Friday, 11 March 2016

Public Cloud vs Private Cloud

The is a lot of debate about Public Cloud vs Private Cloud, yet they are quite different. To me the most significant difference is huge selection of services available with public cloud, services enable devops in small teams to develop very quickly, all accessible through APIs. Public cloud has many services, yet private cloud is mostly just IaaS. Let's investigate the details.....

 Public cloud has following characteristics:
  • Built on low cost commodity infrastructure
  • Pay for what you use with itemised billing, OPEX based.
  • No expensive upfront CAPEX investment
  • No need to project future capacity as public cloud is flexible
  • Low cost compared to Private cloud when Private cloud is built on expensive proprietary infrastructure from legacy IT vendors
  • Self service, no for need many slow silos of IT teams to deploy new application.
  • Everything is software defined as everything can be done with APIs
  • Heavy use of automation is possible as everything is software defined
  • Enabling staff to be far more efficient through heavy use of automation and service
  • IT Ops working closely with Developers to help the business
  • Small, agile Devops teams can build very quickly rather than rely on slow legacy IT with IT silo mentality
  • Hundreds of managed service, not just IaaS like Private Cloud, here are just a few AWS examples:
    • Storage as a Service (SaaS):
      • Amazon Simple Storage Service (S3) provides Web Service based storage.
      • Amazon Glacier provides a low-cost, long-term storage option (compared to S3). High redundancy and availability, but low-frequent access times. Intended for archiving data.
      • AWS Storage Gateway, an iSCSI block storage virtual appliance with cloud-based backup.
    • Database as a Service (DBaaS):
      • Amazon DynamoDB provides a scalable, low-latency NoSQL online Database Service backed by SSDs.
      • Amazon ElastiCache provides in-memory caching for web applications. This is Amazon's implementation of Memcached and Redis.
      • Amazon Relational Database Service (RDS) provides a scalable database server with MySQL, Oracle, SQL Server, and PostgreSQL support.
      • Amazon SimpleDB allows developers to run queries on structured data. It operates in concert with EC2 and S3 to provide "the core functionality of a database".
      • Amazon Aurora is an enterprise grade MySQL-compatible relational database engine that combines the speed and availability of high-end commercial databases
    • Data Warehouse as a Service: Amazon Redshift provides petabyte-scale data warehousing with column-based storage and multi-node compute.
    • AWS Elastic Beanstalk provides quick deployment and management of applications in the cloud, Platform as a Service (Paas) supporting: Ruby, PHP and Python,.NET Framework,Java, Node.js, Docker
    • Directory Service as a Service: AWS Directory Service a managed service that allows you to connect AWS resources with an existing on-premises Microsoft Active Directory or to set up a new, stand-alone directory in the AWS Cloud.
    • Monitoring as a Service: Amazon CloudWatch, provides monitoring for AWS cloud resources and applications, starting with EC2.
    • Orchestration as a Servie: AWS OpsWorks provides configuration of EC2 services using Chef.
    • Business Intelligence as a Service with QuickSight
    • Desktop as a Service - VDI with Workspaces
    • Load Balancing (ELB)
    • DNS as a Service with Route 53
    • Hadoop as a Service with Elastic MapReduce
    • Content Delivery Network with Cloudfront
    • AWS Management Console (AWS Console), A web-based point and click interface to manage and monitor the Amazon infrastructure suite including (but not limited to) EC2, EBS, S3, SQS, Amazon Elastic MapReduce, and Amazon CloudFront. Amazon also makes available a mobile application for Android which has support for some of the management features from the console.
    • Amazon CloudHSM - The AWS CloudHSM service helps you meet corporate, contractual and regulatory compliance requirements for data security by using dedicated Hardware Security Module (HSM) appliances within the AWS cloud.
    • AWS Key Management Service (KMS) a managed service that makes it easy for you to create and control the encryption keys used to encrypt data.
    • Amazon API Gateway is a service for publishing, maintaining and securing web service APIs.
    • Amazon CloudSearch provides basic full-text search and indexing of textual content.
    • Amazon DevPay, currently in limited beta version, is a billing and account management system for applications that developers have built atop Amazon Web Services.
    • Amazon Elastic Transcoder (ETS) provides video transcoding of S3 hosted videos, marketed primarily as a way to convert source files into mobile-ready versions.
    • Amazon Flexible Payments Service (FPS) provides an interface for micropayments.
    • Amazon Simple Email Service (SES) provides bulk and transactional email sending.
    • Amazon Simple Queue Service (SQS) provides a hosted message queue for web applications.
    • Amazon Simple Notification Service (SNS) provides a hosted multi-protocol "push" messaging for applications.
    • Amazon Simple Workflow (SWF) is a workflow service for building scalable, resilient applications.
    • Amazon Cognito a simple user identity and data synchronization service that helps you securely manage and synchronize app data for users across their mobile devices.
    • Amazon AppStream a flexible, low-latency service that lets you stream resource intensive applications and games from the cloud.
    • Amazon Kinesis streams data in real time with the ability to process thousands of data streams on a per-second basis. The service, designed for real-time apps, allows developers to pull any amount of data, from any number of sources, scaling up or down as needed.
    • Code Repository as a Service with git compatible Code Commit
    • Continuous Delivery as a Service with Code Pipeline
  • Public Cloud is completely proprietary, it locks you into the cloud vendor ecosystem
    • Public cloud is most efficient when services are used, yet these services and APIs are completely proprietary
    • The more services you use the more locked in the application becomes
    • Changing from one cloud to another requires rewrite and retesting of application, usually labour cost is more expensive than cloud infrastructure cost hence cost prohibitive to migrate to a different cloud vendor
    • Writing application for multi-cloud is more time consuming and expensive as APIs and services are different, hence you lose agility 
    • For decades we used proprietary infrastructure like Windows, Cisco, F5, EMC, Netapp, Oracle, SQL Server, VMware, HP Openview, Tivoli, etc, etc these were also proprietary and  when we purchased these expensive products we were locked in, yet we accepted this. Why can we not accept the same with applications when we choose proprietary public cloud especially as it is lower in cost?

Private cloud has following characteristics:
  • Virtualization alone is not private cloud
  • Need to build public cloud: 
    • Takes months to evaluate, plan, cost, design and build and test - slow and expensive
    • Need to guesstimate capacity requirements
    • Expensive OPEX as usually based on expensive legacy IT vendor products
    • Can be lower cost than public cloud if products are based on commodity low cost hardware and open source that is easy to manage (not based on complex bleeding edge science project) .....most private clouds are not lower cost than public cloud as they are based on expensive products and require significant resource costs to manage private cloud!
    • Many components from many different vendors, spend significant time integration testing. Many points of support contact.
  • Once you purchase infrastructure you are locked into long term commitment
  • No end-to-end consistent set of APIs 
  • As lacking APIs and services difficult to automate everything
  • Need many silos of IT teams to support, silos are slow and not suitable for agile Devops model
  • Private Cloud lacks wealth of Services that allow projects to be deployed quickly

Why is Private Cloud (mostly) expensive? Because they are usually based on expensive products from legacy IT vendors:

IT Vendors Gross Profit Margin:
Revenue (Billion)Gross Profit (Billion)Gross Profit Margin (%)
IBM http://goo.gl/DVjJn90.1546.451.47%
HP http://goo.gl/y6OBj108.2826.6224.58%
Cisco http://goo.gl/FhjYv48.6827.7757.05%
Dell http://goo.gl/ARBIU56.9412.1921.40%
EMC http://goo.gl/6Selz24.5715.2562.07%
VMware http://goo.gl/EkBWd6.195.1282.71%
NetApp http://goo.gl/1n5Qa6.123.8362.58%
Oracle http://goo.gl/zZbYc38.2331.0481.19%
Microsoft http://goo.gl/37jZ2B94.7459.963.23%
Amazon http://goo.gl/OzKxCg91.9626.2428.53%
SuperMicro http://goo.gl/YoMKKn1.850.2255412.19%
Foxconn https://goo.gl/9gOuYw97.1215.2615.71%

For customer with reasonable scale say 100+ VMs and applications that just need IaaS then I prefer low cost private cloud using low cost commodity infrastructure and open source that is easy to maintain, has simple APIs and easy to use with self service portal. If this is done well then private cloud can be significantly lower cost than public cloud. Yet if you need other services that are not easily available on private cloud then public cloud is far more efficient.....just note that proprietary public cloud services will lock you into vendors public cloud!

 See also:
http://blog.zorangagic.com/2015/03/build-your-own-cloud-for-3000-vms.html
http://blog.zorangagic.com/2016/02/converged-infrastructure-price.html
at

No comments:

Post a Comment