical components deployed both locally in close proximity to devices and globally in the cloud as a fallback option • Localized data processing In many cases AR devices combine data from different local sources ( eg adding live sensor readings from a local piece of equipment to an AR maintenance application) In many cases ingesting data into th e cloud requires high bandwidth and is governed by data security or privacy frameworks A true AR experience requires localized data processing and ingest The m obile edge platform allows you to ingest data from any local source into the AR application as well as execute commands from the AR application to the local data sources (eg perform equipment maintenance tasks) The following diagram shows an example archit ecture for an AR edge application ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 12 Figure 5: Edge assisted AR application The edge assisted AR application has three main domains: • Ultra thin client (eg head mounted display) – G enerates sensor readings of head and eye position location and other relevant data such as live video feed from embedded cameras • Edge services – Part of an AR backend hosted in close proximity to the client on network side These services execute latency critical functions (computing positioning and tracking from AR sensor readings AR graphics rendering) bandwidth hungry functions (eg computer vision models for video recognition) and local data (processing of IoT sensor readings from localized equipment) • Cloud services – Part of AR backend hosted in a traditional centralized cloud These services execute functions centralized in nature (eg authentication and policies command and control center and AR model repository) resource hungry non latency critical functions (computer vision model training) and horizontal cross enterprise functions (eg data lakes integration points with other enterprise systems etc ) This design approach allows client s to offload heavy computations which makes client devices cost efficient lightweight and battery efficient This design also allows local data to be ingested from external sources and contro ls actions to local systems enables offline operation saves costs of WAN connectivity and secures compliance with potential data localization guidelines By working as an integrated part of the mobile network this use case natively supports global mobility telco grade reliability and security ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 13 Connected Vehicle (V2X) Connectivity between vehicles pedestrians roadside infrastructure and other elements in the environment is enabling a tectonic shift in transportation T he full promise of V2X solu tions can only be realized with a new generation of mobile edge applications : • Transportation safety – V2X promises the ability to coordinate actions between vehicles sharing the road (T his ability is sometimes called “Cooperative Cruise Control ”) Informa tion exchange between connected vehicles about intention to change speed or trajectory can significantly improve the safety and robustness of automated or autonomous driving through cooperative maneurvering However due to the very dynamic nature of car traffic these decisions must be made in near real time (with end toend latencies on a millisecond time scale) The m assively distributed nature of road infrastructure near realtime decision making and the requirements for hi ghspeed mobility make the mobile edge platform perfect for host ing the distributed logic of cooperative driving • Transportation efficiency – Cooperative driving promises not only increase d safety o n the road but also a significant boost in transportation efficiency With coordinated vehicle maneuvers the overall capacity of road infrastructure can increase without significant investment in road reconstruction The promise of higher transportation efficiency is further supported by v ehicle toinfrastructure solutions Vehicles can communicate with roadside equipment for speed guidance to coordinate traffic light changes and to reserve parking lots While some information requires only short range communication (eg from a vehicle to a r oadside unit) the coordinated actions of a distributed infrastructure (eg coordinating traffic light changes between multiple intersections) req uires the mobile edge platform to host the logic • Transportation experience – With autonomous driving technologies car infotainment system s are becoming more widespread The mobile edge platform enables the unique possibility of massively distributed content caching with high localization and context awareness as well as the ability to enable location and context based inter actions with vehicle passengers (eg guidance about local ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 14 attractions for travelers time and location limited promotions from local vendors etc) The following diagram shows an example architecture of a V2X edge application Figure 6: Edge assisted connected vehible (V2X) application The V2X solution has three main domains: • Field domain – V ehicle s that generat e data about intended driving maneuvers (eg braking lane change s turn s acceleration) and receive notifications from surroun ding vehicles Road infrastructure that includes all sensors and actuators that are relevant to the driving experience ( eg wind and temperature sensors street lighting connected traffic lights that are controlled via gateway devices such as Road Side Unit) • Edge sites – L ocated in close proximity to the road (eg respective RAN eNodeB sites) and host latency sensitive or highly localized V2X application services Examples of those services include processing and relaying driving maneuver notification s for vehicle coordination processing local sensor readings from road infrastructure dynamic generation of control commands to road infrastructure (eg coordinated traffic lights across several intersections) and caching highly localized infotainment content • Cloud domain – Host s centralized and non latency critical functions such as AAA and policy control historical data collection and ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 15 processing command and control center functions and centralized infotainment content origin With this design approach you can realize low latency and a coordinated exchange of data and control commands between vehicles and surrounding infrastructure This provides a highly specific context for every interaction Conclusion Many technological and market developments are converging to create an opportunity for new applications that take advantage of modern mobile networks and the edge access infrastructure This paper emphasizes the need for an application enablement ecosystem approach and presents a platform to serve multiple edge use cases Contributors The following individuals and organizations contributed to this document: • Shoma Chakravarty WW Technical Leader Telecom Amazon Web Services • Tim Mattison Partner Solution s Architect Amazon Web Services • Alex Rez nik Enterprise Solution Architect and ETSI MEC Chair HPE • Rodion Naurzalin Lead Architect Edge Solutions HPE • Tally Netzer Marketing Leader Saguna • Danny Frydman CTO Saguna Appendix This Appendix gives a more detailed overview of the functional components of the proposed m obile edge platform solution as well as technical characteristics of each component Figure 7 illustrates a functional diagram of the mobile edge platform: ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 16 Figure 7: Mobile edge platform functional diagram Infrastructure Layer The physical infrastructure for a MEC node is based on an edge optimized converged HPE Edgeline EL4000 platform (Figure 8 ) Figure 8: HPE Edgel ine EL4000 chassis and four m710x cartridges The end toend MEC solution gives you the ability to place workloads within any segment of your mobile access network for example at a RAN site backhaul aggregation hub or CRAN hub T he HPE Edge line EL4000 has been optimized for the MEC solution as follows : Compute Density The Edge line EL4000 hosts up to four hot swap SoC cartridges in 1U chassis providing up to 64 Xeon D cores with optimized price/core and watt/core characteristics That design provides 2x – 3x higher compute density compared to a typical traditional data center pl atform while keeping power consumption low These characteristics allow an operator to place a MEC node based on Edge line EL4000 at the deepest edge of access network down to a RAN site ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 17 where space and power constrain ts make other general purpose compute platforms inefficient Workload Specific Compute The diversity of MEC use cases requires that the underlying infrastructure be able to provide different types of compute resources The Edge line EL4000 platform provides diverse compute and hardware acceler ation capabilities which allows you to co locate workloads with different compute needs: • x86 processors that serve general workloads Typical workload example s include a Virtual Network Function virtualized edge application enablement platform and applications that provide fast control actions at the edge for low latency use cases • Builtin GPU that accelerat es graphics processing Typical workload example s are video transcoding at the edge for MEC assisted content distribution and 3 D graphics rendering at the edge for AR/VR streaming application • Plug in dedicated GPU cards that accelerat e deep learning algorithms Enabled by strategic partnership with NVIDIA the Edge line platform can be used for deep learning hardware acceleration at the edge Ty pical workload example s include video analytics and computer vision at the edge and ML inference at the edge for anomaly detection and predictive maintenance • Builtin acceleration of cryptographic operations with QuckAssist Technology (eg accelerating cryptographic or data compression workloads) • Support of up to four PCI E extension slots in a single chassis which provides options for specialized plug in units such as dedicated FPGA boards neuromorphic chips etc Such specialized hardware accelerati on is being evaluated for many network function workloads (such as RAN baseband processing) and applications (efficient deep learning inference) Physical and Operational Characteristics A MEC node should be ready to operate at physical sites and is traditionally used for hosting telco purpose built appliances that are optimized for the physical site environment (eg radio base station equipment at RAN sites ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 18 access routers at traffic hubs etc ) The operational environment of the MEC node sites may be very different from the traditional data center with limited physical space for equipment hosting consumer grade climate control and limited physical accessibility The Edge line EL4000 is optimized to operate in such environments with operational characteristics comparable to the telco purpose built appliances: Parameter RAN Baseband Appliance Typical Data C enter Platform Edge line EL4000 Operating Temperature (oC) +0 …+50 +10 … +35 0 … +55 NonDestructive Shock Tolerance (G) 30 2 30 Expected Mean Time Between Failures ( MTBF ) (years) 3035 1015 >35 On top of enhanced operational characteristics the Edge line EL4000 exposes open iLO interface for the management of highly distributed infrastructure of MEC nodes The iLO interface is compliant with RedFish industry standard It exposes infrastructure management functions via simple RESTful service Saguna OpenRAN C omponents Overview The MEC p latform layer is based on the Saguna OpenRAN solution and consists of the following functions: • Saguna vE dge function located within MEC n ode • Saguna vGate function (optional) located at the core network site • Saguna OMA function (optional) located within a MEC node or at the aggregation point of several MEC n odes Saguna vEdge resides in the MEC node and enables services and applications to operate inside the mobile RAN by providing MEC services such as registration and certification Traffic Offload Function (TOF) real time Radio Network Information Services (RNIS) and optional DNS services The virtualized software node is deployed in the RAN on a server at a RAN site or aggregation point of mobile backhaul traffic It may serve single or multiple ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 19 eNodeB base stations and small cells It can easily be extended to support WiFi and other communications standards in heterogeneous network (HetNet) deployments Saguna vEdge taps the S1 interface (GTP U and S1 AP protocols) and steers the traffic to the appropriate local or remote endpoint based on configured policies Saguna vEdge implements local LTE traffic steering in number of modes (inline steering breakout tap) It has a communication link that connects it to the optional Saguna vGate node using Saguna’s OPTP (Open RAN Transport Protocol) It exposes open REST APIs for managing the platform and providing platform services to the MEC assisted applications Saguna vGate is an optional component that resides in the core network It is responsible for preserving core functionality for RAN generated traffic: l awful interception (LI) charging and policy control The Saguna vGate also enables mobility support for session generated by an MEC assisted application Operating in a v irtual machine Saguna vGate is adjacent to the enhanced packet core (EPC) It has a communication link that connects it to the Saguna vEdge nodes using Saguna’s OPTP (Open RAN Transport Protocol) and m obile network integrations for LI and charging functions Saguna OMA (Open Management and Automation) is an optional subsystem that resid es in the MEC n ode or at the aggregation p oint of several MEC n odes It provides a management layer for the MEC nodes and integrates into the cloud Network Function Virtualization ( NFV ) environment which includes the NFV Orchestrator the Virtual Infrastructure Manager (VIM) and Operations Support Systems (OSS) Saguna OMA provides two management modules: • Virtualized Network Function Manager (VNFM) Provides Life Cycle Management and monitoring for MEC Platform (Saguna vEdge) and MEC assisted applications This is a standard layer of management required within NFV environments It resides at the edge to manage the local MEC environment ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 20 • Mobile Edge Platform Manager (MEPM) – Provides an additional layer of management required for operating and prioritizing MEC applications It is re sponsible for managing the rules and requirements presented by each MEC application rules and resolving conflicts between different MEC assisted applications The Saguna OMA node operates on a virtual machine and manages on boarded MEC assisted application s via its workflow engine using Saguna and third party plugins The Saguna OMA is managed via REST API Saguna OpenRAN Services As a MEC p latform layer Saguna OpenRAN provides the following services: Mobile Network Integration Services • Mobility with Internal Handover support for mobility events between cells connected to the same MEC n ode and External Handover support between two or more MEC n odes and between cells connected to a MEC node and unconnected cells • Lawful Interception (LI) for RAN based generated data It supports X1 (Admin) X2 (IRI) and X3 (CC) interfaces and is pre integrated with Utimaco and Verint LI systems • Charging support using CDR generation for application based charging (based on 3GPP TDF CDR) and charging triggering based on time session and data Supported charging methods are File based (ASN1) and GTP’ • Management vEdge REST API for MEC services discovery and registration MEPM and VNFM let you efficiently operate a MEC solution and integrate it into your existing NFV en vironment Edge Services • Registration for MEC assisted applications The MEC Registration service provides dynamic registration and certification of MEC applications and registration to other MEC services provided by the MEC Platform setting the MEC appli cation type • Traffic Offload Function routes specific traffic flows to the relevant applications as configured by the user The TOF also handles tunneling ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 21 protocols such as GPRS Tunneling Protocol (GTP) for Long Term Evolution (LTE) network Standard A10/A 11 interfaces for 3GPP2 CDMA Network and handles plain IP traffic for WiFi/DSL Network • DNS provides DNS caching service by storing recent DNS addresses locally to accelerate the mobile i nternet and DNS server functionality preconfiguring specific DNS responses for specific domains This lets the User Equipment ( UE) connect to a local application for specific TCP sessions • Radio Network Information Service provided per Cell and per Radio Access Bearer (RAB) The service is vendor independent and can support eNodeBs from multiple RAN vendors simultaneously It supports standard ETSI queries (eg cell info) and notification mechanism (eg RAB establishment events) Additional information based on Saguna proprietary model provides real time feedback on cell congestion level and RAB available throughput using statistical analysis • Instant Messaging with Short Message Service (SMS) provided as a REST API request It offers smart messaging capabilities for example sending SMS to UEs on a specific area ( eg sports stadium) or sending SMS to UE when entering or exiting a specific area (eg shop) Mobile Edge Applications • Throughput guidance application uses the internal RNIS algorithm to deliver throughput guidance for specific IP addresses on the server side or according to domain names of the servers The application can be configured with the period of such Throughput Guidance update per target • DDoS Mitigation application monitors traffic originating from the connected device for specific DDoS attacks on different layers (IP layer for ICMP flooding IP scanning Ping of death; TCP/UDP layer for TCP sync attacks UDP message flooding; Application layer) Devices that are detected as generating DDoS traffic are reported to the network management and traffic from these devices can be locally stopped or the device can be remotely disabled by the network core ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 22 Application Enablement L ayer The Application Enablement layer consists of AWS Greengrass hosted on the MEC node side AWS Greengrass is designed to support IoT solutions that connect different types of devices with the cloud and each other It also runs local functions and parts of applications at the network edge Devices that run Linux and support ARM or x86 architectures can host the AWS Greengr ass Core The AWS Greengrass Core enables the local execution of AWS Lambda code messaging data caching and security Devices running the AWS Greengrass Core act as a hub that can communicate with other devices that have the AWS IoT Device SDK installed such as micro controller based devices or large appliances These AWS Greengrass Core devices and the AWS IoT Device SDK enabled devices can be configured to communicate with one another in a Greengrass Group If the AWS Greengrass Core device loses connection to the cloud devices in the Greengrass Group can continue to communicate with each other over the local network A Greengrass Group represents localized assembly of devices For example it may represent one floor of a building one truck or one home AWS Greengrass builds on AWS IoT and AWS Lambda and it can also access other AWS services It is built for offline operation and greatly simplifies the implementation of local processing Code running in the field can collect filter and aggregate fr eshly collected data and then push it up to the cloud for long term storage and further aggregation Further code running in the field can also take action very quickly even in cases where connectivity to the cloud is temporarily unavailable AWS Greengr ass has two constituent parts : the AWS Greengrass Core and the IoT Device SDK Both of these components run on onpremises hardware out in the field The AWS Greengrass Core is designed to run on devices that have at least 128 MB of memory and an x86 or ARM CPU running at 1 GHz or better and can take advantage of additional resources if available It runs Lambda functions locally interacts with the AWS Cloud manages security and authentication and communicates with the other devices under its purview ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 23 The IoT Device SDK is used to build the applications on devices connected to the AWS Greengrass Core device (generally via a LAN or other local connection) These applications capture data from sensors subscribe to MQTT topics and use AWS IoT device shadows to store and retrieve state information AWS Greengrass features include : • Local support for AWS Lambda – AWS Greengrass includes support for AWS Lambda and AWS IoT d evice shadows With AWS Greengrass you can run AWS Lambda functions right on the device to execute code quickly • Local support for AWS IoT d evice shadows – AWS Greengrass also includes the functionality of AWS IoT d evice shadows The d evice shadow caches the state of your device like a vi rtual version or “shadow” and tracks the device’s current versus desired state • Local messaging and protocol adapters – AWS Greengrass enables messaging between devices on a local network so they can communicate with each other even when there is no connection to AWS With AWS Greengrass devices can process messages and deliver them to other device s or to AWS IoT based on business rules that the user defines Devices that communicate via the popular industrial protocol OPC UA are supported by the AWS Gr eengrass protocol adapter framework and the out ofthebox OPC UA protocol module Additionally AWS Greengrass provides protocol adapter framework to implement support for custom legacy and proprietary protocols • Local resource access – AWS Lambda functions deployed on an AWS Greengrass Core can access local resources that are attached to the device This allows you to use serial ports USB peripherals such as add on security devices sensors and actuators on board GPUs or the local file system to quickly access and process local data • Local machine learning i nference – A llows you to locally run a n MLmodel that’s built and trained in the cloud With hardware acceleration available in the MEC infrastructure layer this feature provides a powerful mec hanism for solving any machine learning task at the local edge eg discovering patterns in data building computer vision systems and running anomaly detection and predictive maintenance algorithms ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 24 AWS Greengrass has a growing list of features Curren t features are shown in Figure 9 Figure 9: AWS Greengrass features AWS Greengrass on the MEC node acts as a pivot point It integrates the MEC platform with the AWS I oT solution and other AWS services providing a powerful application enablement environment for developing deploying and managing MEC assisted applications at scale The figure below illustrates the current portfolio of AWS services that enable a seamless IoT pipeline —from endpoints connecting via Amazon FreeRTOS or the IoT SDK through MQTT or OPC UA to edge gateways that host AWS Greengrass and Lambda functions providing data processing capabilities at the edge up to cloud hosted AWS IoT Core AWS Device Management AWS Device Defender and AWS IoT Analytics services as well as enterprise applications ArchivedAmazon Web Services – A Platform for Computing at the Mobile Edge Page 25 Figure 10: AWS services that enable a seamless IoT pipeline 1 In a telecommunications network the backhaul portion of the network comprises the intermediate links between the core network or backbone network an d the small subnetworks at the "edge"
|
General
|
consultant
|
Best Practices
|
A_Practical_Guide_to_Cloud_Migration_Migrating_Services_to_AWS
|
Archived A Practical Gui de to Cl oud Migration Migratin g Service s to AWS December 2015 This paper has been archived For the latest technical content see: https://docsawsamazoncom/prescriptiveguidance/latest/mrpsolution/mrpsolutionpdfArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 2 of 13 © 2015 Amazon Web Services Inc or its affiliates All rights reserved Notices This document is provided for informational purposes only It represents AWS’s current product offerings and practices as of the date of issue of this document which are subject to change without notice C ustomers are responsible for making their own independent assessment of the information in this document and any use of AWS’s products or services each of which is provided “as is” without warranty of any kind whether express or implied This document do es not create any warranties representations contractual commitments conditions or assurances from AWS its affiliates suppliers or licensors The responsibilities and liabilities of AWS to its customers are controlled by AWS agreements and this docum ent is not part of nor does it modify any agreement between AWS and its customers ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 3 of 13 Contents Abstract 3 Introduction 4 AWS Cloud Adoption Framework 4 Manageable Areas of Focus 4 Successful Migrations 5 Breaking Down the Economics 6 Understand OnPremises Costs 6 Migration Cost Considerations 8 Migration Options 10 Conclusion 12 Further Reading 13 Contributors 13 Abstract To achieve full benefits of moving applications to the Amazon Web Services (AWS) platform it is critical to design a cloud migration model that delivers optimal cost efficiency This includes establishing a compelling business case acquiring new skills within the IT organization implemen ting new business processes and defining the application migration methodology to transform your business model from a traditional on premises computing platform to a cloud infrastructure ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 4 of 13 Perspective Areas of Focus Introduction Cloudbased computing introduces a radical shift in how technology is obtained used and managed as well as how organizations budget and pay for technology services With the AWS cloud platform project teams can easily configure the virtual network using t heir AWS account to launch new computing environments in a matter of minutes Organizations can optimize spending with the ability to quickly reconfigure the computing environment to adapt to changing business requirements Capacity can be automatically sc aled —up or down —to meet fluctuating usage patterns Services can be temporarily taken offline or shut down permanently as business demands dictate In addition with pay peruse billing AWS services become an operational expense rather than a capital expense AWS Cloud Adoption Framework Each organization will experience a unique cloud adoption journey but benefit from a structured framework that guides them through the process of transforming their people processes and technology The AWS Cloud Adopt ion Framework (AWS CAF) offers structure to help organizations develop an efficient and effective plan for their cloud adoption journey Guidance and best practices prescribed within the framework can help you build a comprehensive approach to cloud comput ing across your organization throughout your IT lifecycle Manageable Areas of Focus The AWS CAF breaks down the complicated planning process into manageable areas of focus Perspectives represent top level areas of focus spanning people process and te chnology Components identify specific aspects within each Perspective that require attention while Activities provide prescriptive guidance to help build actionable plans The AWS Cloud Adoption Framework is flexible and adaptable allowing organizations to use Perspectives Components and Activities as building blocks for their unique journey Business Perspective Focuses on identifying measuring and creating business value using technology services The Components and Activities within the Business Perspective can help you develop a business case for cloud align ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 5 of 13 business and technology strategy and support stakeholder engagement Platform Perspective Focuses on describing the structure and relationship of technology elements and services in complex IT environments Components and Activities within the Perspective can help you develop conceptual and functional models of your IT environment Maturity Perspective Focuses on defining the target state of an organization's capabilities measuring maturity and optimizing resources Components within Maturity Perspective can help assess the organization's maturity level develop a heat map to prioritize initiatives and sequence initiatives to develop the roadm ap for execution People Perspective Focuses on organizational capacity capability and change management functions required to implement change throughout the organization Components and Activities in the Perspective assist with defining capability and skill requirements assessing current organizational state acquiring necessary skills and organizational re alignment Process Perspective Focuses on managing portfolios programs and proj ects to deliver expected business outcome on time and within budget while keeping risks at acceptable levels Operations Perspective Focuses on enabling the ongoing operation of IT environments Components and Activities guide operating procedures service management change management and recovery Security Perspective Focuse s on helping organizations achieve risk management and compliance goals with guidance enabling rigorous methods to describe structure of security and compliance processes systems and personnel Components and Activities assist with assessment control selection and compliance validation with DevSecOps principles and automation Successful Migrations The path to the cloud is a journey to business results AWS has helped hundreds of customers achieve their business goals at every stage of their journey While every organization’s path will be unique there are common patterns approaches and best pract ices that can be implemented to streamline the process 1 Define your approach to cloud computing from business case to strategy to change management to technology 2 Build a solid foundation for your enterprise workloads on AWS by assessing and validating yo ur application portfolio and integrating your unique IT environment with solutions based on AWS cloud services Perspective Areas of Focus ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 6 of 13 3 Design and optimize your business applications to be cloud aware taking direct advantage of the benefits of AWS services 4 Meet your internal and external compliance requirements by developing and implementing automated security policies and controls based on proven validated designs Early planning communication and buy in are essential Understanding the forcing function (tim e cost availability etc) is key and will be different for each organization When defining the migration model organizations must have a clear strategy map out a realistic project timeline and limit the number of variables and dependencies for trans itioning on premises applications to the cloud Throughout the project build momentum with key constituents with regular meetings and reporting to review progress and status of the migration project to keep people enthused while also setting realistic ex pectations about the availability timeframe Breaking Down the Economics Understand On Premises Costs Having a clear understanding of your current costs is an important first step of your journey This provides the baseline for defining the migration model that delivers optimal cost efficiency Onpremises data centers have costs associated with the servers storage networking power cooling physical space and IT labor required to support applications and services running in the production environment Although many of these costs will be eliminated or reduced after applications and infrastructure are moved to the AWS platform knowing your current run rate will help determine which applications are good candidates to move to AWS which applications need to be rewrit ten to benefit from cloud efficiencies and which applications should be retired The following questions should be evaluated when calculating the cost of on premises computing: Understanding Costs To build a migration model for optimal efficiency it is important to accurately understand the current costs of running onpremises applications as well as the interim costs incurred during the transition ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 7 of 13 “Georgetown’s modernization strategy is not just about upgrading old systems; it is about changing the way we do business building new partnerships with the community and working to embrace innovation Cloud has been an important component of this Although we thought the primary driver would be cost savings we have found that agility innovation and the opportuni ty to change paths is where the true value of the cloud has impacted our environment “Traditional IT models with heavy customization and sunk costs in capital infrastructures —where 90% of spend is just to keep the trains running —does not give you the opp ortunity to keep up and grow” Beth Ann Bergsmark Interim Deputy CIO and AVP Chief Enterprise Architect Georgetown University Labor How much do you spend on maintaining your environment (broken disks patching hosts servers going offline etc)? Network How much bandwidth do you need? What is your bandwidth peak to average ratio? What are you assuming for network gear? What if you need to scale beyond a single rack? Capacity What is the cost of over provisioning for peak capacity? How do you plan for capacity? How much buffer capacity are you planning on carrying? If small what is your plan if you need to add more? What if you need less capacity? What is your plan to be abl e to scale down costs? How many servers have you added in the past year? Anticipating next year? Availability / Power Do you have a disaster recovery (DR) facility? What was your power utility bill for your data center(s) last year? Have you budgeted for both average and peak power requirements? Do you have separate costs for cooling/ HVAC? Are you accounting for 2N power? If not what happens when you have a power issue to your rack? Servers What is your average server utilization? How much do you overpr ovision for peak load? What is the cost of over provisioning? Space Will you run out of data center space? When is your lease up? ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 8 of 13 Migration Cost Considerations To achieve the maximum benefits of adopting the AWS cloud platform new work pract ices that drive efficiency and agility will need to be implemented: IT staff will need to acquire new skills New business processes will need to be defined Existing business processes will need to be modified Migration Bubble AWS uses the term “migration bubble” to describe the time and cost of moving applications and infrastructure from on premises data centers to the AWS platform Although the cloud can provide significant savings costs may increase as you move into the migration bubble It i s important to plan the migration to coincide with hardware retirement license and maintenance expiration and other opportunities to reduce cost The savings and cost avoidance associated with a full all in migration to AWS will allow you to fund the mig ration bubble and even shorten the duration by applying more resources when appropriate Time Figure 1: Migration Bubble Level of Effort The cost of migration has many levers that can be pulled in order to speed up or slow down the process including labor process tooling consulting and technology Each of these has a corresponding cost associated with it based on the level of effort required to move the application to the AWS platform Migration Bubble Planning • • • • • • Planning and Assessment Duplicate Environments Staff Training Migration Consulting 3rd Party Tooling Lease Penalties Operation and Optimization Cost of Migration $ ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 9 of 13 To calculate a realistic total cost of ownership (TCO) you need to understand what these costs are and plan for them Cost considerations include items such as: Labor During the transition existing staff will need to continue to maintain the production environment learn new skills and decommission the old infrastructure once the migration is complete Additional labor costs in the migration bubble include: Staff time to plan and assess project scope and project plan to migrate applications and infrastructure Retaining consulting partners with the expertise to streamline migration of applications and infrastructure as well as training staff with new skills Due to the general lack of cloud experience for most organization s it is necessary to bring in outside consulting support to help guide the process Process Penalty fees associated with early termination of contracts may be incurred (facilities software licenses etc) once applications or infrastructure are decommissioned The cost of tooling to automate the migration of data and virtual machines from on premises to AWS Technology Duplicate environments will be required to keep production applications/infrastructure available while transitioning to the AWS platform Cost considerations include: Cost to maintain production environment during migration Cost of AWS platform comp onents to run new cloud based applications Licensing of automated migration tools license to accelerate the migration process ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 10 of 13 “I wanted to move to a model where we can deliver more to our citizens and r educe the cost of delivering those services to them I wanted a product line that has the ability to scale and grow with my department AWS was an easy fit for us and the way we do business” Chris Chiancone CIO City of McKinney City of McKinney City of McKinney Texas Turns to AWS to Deliver More Advanced Services for Less Money The City of McKinney Texas about 15 miles north of Dallas and home to 155000 people was ranked the No 1 Best Place to live in 2014 by Money Magazine The city’s IT department is going all in on AWS and uses the platform to run a wide range of services and applications such as its land management and records management systems By using AWS the city’s IT department can focus on delivering new and better services for its fast growing population and city employees instead of spending resources buying and maintaining IT infrastructure City of McKinney chose AWS for our ability to scale and grow with the needs of the city’s IT department AWS provides an easy fit for the way the city does business Without having to own the infrastructure the C ity of McKinney has the ability to use cloud resources to address business needs By moving from a CapEx to an OpEx model they can now return funds to critical city projects Migration Options Once y ou understand the current costs of an on premises production system the next step is to identify applications that will benefit from cloud cost and efficiencies Applications are either critical or strategic If they do not fit into either category they should be taken off the priority list Instead categorize these as legacy applications and determine if they need to be replaced or in some cases eliminated Figure 2 illustrates decision points that should be considered in ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 11 of 13 “A university is really a small city with departments running about 1000 diverse small services across at the university We made the decision to go down the cloud journey and have been working with AWS for the past 4 years In building our business case we wanted the ability to give our customers flexible IT services th at were cost neutral “We embraced a cloud first strategy with all new services a built in the cloud In parallel we are migrating legacy services to the AWS platform with the goal of moving 80% of these applications by the end of 2017” Mike Chapple P hD Senior Director IT Services Delivery University of Notre Dame selecting applications to move to the AWS platform focusing on the “6 Rs” — retire retain re host re platform re purchase and re factor Decommission Refactor for AWS Rebuild Application Architecture AWS VM Import Org/Ops Change Do Not Move Move the App Infrastructure Design Build AWS Lift and Shift (Minimal Change) Determine Migration 3rd Party Tools Impact Analysis Management Plan Identify Environment Process Manually Move App and Data Ops Changes Migration and UAT Testing Signoff Operate Discover Assess (Enterprise Architecture and Determine Migration Path Application Lift and Shift Determine Migration Process Plan Migration and Sequencing 3rd Party Migration Tool Tuning Cutover Applications) Vendor S/PaaS (if available) Move the Application Refactor for AWS Recode App Components Manually Move App and Data Architect AWS Environment Replatform (typically legacy applications) Rearchitect Application Recode Application and Deploy App Migrate Data Figure 2: Migration Options Applications that deliver increased ROI through reduced operation costs or deliver increased business results should be at the top of the priority list Then you can determine the best migration path for each workload to optimize cost in the migration process ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 12 of 13 Conclusion Many organizations are extending or moving their business applications to AWS to simplify infrastructure management deploy quicker provide greater availability increase agility allow for faster innovation and lower cost Having a clear understanding of existing infrastructure costs the components of your migration bubble and their corresponding costs and projected savings will help you calculate payback time and projected ROI With a long history in enabling enterprises to successfully adopt cloud computing Amazon Web Services delivers a mature set of services specifically designed for the unique security compliance privacy and governance requirements of large organizations With a technology platform that is both broad and deep Professional Services and Support organizations robust training programs and an ecosystem tens ofthousands strong AWS can help you move faster and do more With AWS you can: Take advantage of more services storage options and security controls than any other cloud platform Deliver on stringent standards with the broadest set of certifications accreditations and controls in the industry Get deep assistance with our global cloud focused enterprise professional services support and training teams ArchivedAmazon Web Services – A Practical Guide to Cloud Migration December 2015 Page 13 of 13 Further Reading For additional help please consult the following sources: The AWS Cloud Adoption Framework http://d0awsstaticcom/whitepapers/aws_cloud_adoption_frameworkp df Contributors The following individuals and organizations contributed to this document: Blake Chism Practice Manager AWS Public Sector Sales Var Carina Veksler Public Sector Solutions AWS Public Sector Sales Var
|
General
|
consultant
|
Best Practices
|
Active_Directory_Domain_Services_on_AWS
|
This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlActive Di rectory Domain Services on AWS Design and Planning Guide November 20 2020 This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlNotices Customers are responsible for making their own independent assessment of the information in this document This document: (a) is for informational purposes only (b) represents current AWS product offerings and practices which are subject to change withou t notice and (c) does not create any commitments or assurances from AWS and its affiliates suppliers or licensors AWS products or services are provided “as is” without warranties representations or conditions of any kind whether express or implied The responsibilities and liabilities of AWS to its customers are controlled by AWS agreements and this document is not part of nor does it modify any agreement between AWS and its customers © 20 20 Amazon Web Services Inc or its affiliates All rights reserved This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlContents Importance of Active Directory in the cloud 1 Terminology and definitions 1 Shared responsibility model 3 Direct ory services options in AWS 4 AD Connector 4 AWS Managed Microsoft Active Directory 5 Active Directory on EC2 7 Comparison of Active Directory Services on AWS 7 Core infrastructure design on AWS for Windows Workloads and Directory Services 9 Planning AWS accounts and Organization 9 Network design considerations for AWS Managed Microsoft AD 9 Design consideration for AWS Managed Micro soft Active Directory 12 Single account AWS Region and VPC 12 Multiple accounts and VPCs in one AWS Region 13 Multiple AWS Regions deploymen t 14 Enable Multi Factor Authentication for AWS Managed Microsoft AD 16 Active Directory permissions delegation 17 Design considerations for running Active Directory on EC2 instances 18 Single Region deployment 18 Multi region/global deployment of self managed AD 20 Designing Active Directory sites and services topology 21 Security considerations 22 Trust relationships with on premises Active Directory 22 Multi factor authentication 24 AWS account security 24 Domain controller security 24 This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlOther considerations 25 Conclusion 26 Contributors 26 Further Reading 27 Document Revisions 27 This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAbstract Cloud is now the center of most enterprise IT strategies Many enterprises find that a wellplanned move to the cloud results in an immediate business payoff Active Directory is a foundation of the IT infrastructure for many large enterprises This whitepaper covers best practices for designing Active Directory Domain Services (AD DS) architecture in Amazon Web Services (AWS) including AWS Managed Microsoft AD Active Directo ry on Amazon Elastic Compute Cloud (Amazon EC2) instances and hybrid scenarios This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 1 Importance of Active Directory in the cloud Microsoft Active Directory was introduced in 1999 and became de facto standard technology for centralized management of Microsoft Windows computers and user authentications Active Directory serves as a distributed hierarchical data storage for information about corporate IT infrastructure including Domain Name System (DNS) zones and records devices and users user credentials and access rights based on groups membership Currently 95% of enterprises use Active Directory for authentication Successful adoption of cloud technology requires considering existing IT infr astructure and applications deployed on premises Reliable and secure Active Directory architecture is a critical IT infrastructure foundation for companies running Windows workloads Terminology and definitions AWS Managed Microsoft Active Directory AWS Directory Service for Microsoft Active Directory also known as AWS Managed Microsoft AD is Microsoft Windows Server Active Directory Domain Services (AD DS) deployed and managed by AWS for you The service runs on actual Windows Server for the highest po ssible fidelity and provides the most complete implementation of AD DS functionality of cloud managed AD DS services available today Active Directory Connector (AD Connector) is a directory gateway (proxy) that redirects directory requests from AWS applic ations and services to existing Microsoft Active Directory without caching any information in the cloud It does not require any trusts or synchronization of user accounts Active Directory Trust A trust relationship (also called a trust) is a logical rel ationship established between domains to allow authentication and authorization to shared resources The authentication process verifies the identity of the user The authorization process determines what the user is permitted to do on a computer system or network Active Directory Sites and Services In Active Directory a site represents a physical or logical entity that is defined on the domain controller Each site is associated with an Active Directory domain Each site also has IP definitions for what IP addresses and ranges belong to that site Domain controllers use site information to inform Active Directory clients about domain controllers present within the closest site to the client This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 2 Amazon V irtual Private Cloud ( Amazon VPC) lets you provision a logically isolated section of the AWS Cloud where you can launch AWS resources in a virtual network that you define You have complete control over your virtual networking environment including the selection of your own private IP address ranges creation of subnets and configuration of route tables and network gateways You can also create a hardware Virtual Private Network (VPN) connection between your corporate data center and your VPC to leverage the AWS Cloud as an extension of your corporate data ce nter AWS Direct Connect is a cloud service solution that makes it easy to establish a dedicated network connection from your premises to AWS Using AWS Direct Connect you can establish private connectivity between AWS and your data center office or colocation environment AWS Single Sign On (AWS SSO) is a cloud SSO service that makes it easy to centrally manage SSO access to multiple AWS accounts and business applications With AWS SSO you can easily manage SSO access and user permissions to all of your accounts in AWS Organi zations centrally AWS Transit Gateway is a service that enables customers to connect their VPCs and their on premises networks to a single gateway Domain controller (DC) – an Active Directory server that responds to authentication requests and store a re plica of Active Directory database Flexible Single Master Operation (FSMO) roles In Active Directory some critical updates are performed by a designated domain controller with a specific role and then replicated to all other DCs Active Directory uses r oles that are assigned to DCs for these special tasks Refer to the Microsoft documentation web site for more information on FSMO roles Global Catalog A glob al catalog server is a domain controller that stores partial copies of all Active Directory objects in the forest It stores a complete copy of all objects in the directory of your domain and a partial copy of all objects of all other forest domains Read Only Domain Controller (RODC ) Read only domain controllers (RODCs) hold a copy of the AD DS database and respond to authentication requests but applications or other servers cannot write to them RODCs are typically deployed in locations where physical s ecurity cannot be provided VPC Peering A VPC peering connection is a networking connection between two VPCs that enables you to route traffic between them using private IPv4 or IPv6 This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 3 addresses Instances in either VPC can communicate with each other as if they are within the same network Shared responsibility model When operating in the AWS Cloud Security and Compliance is a shared responsibility between AWS and the custome r AWS is responsible for security “of” the cloud whereas customers are responsible for security “in” the cloud Figure 1 Shared Responsibility Model when operating in AWS Cloud AWS is responsible for securing its software hardware and the facilities where AWS services are located including securing its computing storage networking and database services In addition A WS is responsible for the security configuration of AWS Managed Services like Amazon DynamoDB Amazon Relational Database Service (Amazon RDS) Amazon Redshift Amazon EMR Amazon WorkSpaces and so on Customers are responsible for implementing appropria te access control policies using AWS Identity and Access Management ( IAM) configuring AWS Security Groups (Firewall) to prevent unauthorized access to ports and enabling AWS CloudTrail Customers are also responsible for enforcing appropriate data loss p revention policies to ensure compliance with internal and external policies as well as detecting and This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 4 remediating threats arising from stolen account credentials or malicious or accidental misuse of AWS If you decide to run your own Active Directory on Am azon EC2 instances you have full administrative control of the operating system and the A ctive Directory environment You can set up custom configurations and create a complex hybrid deployment topology However you must operate and support it in the sam e manner as you do with onpremises Active Directory If you use AWS Managed Microsoft AD AWS provides instance deployment in one or multiple regions operational management of your directory monitoring backup patching and recovery services You confi gure the service and perform administrative management of users groups computers and policies AWS Managed Microsoft AD has been audited and approved for use in deployments that require Federal Risk and Authorization Management (FedRAMP) Payment Card Industry Data Security Standard (PCI DSS) US Health Insurance Portability and Accountability Act (HIPAA) or Service Organizational Control (SOC) compliance When used with compliance requirements it is your responsibility to configure the directory password policies and ensure that the entire application and infrastructure deployment meets your compliance requirements For more information see Manag e Compliance for AWS Managed Microsoft AD Directory services options in AWS AWS provides a comprehensive set of services and tools for deploying Microsoft Windows workloads on its rel iable and secure cloud infrastructure AWS Active Directory Connector (AD Connector) and AWS Managed Microsoft AD are fully managed services that allow you to connect AWS applications to an existing Active Directory or host a new Active Directory in the cl oud Together with the ability to deploy selfmanaged Active Directory in Amazon EC2 instances these services cover all cloud and hybrid scenarios for enterprise identity services AD Connector AD Connector can be used in the following scenarios: • Sign in to AWS applications such as Amazon Chime Amazon WorkDocs Amazon WorkMail or Amazon WorkSpaces using corporate credentials (See the list of compatible applications on the AWS Documentation site) This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 5 • Enable Access to the AWS Management Console with AD Crede ntials For large enterprises AWS recommends us ing AWS Single Sign On • Enable multi factor authentication by integrating with your existing RADIUS based MFA infrastructure • Join Windows EC2 instances to your on premises Active Directory Note: Amazon RDS for SQL Server and Amazon FSx for Windows File Server are not compatible with AD Connector Amazon RDS for SQL Server compatible with AWS Managed Microsoft AD only Amazon FSx for Windows File Server can be deployed with AWS Managed Microsoft AD or self managed Active Directory AWS Managed Microsoft Active Directory AWS Directory Service lets you run Microsoft Active Directory as a managed service By default each AWS Managed Microsoft AD has a minimum of two domain controllers each deployed in a separate Availability Zone (AZ) for resiliency and fault tolerance All domain controllers are exclusively yours with nothing shared with any oth er AWS customer AWS provides operational management to monitor update backup and recover domain controller instances You administer users groups computer and group policies using standard Active Directory tools from a Windows computer joined to the AWS Managed Microsoft AD domain AWS Managed Microsoft AD preserves the Windows single sign on (SSO) experience for users who access AD DS integrated applications in a hybrid IT environment With AD DS trust support your users can sign in once on premises and access Windows workloads runnin g onpremises and in the cloud You can optionally expand the scale of the directory by adding domain controllers thereby enabling you to distribute requests to meet your performance requirements You can also share the directory with any account and VPC Multi Region replication can be used to automatically replicate your AWS Managed Microsoft AD directory data across multiple Regions so you can improve performance for users and applications in disperse geographic locations AWS Managed Microsoft AD uses native AD replication to replicate your directory’s data securely to the new Region Multi Region replication is only supported for the Enterprise Edition of AWS Managed Microsoft AD AWS Managed Microsoft AD enables you to forward all domain controller’s Windows Security event log to Amazon CloudWatch giving you the ability to monitor your use of the directory and any administrative intervention performed in the course of AWS This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 6 operating the service It is also approved for applications in the AWS Cloud tha t are subject to compliance by the US Health Insurance Portability and Accountability Act (HIPAA) Payment C ard Industry Data Security Standard (PCI DSS) Federal Risk and Authorization Management (FedRAMP) or Service Organizational Control (SOC) when you enable compliance for your directory You can also tailor security with features that enable you to manage password policies and enable secure LDAP communications through Secure Socket Layer (SSL)/Transport Layer Security (TLS) You can also enable multi factor authentication (MFA) for AWS Managed Micros oft AD This authentication provides an additional layer of security when users access AWS applications from the internet such as Amazon WorkSpaces or Amazon QuickSight AWS Managed Microsoft AD enables you to extend your schema and perform LDAP write operations These features combined with advanced security features such as Kerberos Constrained Deleg ation and Group Managed Service Account provide the greatest degree of compatibility for Active Directory aware applications like Microsoft SharePoint Microsoft SQL Server Always On Availability Groups and many NET applications Because Active Directo ry is an LDAP directory you can also use AWS Managed Microsoft AD for Linux Secure Shell (SSH) authentication and other LDAP enabled applications The full list of supported AWS applications is available on the AWS Documentation site AWS Managed Microsoft AD runs actual Window Server 2012 R2 Active Directory Domain Services and operates at the 2012 R2 functional level AWS Managed Microsoft AD is available in two editions: Standard and Enterprise These editions have different storage capacity ; Enterprise Edition also has multi region features Edition Storage capacity Approximate number of objects that can be stored* Approximate number of users in domain* Standard 1 GB ~30000 Up to ~5000 users Enterprise 17 GB ~500000 Over 5000 users * The number of objects varies based on type of objects schema extensions number of attributes and data stored in attributes This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 7 Note: AWS Domain Administrators have full administrative access to all domains hosted on AWS See your agreement with AWS and the AWS Data Privacy FAQ for more information about how AWS handles content that you store on AWS systems including directory informat ion You do not have Domain or Enterprise Admin permissions and rely on delegated groups for administration AWS Managed Microsoft AD can be used for following scenarios: managing access to AWS Management Console and cloud services joining EC2 Windows ins tances to Active Directory deploying Amazon RDS databases with Windows authentication using FSx for Windows File Services and signing in to productivity tools like Amazon Chime and Amazon WorkSpaces For more information on this solution see Design consideration for AWS Managed Microsoft Active Directory in this document Active Directory on EC2 If you prefer to extend your Active Directory to AWS and manage it yourself for flexibility or other reasons you h ave the option of running Active Directory on EC2 For more information s ee Design considerations for running Active Directory on EC2 instances in this document Comparison of Active Directory Services on AWS The following table compares the features and functions between various Directory Services options available on AWS Many features are not applicable directly to AWS AD Connector because it is actins only as a proxy to the existing Active Directory domain Function AWS AD Connector AWS Managed Microsoft AD Active Directory on EC2 Managed service yes yes no Multi Region deployment n/a yes Enterprise Edition yes Share directory with multiple accounts no yes no Supported by AWS applications (Amazon Chime Amazon WorkSpaces AWS Single Sign On & etc) yes yes yes (through federation or AD Connector) This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 8 Function AWS AD Connector AWS Managed Microsoft AD Active Directory on EC2 Supported by RDS (SQL Server Oracle MySQL PostgreSQL and MariaDB) n/a yes no Supported by FSx for Windows File Server n/a yes yes Creating users and groups yes yes yes Joining computers to the domain yes yes yes Create trusts with existing Active Directory domains and forests n/a yes yes Seamless domain join for Windows and Linus EC2 instances yes yes yes with AWS AD Connector Schema extensions n/a yes yes Add domain controllers n/a yes yes Group Managed Service Accounts n/a yes Depends on the Windows Server version Kerberos constrained delegation n/a yes yes Support Microsoft Enterprise CA n/a yes yes Multi Factor Authentication yes through RADIUS yes through RADIUS yes with AD Connector Group policy n/a yes yes Active Directory Recycle bin n/a yes yes PowerShell support n/a yes yes This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 9 Core infrastructure design on AWS for Windows Workloads and Directory Services Planning AWS accounts and Organization AWS Organizations helps you centrally manage your AWS accounts identity services and access policies for your workloads on AWS Whether you a re a growing startup or a large enterprise Organizations helps you to centrally manage billing; control access compliance and security; and share resources across your AWS accounts For more information refer to the AWS Organizations User Guide With AWS Organization s you can centrally define critical resources and make them available to accounts across your organization For example you can authenticate against your central identity store and enable applications deployed in other accounts to access it If your users need to manage AWS services and access AWS applications with their Active Directory credentials we recommend integrating your identity servi ce with the management account in AWS Organization s • Deploy AWS Managed AD in the management account with trust to your on premises A ctive Directory to allow users from any trusted domain to access AWS Applications Share AWS Managed AD to other accounts across your organization • Deploy AWS Single Sign On in the management account to centrally manage access to multiple AWS accounts and business applic ations and provide users with single sign on access to all their assigned accounts and applications from one place AWS SSO also includes built in integrations to many business applications such as Salesforce Box and Microsoft Office 365 Network design considerations for AWS Managed Microsoft AD Network design for Microsoft workloads and directory services consist s of network connectivity and DNS names resolution This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 10 To plan the network topology for your organization refer to the whitepaper Building a Scalable and Secure Multi VPC AWS Network Infrastructure and consider the following recommendations: • Plan your IP networks for Microsoft workloads without overlapping address spaces Microsoft does not recommend using Active Directory over NAT • Place directory services into a centralized VPC that is reachable from any other VPC with workloads depending on Active Directory • By default instances that you launch into a VPC cannot communicate with your onpremises network To extend your existing AD DS i nto the AWS Cloud you must connect your on premises network to the VPC in one of two ways: by using Virtual Private Network (VPN) tunnels or by using AWS Direct Connect To connect multiple VPCs in AWS you can use VPC peering or AWS Transit Gateway Network port requirements and security groups Active Directory requires certain network ports to be open to allow traffic for LDAP AD DS replication user authentication Windows Time services Distributed File System (DFS) and many more When you deploy Active Directory on EC2 instances using the AWS Quick Start or AWS Managed Microsoft AD it automatically creates a new security group with all required por t rules If you manually deploy your Active Directory you need to create a security group and configure rules for all required network protocols For a complete list of ports see Active Directory and Active Directory Domain Services Port Requirements in the Microsoft TechNet Library DNS names resolution Active Directory heavily relies on DNS services and hosts its own DNS services on domain controllers To establish seamless name resolution in all your VPCs and your onpremises network create a Route 53 Resolver deploy inbound/ outbound endpoints in your VPC and configure conditional forwarders to all of your Active Directory domai ns (including AWS Managed AD and on premises A ctive Directory ) in the Route 53 Resolver Share centralized Route 53 Resolver endpoints across all VPC in your organization Create conditional forwarders on your on premises DNS servers for all Route 53 DNS This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 11 zones and DNS zones on AWS Managed AD and point them to Route 53 Resolver Endpoints Figure 2 Route 53 Resolver configuration for hybrid network Here are design considerations for DNS resolution : • Make all Active Directory DNS domain s resolvable for all clients because they are using it to locate Active Directory services and register their DNS names using dynamic updates • Try to keep the DNS name resolution local to the AWS Region to reduce latency • Use Amazon DNS Server (2 resolve r) as a forwarder for all other DNS domains that are not authoritative on your DNS Servers on A ctive Directory domain controllers This setup allows your DCs to recursively resolve records in Amazon Route 53 private zone and use Route 53 Resolver condition al forwarders • Use Route 53 Resolver Endpoints to create DNS resolution hub and manage DNS traffic by creating conditional forwarders For more information on designing a DNS name resolution strategy in a hybrid scenario see the Amazon Route 53 Resolver for Hybrid Clouds blog post This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 12 Note: The Amazon EC2 instance limits the numbe r of packets that can be sent to the Amazon provided DNS server to a maximum of 1024 packets per second per network interface This limit cannot be increased If you run into this performance limit you must set up conditional forwarding for Amazon Route 5 3 private zones to use the Amazon DNS Server (2 resolver) and use root hints for internet name resolution This setup reduces the chances of you exceeding the 1024 packet limit on AWS DNS resolver Design consideration for AWS Managed Microsoft Active Dir ectory Active Directory depends on the network and accounts design Before you select the right Active Directory topology you must choose your network and organizational design Although there is no one sizefitsall answer for how many AWS accounts a par ticular customer should have most companies create more than one AWS account as multiple accounts provide the highest level of resource and billing isolation in the following cases: • The business requires strong fiscal and budgetary billing isolation betw een specific workloads business units or cost centers • The business requires administrative isolation between workloads • The business requires a particular workload to operate within specific AWS service limits and not impact the limits of another workload • The business’s workloads depend on specific instance reservations to support high availability (HA) or disaster recovery (DR) capacity requirements Single account AWS Region and VPC The simplest case is when you need to deploy a new solution i n the cloud from scratch You can deploy AWS Managed Microsoft AD in minutes and use it for most of the services and applications that require Active Directory This solution is ideal for scenarios with no additional requirements for logical isolation betw een application tiers or administrat ors This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 13 Figure 3 Managed A ctive Directory architecture deployed by Quick Start Multiple accounts and VPCs in one AWS Region Large organizations use multiple AWS accounts for administrative delegation and billing purpose s You can share a single AWS Managed Microsoft AD with multiple AWS accounts within one AWS Region This capability makes it easier and more cost effective for you to manage directory aware workloads from a single directory across accounts and VPCs This option also allows you seamless ly join your Amazon EC2 Windows instances to AWS Managed Microsoft AD This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 14 Figure 4 Sharing single AWS Managed Microsoft AD with another account AWS recommends that you create a separate account for identity services like Active Directory and only allow a very limited group of administrators to have access to this account Generally you should treat Active Directory in the cloud in the same manner as on premises A ctive Directory Just as you would limit access to a physica l data center make sure to limit administrative access to the AWS account control Create additional AWS accounts as necessary in your organization and share the AWS Managed Microsoft AD with them After you have shared the service and configured routing these users can use A ctive Directory to join EC2 Windows instances but you maintain control of all administrative tasks Deploy AWS Managed AD in your management account of AWS Organization s This allow s you to use Managed AD for authentication with AWS Identity and Access Management (IAM) to access the AWS Management Console and other AWS applications using your Active Directory credentials Multiple AWS Regions deployment AWS Managed Microsoft AD Enterprise Edition supports Multi Region deployment You can use automated multi Region replication in all Regions where AWS Managed Microsoft AD is available AWS services such as Amazon RDS for SQL Server and Amazon FSx connect to the local instances of the global directory This allows your users to sign in once to AD aware applications running in AWS as well as AWS services like Amazon RDS for SQL Server in a ny AWS Region – using credentials from AWS Managed Microsoft AD or a This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 15 trusted AD domain or forest Refer to AWS Directory Service documentation for the current list of AWS Services supporting Mu ltiRegion replication feature With multi Region replication in AWS Managed Microsoft AD AD aware applications such as SharePoint SQL Server Always On AWS services such as Amazon RDS for SQL Server and Amazon FSx for Windows File Server use the dire ctory locally for high performance and are multi Region for high resiliency The following list comprises additional benefits of Multi Region replication • It enables you to deploy a single AWS Managed Microsoft AD instance globally quickly and eliminates the heavy lifting of self managing a global AD infrastructure • Optimal performance for workloads deployed in multiple regions • Multi Region resiliency AWS Managed Microsoft AD handles automated software updates monitoring recovery and the security of the underlying AD infrastructure across all Regions • Disaster recovery In the event that all domain controllers in one Region are down AWS Managed Microsoft AD recovers the domain controllers and replicates the directory data automatically Meanwhile do main controllers in other Regions are up and running To deploy AWS Managed Microsoft AD across multiple Regions you must create it in Primary region and after that a dd one or more Replicated regions Consider following factors for your Active Directory d esign: • When you deploy a new Region AWS Managed Microsoft AD creates two domain controllers in the selected VPC in the new Region You can add more domains controllers later for scalability • AWS Managed Microsoft AD uses a backend network for replication and communications between domain controllers • AWS Managed Microsoft AD creates a new Active Directory Site and names it the same name of the Region For example us east1 You can also rename this later using the Active Directory Sites & Services tool • AWS Managed AD is configured to use change notifications for inter site replications to eliminate replication delays This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 16 After you add your new Region you can do any of the following tasks : • Add more domain controllers to the new Region for horizontal scala bility • Share your directory with more AWS accounts per Region Directory sharing configurations are not replicated from the primary Region and you may have different sharing configuration in different region based on your security requirements • Enable log forwarding to retrieve your directory’s security logs using Amazon CloudWatch Logs from the new Region When you enable log forwarding you must provide a log group name in each Region where you replicated your directory • Enable Amazon Simple Notification Service (Amazon SNS) monitoring for the new Region to track your directory health status per Region Enable Multi Factor Authentication for AWS Managed Microsoft AD You can enable multi factor authentication (MFA) for your AWS Managed Microsoft AD to increase security when your users specify their A ctive Directory credentials to access supported Amazon enterprise applications When you enable MFA your users enter their user name and password (first factor) and then e nter an authentication code (second factor) that they obtain from your virtual or hardware MFA solution These factors together provide additional security by preventing access to your Amazon enterprise applications unless users supply valid user credenti als and a valid MFA code To enable MFA you must have an MFA solution that is a remote authentication dial in user service (RADIUS) server or you must have an MFA plugin to a RADIUS server already implemented in your on premises infrastructure Your MFA solution should implement onetime passcodes (OTP) that users obtain from a hardware device or from software running on a device (such as a mobile phone) This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 17 Figure 6 Using AWS Managed Microsoft Active Directory with MFA for access to Amazon Work Spaces A more detailed description of this solution is available on the AWS Security Blog Active Directory permissions delegation When you use AWS Managed Microsoft AD AWS assumes responsibility for some of the service level tasks so that you may focus on other business critical tasks The following service level tasks are a utomatically performed by AWS • Taking snapshots of the Directory Service and providing the ability to recover data • Creating trusts by administrator request • Extending Active Directory schema by administrator request • Managing Active Directory forest config uration • Managing monitoring and updating domain controllers • Managing and monitoring DNS service for Active Directory • Managing and monitoring Active Directory replication • Managing Active Directory sites and networks configuration This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 18 With AWS Managed Microsoft AD you also may delegate administrative permissions to some groups in your organization These permissions include managing user accounts joining computers to the domain managing group policies and password policies managing DNS DHCP DFS RAS CA and other services The full list of permissions that can be delegated is described in the AWS Directory Service Administration Guide Work with all teams that are using Active Directory services in your organization and create a li st with all of the permissions that must be delegated Plan security groups for different administrative roles and use AWS Managed Microsoft AD delegated groups to assign permissions Check the AWS Directory Service Administration Guide to make sure that it is possible to delegate all of the required permissions Design considerations for running Active Directory on EC2 instances If you cannot use AWS Managed Microsoft AD and you have Windows workloads you want to deploy on AWS you can still run Active Directory on EC2 instances in AWS Depending on the number of Regions where you are deploying your solution your Active Directory design may slightly differ The following section provides a deployment guide and recommendation on how you can deploy Active Directory on EC2 instances in AWS Single Region deployment This deployment scenario is applicable if you are operating in a singl e Region or you do not need Active Directory to be in more than a single Region The deployment options or architecture patterns are not significantly different whether you are operating in a single VPC or multiple VPCs If you are using multiple VPCs you must ensure that network connectivity between the VPCs is available through VPC peering VPN or AWS Transit Gateway The following diagrams depict how Active Directory can be deployed in a single Region in a single VPC or multiple VPCs This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Activ e Directory Domain Services on AWS 19 Figure 7 Deplo ying Active Directory on EC2 instances in a single Region for single VPC Figure 8 Deploying Active Directory on EC2 instances in a single Region for multiple VPCs Consider the following points when deploying Active Directory in this architecture: • We recommend deploying at least two domain controllers (DCs) in a Region These domain controllers should be placed in different AZs for availability reasons • DCs and other non internet facing servers should be placed in private subnets • If you require additional DCs due to performance you can add more DCs to existing AZs or deploy to another available AZ This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 20 • Configure VPCs in a Region as a single A ctive Directory site and define A ctive Directory subnets accordingly This configuration ensures that all of your clients correctly select the closest available DC • If you have multiple VPCs you can centralize the Active Directory services in one of the existing VPCs or create a shared services VPC to centralize the domain controllers • You must ensure you have highly available network connectivity between VPCs such as VPC peering If you are connecting the VPCs using VPNs or other methods ensure connectivity is highly available • If you want to use your self managed Active Directory credentials to acc ess AWS Services or thirdparty services you can integrate your self managed AD with AWS IAM and AWS Single Sign On using AWS AD Connector or AWS Managed AD through a trust relationship In these cases AD Connector or AWS Managed AD must be deployed in t he management account of your organization Multi region/global deployment of self managed AD If you are operating in more than one Region and require Active Directory to be available in these Regions use the multi region/global deployment scenario Withi n each of the Regions use the guidelines for single Region deployment as all of the single Region best practices still apply The following diagrams depict how Active Directory can be deployed in multiple Regions In this example we are showing Active Di rectory deployed in three Regions that are interconnected to each other using cross Region VPC peering In addition these Regions are also connected to the corporate network using AWS Direct Connect and VPN This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 21 Figure 9 Deploying Active Directory on EC2 instances in multiple Regions with multiple VPCs Consider the following recommendations when deploying Active Directory in this architecture : • Deploy a t least two domain controllers in each Region These domain controllers should be placed in different AZs for availability reasons • Configure VPCs in a reg ion as a single A ctive Directory site and define A ctive Directory subnets accordingly This configuration ensures all of your clients will correctly select the closest available domain controller • Ensure robust inter Region connectivity exists between all of the Regions Within AWS you can leverage cross Region VPC peering to achieve highly available private connectivity between the Regions You can also leverage the Transit VPC solution to interconnect multiple regions Designing A ctive Directory sites an d services topology It’s important to define A ctive Directory sites and subnets correctly to avoid clients from using domain controllers that are located far away as this would cause increased latency See How Domain Controllers are Located in Windows This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Ser vices on AWS 22 Follow these best practices for configuring sites and services: • Configure one A ctive Directory site per AWS Region If you are operating in multiple AWS Regions we recommend configuring one A ctive Directory site for each of these Regions • Define the entire VPC as a subnet and assign it to the A ctive Directory site defined for this Region • If you have multiple VPCs in the same Region define each of these VPCs as separate subnets and assign it to the single A ctive Directory site set up for this Region This setup allows you to use domain controllers in that Region to service all clients in that region • If you have enabled IPv6 in your Amazon VPC create the necessary IPv6 subnet definition and assign it to this A ctive Directory site • Define all IP address ranges If clients exist in undefined IP address ranges the clients might not be associated with the correct A ctive Directory site • If you have reliable high speed connectivity between all of the sites you can use a single site link for all of your AD sites and maintain a single replication configuration • Use consistent sites names in all AD forests connected by trusts Security considerations Trust relationships with on premises A ctive Directory Whether you are deploying Active Directory on EC2 instances or using AWS Managed Microsoft AD these are the three common deployment patterns seen on AWS 1 Deploy a standal one forest/domain on AWS with no trust In this model you set up a new forest and domain on AWS which is different and separate from the current Active Directory that is running on premises In this deployment both accounts (user credentials service acc ounts) and resources (computer objects) reside in Active Directory running on AWS and most or all of the member servers run on AWS in single or multiple Regions For this deployment there is no network connectivity requirement between on premises and AWS for the purposes of Active Directory as nothing is shared between the two A ctive Directory forests This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 23 2 Deploy a new forest/domain on AWS with one way trust If you are planning on leveraging credentials from an on premises A ctive Directory on AWS member serve rs you must establish at least a one way trust to the Active Directory running on AWS In this model the AWS domain becomes the resource domain where computer objects are located and on premises domain becomes the account domain Note: You must have robust connectivity between your data center and AWS A connectivity issue can break the authentication and make the whole solution not accessible for users Consider to extend your Active Directory domains to AWS to eliminate dependency on connectivity with onpremises infrastructure or deploy a multi path AWS Direct Connect or VPN connection 3 Extend your existing domain to AWS In this model you extend your existing Active Directory deployment from on premises to AWS which means adding additional domain controllers (running on Amazon EC2) to your existing domain and placing them in multiple AZs within your Amazon VPC If you are operating in multiple Regions add domain controllers in each of these Regions This deployment is easy flexible and provides the following advantages: o You are not required to set up additional trusts o DCs in AWS are handling both accounts and resources o More resilient to network connectivity issues o You can seamlessly set up and use AWS Cloud in a hybrid scenario with least impact to the applications (Note that network connectivity is required between your data center and AWS for initial and on going replication of data between the domain controllers) When you use cross forest trust relationships in Active Direct ory you need to use consistent Active Directory site names in both forests to have optimal performance Refer to the article Domain Locator Across a Forest Trust for more information See How Domain and Forest Trusts Work on the Microsoft Doc umentation website for more information This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 24 Multifactor authentication Multi factor authentication (MFA) is a simple best practice that adds an extra layer of protection on top of your user name and password With MFA enabled when users sign in to the AWS Management Console they are prompted for thei r user name and password (the first factor —what they “know ”) then prompted for an authentication response from their AWS MFA device (the second factor —what they “have ”) Taken together these multiple factors provide increased security for your AWS accoun t settings and resources We recommend enabling MFA on all of your privileged accounts regardless of whether you are using IAM or federating through SSO AWS account security Since you are running your domain controllers on Amazon EC2 securing your AWS account is an important process in securing your Active Directory domain Follow these recommendations to make sure your AWS account is secure • Enable MFA and then lock away your AWS root user credential • Use IAM groups to manage permission if you are using IAM users • Grant least privilege to all your users within AWS • Enable MFA for all privileged users • Use EC2 roles for applications that run on EC2 instances • Do not share access keys • Rotate credentials regularly • Turn on and analyze log files in AWS CloudTrail VPC Flow Logs and Amazon S3 bucket logs • Turn on encryption for data at rest and in transit where necessary Domain controller security Domain controllers provide the physical storage for the AD DS database i n addition to providing the services and data that allow enterprises to effectively manage their servers workstations users and applications If privileged access to a domain controller is obtained by a malicious user that user can modify corrupt or destroy the AD DS database and by extension all of the systems and accounts that are managed This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 25 by Active Directory Make sure your domain controller is secure to avoid compromising your Active Directory data The following points are some of the best pract ices to secure domain controllers running on AWS: • Secure the AWS account where the domain controllers are running by following least privilege and role based access control • Ensure unauthorized users don’t have access in your AWS account to create/access A mazon Elastic Block Store (Amazon EBS) snapshots launch or terminate EC2 Instances or create/copy EBS volumes • Ensure you are deploying your domain controllers in a private subnet without internet access Ensure that subnets where domain controllers are running don’t have a route to a NAT gateway or other device that would provide outbound internet access • Keep your security patches up todate on your domain controllers We recommend you first test the security patches in a non production environment • Restrict ports and protocols that are allowed into the domain controllers by using security groups Allow remote management like remote desktop protocol (RDP) only from trusted networks • Leverage the Amazon EBS encryption feature to encrypt the root and addit ional volumes of your domain controllers and use AWS Key Management Service (AWS KMS) for key management • Follow Microsoft recommended security configuration baselines and Best Practices for Securing Active Directory Other considerations FSMO Roles You can follow the same recommendation you would follow for your on premises deployment to determine FSMO roles on DCs See also best practices from Microsoft In the case of AWS Managed Microsoft AD all domain controllers and FSMO roles assignments are managed by AWS and do not require you to manage or change them Global Catalog Unless you have slow connections or an extremely large A ctive Directory database w e recommend adding global catalog role to all of your domain This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 26 controllers in multi domain forests (except the domain controller with the Infrastructure Master role) If you are hosting Microsoft Exchange in AWS Cloud at least one global catalog server is required in a site with Exchange servers For more information about global catalog see Microsoft documentation Since there is only one domain in the forest for AWS Managed Microsoft AD all domain controllers are configured as global catalog and will have full informatio n about all objects Read Only Domain Controllers (RODC) It’s possible to deploy RODC on AWS if you are running A ctive Directory on EC2 instances and require it and there are no special considerations for doing so AWS Managed Microsoft AD does not suppo rt RODCs All of the domain controllers that are deployed as a part of AWS Managed Microsoft AD are writable domain controllers Conclusion AWS provides several options for deploying and managing Active Directory Domain Services in the cloud and hybrid env ironments You can leverage AWS Managed Microsoft AD if you no longer want to focus on heavy lifting like managing the availability of the domain controllers patching backups and so on Or you can run Active Directory on EC2 instances if you need to have full administrative control on your Active directory In this whitepaper we have discussed these two main approaches of deploying Active Directory on AWS and have provided you with guidance and consideration for each of the de sign Depending on our deployment pattern scale requirements and SLA you may select one of these options to support your Windows workloads on AWS Contributors Contributors to this document include : • Vladimir Provorov Senior Solutions Architect Amazon Web Services • Vinod Madabushi Enterprise Solutions Architect Amazon Web Services This version has been archived For the latest version of this document visit: https://docsawsamazoncom/whitepapers/latest/ activedirectorydomainservices/activedirectory domainserviceshtmlAmazon Web Services Active Directory Domain Services on AWS 27 Further Reading For additional information see: • AWS Whitepapers • AWS Directory Service • Microsoft Workloads on AWS • Active Directory Domain Services on the AWS Cloud: Quick Start Reference Deployment • AWS Documentation Document Revisions Date Descript ion November 2020 AWS Managed Microsoft AD multi region feature update August 2020 Numerous updates throughout December 2018 First publication
|
General
|
consultant
|
Best Practices
|
Amazon_Aurora_Migration_Handbook
| "This paper has been archived For the latest Ama zon Aurora Migration content refer to: https://d1aw(...TRUNCATED)
|
General
|
consultant
|
Best Practices
|
Amazon_Aurora_MySQL_Database_Administrators_Handbook_Connection_Management
| "This version has been archived For the latest version of this document visit: https://docsawsamazon(...TRUNCATED)
|
General
|
consultant
|
Best Practices
|
Amazon_EC2_Reserved_Instances_and_Other_Reservation_Models
| "Amazon EC2 Reserved Instances and Other AWS Reservation Models AWS Whitepaper Amazon EC2 Reserved I(...TRUNCATED)
|
General
|
consultant
|
Best Practices
|
Amazon_Elastic_File_System_Choosing_Between_Different_Throughput_and_Performance_Mode
| "This paper has been archived For the latest technical content refer t o the AWS Wh i t epapers & Gu(...TRUNCATED)
|
General
|
consultant
|
Best Practices
|
Amazon_Virtual_Private_Cloud_Network_Connectivity_Options
| "Amazon Virtual Private Cloud Connectivity Options January 2018 © 2018 Amazon Web Services Inc or i(...TRUNCATED)
|
General
|
consultant
|
Best Practices
|
End of preview. Expand
in Data Studio
No dataset card yet
- Downloads last month
- 27