Elastic cloud hosting

Written by: , Apr. 24, 2018
SECTION 1

Overview

Nginx with IBM Cloud Private architecture diagram

 

What is elastic cloud hosting?

Elastic cloud hosting plans allow for web/mobile applications to scale past single web server limitations in production or for entire data center operations to be outsourced to remote facilities for cost savings and greater efficiency. Elastic cloud hosting is based on cluster web server networks that auto-scale according to resource provisioning limits to meet the code/database processing requirements of high traffic web/mobile software applications. Elastic cloud hosting includes network traffic load balancing between hardware and data center locations similar to CDN routing capabilities through virtualization and orchestration engines that synchronize file storage & database changes across multiple web servers.

Elastic web servers are used to manage user traffic spikes or "slashdotting" on public cloud hosting plans, allowing software companies, social networks, media publishers, and ecommerce sites to manage web servers through data center resources that auto-scale with user demand. Elastic cloud hosting operates on the model of "pay-per-unit" of CPU multi-processing power, SSD/HDD storage, fiber bandwidths, or dedicated RAM consumed through bulk wholesale rates for IT business services available to companies that switch to public cloud solutions. The leading elastic cloud server platforms are Google Cloud/Compute Engine, Kubernetes, Amazon Web Services (AWS) EC2, LiquidWeb CloudSites, Mesosphere DC/OS, and Microsoft Azure.

Varnish http cache

Elastic cloud hosting vs. Dedicated servers

The elastic cloud cluster server is designed to challenge the use of dedicated web servers in data center environments for web hosting at scale for companies through better load balancing of network traffic between hardware units, avoiding "single point of failure" models. Many companies require multiple data centers for backup and storage requirements that can also be used for elastic cloud web server orchestration. Similarly, a CDN or Content Delivery Network can be used to distribute cached versions of a website's files to over 100 data centers internationally, which takes the major strain away from local dedicated hardware. VPS plans have all of the advantages of dedicated servers, including account isolation, SSD speed, low density domain name provisioning, & advanced programming language or database customization, while being available at a lower overall cost than dedicated servers for the same hardware configurations.

"VPS, VMs, & containers can be scaled through elastic cluster web servers most efficiently in data center operations. Public, private, & hybrid cloud constructs all take advantage of elastic cluster server resources for web/mobile application support in production."

Virtualization allows systems administrators to make better use of hardware, where containers can increase software app portability between cloud hosts or allow full web server software stack packages to be installed in seconds from disk images. Businesses need to build on tools that will scale seamlessly to support the higher levels of traffic without requiring a platform rebuild when growth patterns are at their highest rates. There are open source and proprietary licensed software tools from a number of IT companies that assist businesses with elastic web server orchestration in data center management, with OpenStack, CloudStack, Mesosphere DC/OS, Microsoft Azure, Kubernetes, Docker Swarm, CoreOS, Virtuozzo/OpenVZ, & VMware solutions all competing in the industry. Platform-as-a-Service (PaaS) plans challenge private cloud investment and maintenance costs for companies as the dividing line between open source and proprietary-licensed elastic web server software, fueling industry-wide adoption for enterprise-level data center management requirements.

end to end cloud management

Elastic cloud hosting - advantages

Elastic cloud hosting can be favored for speed, reliability, efficiency, billing, security, responsibility, innovation, CI/CD requirements, DevOps best practices, and ability to scale without limit to the highest levels of user traffic in web/mobile application support. Elastic web hosting plans guarantee that users will receive a fast download and page processing time when accessing a website no matter how many simultaneous browsers are visiting, and that code will not be intentionally throttled by shared resource hosting plans. Any website on shared web hosting plans will not be able to support more than 100 to 150 simultaneous users regularly without exceeding account limits for CPU processing or I/O connections. VPS and dedicated server plans allow for more RAM, CPU cores, SSD storage, etc. to be allocated to a website in production, but elastic server plans allow for these resources to be auto-scaled or scheduled in advance at different rates of total deployment.

"Because the public cloud hosting company manages overcapacity hardware requirements more efficiently than a private data center with expert staff, security teams, & better wholesale pricing on web server hardware upgrades, these Platform-as-a-Service (PaaS) and Infrastructure-as-a-Service (IaaS) plans are complete data center solutions for complex business organizations from the highest levels of corporate enterprise to SMEs & NGOs."

Companies have a choice of using Kubernetes, Docker, CoreOS, Rancher, OpenShift, Mesosphere DC/OS, CloudStack, Cloud Foundry, or Heroku to compete with AWS EC2, or using a PaaS/IaaS/CaaS plan from Google Cloud, AWS, Microsoft Azure, IBM BlueMix, HP, Oracle, VMware, or another IT major for better bulk-rate "pay as you go" pricing. Rackspace developed and then sold their CloudSites platform to LiquidWeb, exiting the elastic server market for retail hosting entirely for a business model based on consulting. Many specialist web hosting start-up companies are developing innovative elastic web server technology or Kubernetes & container orchestration platform utilities. The elastic web server is a quickly changing technology that develops through continual innovation from all of the major cloud hosting solutions providers, available in the marketplace today at prices targeting every project's budget.

DevOps Enabler Tools

 

SECTION 2

Elastic cloud hosting types

From 1998 to 2008, Google led the advancement of cloud computing by developing internal solutions based on containers, elastic database search, virtualization, microservices, etc. that enabled the operation of web/mobile apps at scale on the world's largest data centers. From these computer science & industry fundamentals, Docker containers, Kubernetes cluster orchestration, Docker Swarm, CoreOS Tectonic (rkt), RancherOS, RHEL OpenShift, OpenStack, CloudStack, & Mesosphere DC/OS all offer competitive functionality for use in public/private/hybrid cloud network orchestration. The other pivotal developments in elastic cloud web server evolution were the release by AWS of the EC2 platform in 2008-2009 and the development by Rackspace of the CloudSites service at around the same time which led to a collaboration with NASA on the OpenStack project. These two services standardized what elastic cluster web servers could do for corporate enterprise (AWS) and small business/independent publishers (Rackspace).

"There are still not many public cloud alternatives to AWS EC2 and CloudSites for elastic, auto-scaling web server products outside of the Container-as-a-Service (CaaS) , OpenStack Linux, Mesosphere DC/OS Marathon, Docker Swarm, CoreOS Tectonic, & Kubernetes hosting plans."

AWS EC2 is by far the most popular elastic cluster server platform in the world and established market dominance when there were no other competing product alternatives for the largest web publishers, software, & media companies. Businesses like Netflix, Verizon, or Visa can host their entire data center capacity on a public cloud host like AWS EC2 or Google Cloud, while specialist web hosting companies like WPengine, Kinsta, Acquia, & Pantheon can also contract with these large IT companies for the hardware required to operate cloud hosting at high volumes of web traffic for managed corporate clients and retail customers. AWS advanced the EC2 elastic web server platform and grew rapidly to become the leading solutions provider in public cloud hosting, where Google has subsequently open-sourced the Kubernetes project code and container orchestration framework. Kubernetes is now the main private cloud or PaaS/IaaS/CaaS alternative to AWS EC2 for large scale organizations, popular blogs, mobile apps, ecommerce websites, & social networks.

"The Rackspace sale of CloudSites to LiquidWeb is an example of elastic cloud server solutions not being viable for the retail hosting market over the last 10 years where the platform could not retain customers effectively vs. larger IT cloud competitors."

Multi-cloud solutions and microservices allow DevOps teams to take advantage of the best IT products, professional services, & web hosting prices across many quality vendors.

DevOps - a living example from the Accenture Open Web Platform

Public cloud

The major public cloud hosting companies are industry leading because of their scale, where the number one corporation, Google, famously managed over 2 billion containers per week in production with "Borg". Social networks like Facebook, Twitter, & Snapchat, media companies like Netflix, and telecoms like Verizon can all build web support operations on public cloud hardware platforms to save money on data center costs. Small businesses and start-up software companies can access the same tools enterprise corporations employ to support the highest rates of user traffic on their web servers or integrate API services to add third-party features to applications.

"Kubernetes, Docker container orchestration, Serverless  or Function-as-a-Service (FaaS) platforms, LiquidWeb CloudSites, OpenStack, & AWS EC2 are considered the leading edge solutions for elastic web server support on public cloud hardware plans, each with different advantages as a hosting solution. IBM, HP, & Oracle compete with the three majors for these services."

The major public cloud service providers like Google, AWS, Microsoft Azure, IBM BlueMix, HP, & Oracle all have extensive suites of database tools, custom development platforms for programming languages, search tools, data analytics, network monitoring, & web security options that are unique to their brands. Some companies like Rackspace, VMware, & Parallels do extensive cloud consulting and product support for enterprise clients using AWS, Google, & Microsoft cloud data center hardware in their business solutions. Public cloud service providers offer the most advanced new platform technology for elastic cluster servers, as well as the resources for even the largest IT companies to scale web/mobile applications using it, but at prices higher than other open source software solutions.

Small businesses, NGOs, education groups, & web publishers all get low cost, introductory services on advanced cloud platform tools to build more innovative apps & IT service products for their local requirements. Public cloud models can also support Software-as-a-Service apps like email, spreadsheets, word processing, billing, accounting, invoicing, CRM, logistics, engineering, graphic design, photography, DTP, law, & medical industry tools that operate on a subscription basis or through licensed SaaS purchasing agreements.

"Companies like AWS, Google, & Microsoft have so many developer and programmer tools on their platforms that hundreds of thousands of technical professionals are being supported with jobs working in the downstream ecosystems."

Serverless or FaaS platforms operate in parallel to web/mobile app code, whereas cloud cluster servers require provision of over-capacity limits to be set for billing and automation, but both include better real-time network data analysis for online businesses.

Mesosphere DC/OS

Private cloud

The main option to a public cloud or PaaS elastic cloud server product is private cloud architecture that is managed by either open source or proprietary-licensed software platform tools. Microsoft Windows, Hyper-V, VMware, & Parallels are best known, while Citrix XenServer, KVM, OpenStack, Docker, Rancher, CoreOS, Mesosphere OS, CloudStack, Kubernetes, OpenVZ/Virtuozzo, & other open source software is used in other solutions. Large companies are working to solve challenges related to supporting complex code for web & mobile applications in production. This means cloud optimizing data center solutions, upgrading to keep current with the latest web standards, implementing advanced web page caching, & CDN integration for high traffic websites, as well as training & recruiting expert staff, managing cloud data center hardware, DevOps teams with CI/CD requirements, complex apps with many registered or simultaneous users, etc.

"OpenStack was developed in coordination between Rackspace & NASA as one of the most advanced cloud data center management frameworks. "

Mesosphere DC/OS (the Datacenter Operating System) builds on Apache Mesos with its own container standard and orchestration engine Marathon for cluster web server management. This solution is used by major industry data centers like eBay, PayPal, & Twitter. Kubernetes is the open source version of Google's "Borg" software and is popular for use with Docker container orchestration. Docker Swarm and CoreOS Tectonic provide various solutions for Kubernetes hosting in private cloud networks and OpenStack Linux distros from Red Hat, Ubuntu, & SUSE are the main private cloud orchestration packages competing in this area of open source virtualization solutions. VMware, Microsoft, & Parallels are the top platform software developers in Windows virtualization solutions. Companies like Rackspace, IBM, HP, Oracle, Cognizant, Infosys, Wipro, Tata Communications, Accenture, & many others provide professional consulting for private cloud construction and data center management to the world's largest organizations in business, governance, education, and other sectors.

NetApp Docker Volume Plugin (nDVP)

Hybrid cloud

Hybrid cloud solutions allow complex organizations to keep some data center operations "in-house" at a local facility while making use of remote hosting, hardware, or API services from IT companies in other departments. In practice, there are almost unlimited ways in which complex organizations can develop hybrid cloud solutions for their business or communication needs, and many of them are already well adopted today. For example, knowledge sharing, business intelligence, data analytics, network monitoring, & CDNs with anti-DDoS & firewall security are all common. Even many of the largest Fortune 500 and government organizations have switched to cloud solutions over the last 10 years internationally. This represents a major change in corporate IT buying patterns, hiring, recruiting, training, etc. with more competition between IT majors and web start-up companies specializing in PaaS/SaaS solutions.

"The major programmers with experience in this field are well compensated and highly recruited, where these developers regularly build entirely new platform solutions, i.e. with containers, Kubernetes, & elastic web server frameworks, the OpenStack ecosystem, CMS scripts, Linux distros, cloud load balancing utilities, etc."

Other third party tools manage disk images, SSL/TLS certificates, WAFs, page caching, database frameworks, programmer developer kits, and version control so that Agile teams can bring complex web/mobile software to market under CI/CD guidelines.

With hybrid cloud constructs, businesses and other organizations can combine products & services across IT vendors to build solutions for their work groups that solves problems according to budgetary restraints. This can involve seeking the best prices on cloud data center hardware in web hosting or choosing a Kubernetes hosting plan from AWS, Google, Azure, etc. The introduction of Kubernetes hosting by public cloud companies only occurred over the past year, whereas previously the main options were AWS EC2 and Mesosphere DC/OS. CoreOS and Red Hat Open Shift offered other options for container orchestration with OpenStack, Docker, rkt, & ContainerLinux.

Companies can also contract Database-as-a-Service (DBaaS) plans from major public cloud providers like AWS, Microsoft Azure, or Google App Engine. CRM software like Salesforce.com deploys in the cloud for better efficiency or cheaper licensing costs vs. desktop software. Businesses of all sizes commonly outsource web hosting services to data centers, where elastic cloud hosting developed over the last 10 years to offer new potential to SaaS start-ups and large enterprise corporations on public cloud hardware. Programmers are only now opening up new types of "big data" applications using Apache Hadoop, Kafka, Ignite, Cassandra, & Mesos as hybrid cloud tools.

Docker Stats since DockerCOn '14

Elastic cloud hosting – cost

The cost of elastic cloud hosting can range from less than $100 per month for SMEs to $200+ million per year for the largest business organizations and media companies. "Pay-as-you-go" billing means you only get charged per second of CPU processing power consumed, or on the basis of exact bandwidth, storage space, RAM allocations, etc. Elastic web servers require over-provisioning of resources for peak time auto-deployment that are used as extra capacity of the public cloud host but then powered down without cost when not needed by web traffic flows. This method of provisioning saves money on large scale data center traffic. For smaller websites, the same technology keeps a website online during web traffic spikes, i.e. special promotions or "slashdotting" events.

"Elastic cloud solutions at the retail level may be available as low as $30-$60 per month through LiquidWeb CloudSites. Under "pay-as-you-go" billing and over-provisioning at AWS, only the exact amount of web server hardware deploys to match a web/mobile app's hosting requirements."

This also provides a higher level of resource provisioning efficiency for process-intensive scripts. Elastic web server solutions bill at 20% to over 1000+% more expensive than other web hosting plans, but no other platforms offer these options to businesses that need high scale web traffic support, where these companies in turn seek to save on hardware operating costs through greater data center outsourcing with cloud web server platform optimization pre-installed and managed remotely.

AWS EC2 sets the market standard for elastic cloud servers at scale, which is important for many web start-up companies, mobile apps, popular blogs, & Fortune 500 companies to operate. AWS also offers managed Kubernetes and elastic container orchestration frameworks as other parts of their enterprise cloud services. AWS has thousands of web hosting, database, and API products for developers to build new apps with professionally, such as voice recognition and AI/ML/DL, although the main problem for companies perceived as being vendor lock-in (i.e. Alexa). The opposite point is that most organizations cannot develop equivalent applications for their software apps at prices that can compete with the API services from AWS, Google, Azure, IBM BlueMix, etc.

"Many of the independent web hosting companies outside of LiquidWeb CloudSites do not offer elastic web server frameworks at all on their platforms. Yet, these are still viewed as the most advanced platforms for cloud data center management at scale, and the determining reason why so many companies choose AWS EC2."

Now the same companies have adopted Kubernetes/Docker solutions due to: better virtualization; smaller OS installs (i.e. Rancher, Windows Nano, SUSE MicroOS, VMware Photon, & CoreOS); cloud orchestration engines for cluster web servers; reverse proxy load balancing on network traffic (NGINX); and improved HTTP/S caching with Varnish Cache. These features allow businesses with high traffic data center requirements to manage elastic web server solutions at lower costs.

Apache Kafka architecture diagram

SECTION 3

Solutions

Elastic cloud hosting solutions appeal to four main market sectors with different needs, pricing, & use case scenarios. These are: Enterprise Corporations, Small Businesses, Ecommerce, & Web/Mobile Applications. In cloud hosting, it is the total volume of web traffic and number of I/O requests that must be taken into consideration; how these requirements are processed on hardware at scale; what the price for hardware with various web server software stack platforms is at the various public cloud hosts; and how businesses can save money across a wide range of sectors and verticals by adopting cloud solutions.

"Most business organizations adopt elastic web server platforms for better performance of web/mobile applications in production for their ecommerce sites, game portals, video streaming, or other industry requirements."

Google, Microsoft, AWS, HP, IBM, etc. compete for business offering cloud data center services to other corporations and government organizations at bulk-rate prices with developer support. SaaS, PaaS, IaaS, & CaaS plans are the best ways for organizations of all sizes to adopt elastic web server container orchestration & automated deployment.

apache spark architecture

Enterprise corporations are doing billions of dollars of business per year and regularly investing hundreds of millions of dollars in IT services. The largest cloud hosting companies compete for Fortune 500 and government contract business, offering these groups savings of up to 40% to 50% by transferring their "in-house" IT departments to the cloud. With web development, hosting, or elastic servers, the savings can be even more.

"After the initial security fears of the major companies were overcome by the performance of public cloud solutions with near equivalency in data security over private cloud constructs, there has been a mass migration of the largest business and government organizations to the cloud model, where most of them are going to AWS, Google, & Microsoft Azure."

Only IBM, HP, & Oracle compete with the top three in cloud development, although Cloudflare & DigitalOcean are leading in CDN/Edge computing sectors vs. major web hosting brands.

Open source solutions and proprietary code compete in enterprise for adoption in cloud hosting and elastic cluster server management. Red Hat Enterprise Linux and SUSE Linux package OpenStack distributions to data centers that need complex cloud orchestration to support cluster servers for web/mobile apps. Mesosphere DC/OS is available hosted at Microsoft Azure Cloud. AWS, Google, & Azure compete with Kubernetes & Docker orchestration solutions for the largest enterprise companies to host their SaaS applications online.

"These 5-star Kubernetes services & elastic container orchestration engines are trusted because they are the same tools and facilities Google, Amazon, Rackspace, Red Hat, & Microsoft utilize for their own internal business operations."

Other web start-ups like Netflix, Uber, or Acquia Cloud host on AWS EC2 & Google Compute Engine to rebrand the hardware with their proprietary software stack compositions for cloud hosting, adding consulting services, mass media publication content, or ecommerce features. Elastic web server platforms allow enterprise corporations to offer the most advanced new software features to users on their web/mobile applications.

FaaS - Function as a Service

Small business

Small businesses have adopted AWS EC2 in major ways and used Google Compute Engine successfully in millions of applications. Millions of developers use Microsoft Cognitive Services or AWS SageMaker & DeepLens options for remote AI/ML/DL processing on EC2 P3 instances where the main competitors are Google's TensorFlow TPUs or IBM's Power 9 units. It is not that elastic cloud and machine learning plans are only useful for the largest enterprise corporations or highly-funded VC startups. The public cloud solutions providers excel at putting powerful developer and programming tools in the small business range for the construction of innovative new web/mobile apps.

'The "pay-as-you-go" approach is designed to be broadly competitive with VPS and dedicated server costs. Elastic web server platforms allow small businesses to scale to larger user traffic rates on any public cloud host without changing standards.'

LiquidWeb's CloudSites platform is the most user-friendly entry path to cloud cluster web server orchestration, which is a migration of original Rackspace customers. AWS and Google Cloud have low cost billing for independent publishers that competes with shared hosting & VPS plans at major brands.

One of the advantages of elastic cluster scaling for small businesses is the ability to schedule more system resources for weekends, holidays, or special promotions. Still many small businesses have overpaid for Rackspace elastic web servers when compared to other shared hosting plans, then migrate to legacy solutions that are more cost-effective.

"It largely depends on a changing web development industry, cloud web hosting company investments, and quality of programming talent what solutions are built for local/international clients, but the availability of open source tools and PaaS products gives small businesses a quicker means to adopt new technological advances as features in products."

SaaS companies like AirBNB, Uber, Groupon, Instagram, & Netflix quickly change the marketplace for goods and services, so web/mobile app startup companies continually challenge sectors for disruption using elastic cloud clusters to scale user traffic across devices like smart-phones, iPads, desktop computers, & tablets. The small business startup, web development firm, contracting agency, independent/freelance programmer, or software development company will be among the first to support elastic web servers.

 

Ecommerce

Ecommerce is the sector where elastic cloud servers are most common, as these applications have more registered users, financial transactions, encrypted connections, product recommendations, etc. that cannot be effectively cached and require large amounts of data center resources to process successfully at scale with high performance standards. Consider just 1 second of difference in page load times over millions of visitors can lead to a loss of 30% to 40% of customers. Consequently, the largest ecommerce websites like Amazon.com, eBay, Salesforce, AliBaba, etc. all need website support for complex software applications and databases with the fastest processing speeds and performance at scale for users, otherwise, their cloud service model will fail with browser lag, user downtime, long downloads, or slow API connection speeds.

'AWS gives developers the chance to build using the same database and analytic software they use to operate Amazon.com. Google offers Kubernetes hosting with "big data" search products that are near-equivalent to their own internal data center platforms.'

Innovative companies can take advantage of these tools to build custom web/mobile apps for their own products & services that offer new software features to clients. Academics can use these tools for technical programming requirements, IT departments, databases, or super-cluster research.

Where Google spent the last 10 years in a "Mobile First!" concentration, the next 10 years will be "AI First!". Google has already integrated AI/ML/DL servers across their business infrastructure, as have all of the largest ecommerce companies, social networks, and mobile software apps. Mostly these are for product and content recommendations based in "big data" search or ML algorithms, but it simply shows how far behind the industry majors most businesses are in their ecommerce platform development. Small businesses have a variety of IT needs, of which elastic cluster cloud servers have become a major service option in the marketplace.

"Millions of businesses already use the major public cloud hosting companies where ecommerce is the most difficult sector to manage in web development, custom programming, and cloud hosting."

Many custom third-party software tools and utilities for web servers and data center management continue to innovate in the sector around Kubernetes, Docker, OpenStack, & Apache Mesos. Big data pipelines enable better search analytics and network monitoring for app developers which can be synched effectively with elastic cloud cluster servers for ecommerce websites with many registered users.

Compuware ISPW

Web/mobile apps

Elastic cloud hosting was made for complex web & mobile applications that need more resources than a single dedicated web server can provide to support social networking communities, ecommerce, games, streaming media, & "big data" search analytics. These companies regularly scale from VPS or dedicated server plans to elastic cloud cluster orchestration as a means to manage operations more efficiently through AWS, Google Cloud, or Microsoft Azure. More than 50% of web traffic by volume of the top 1 million apps goes through the top three cloud company data centers. Many web hosting companies with optimized platform services for WordPress, Drupal, & Joomla host on top of Google Cloud & AWS hardware.

"There are few competitors to the major IT companies at this size of operations. These companies also have the advantage of providing a huge suite of developer tools, database frameworks, programming language support, and elastic server orchestration engines for different sized organizations under public cloud plans. Disadvantages are data privacy related & market centralization."

The public cloud companies are used successfully for programming and web/mobile app development from hobbyists to the largest corporate enterprise data centers. Each concern has different advantages of accessing the tools through the service provider. For APIs or AI/ML/DL integration, there are also few companies that can compete with Google, Microsoft, AWS, & IBM at this time. Small businesses can use the services provided by these majors to add voice recognition, image search, text translation, and AI-driven product or content recommendations into their apps at cheap monthly licensing fees. Academic researchers can use their AI/ML/DL platforms to create supercomputers.

"One of the most characteristic aspects of the public cloud hosts is their ability to scale products and services to projects of all sizes."

Web and mobile apps with custom database and programming language support use these services to meet their requirements with reliability in production. Google's AI server farms run on the TensorFlow platform, including elastic scale for bulk processing requirements in AI/ML/DL. These solutions work in parallel to web hosting platforms.

SECTION 4

Conclusion

The top three choices for elastic cloud cluster web server orchestration are:

  • AWS EC2;
  • Docker & Kubernetes platforms like Docker Swarm, CoreOS Tectonic (rkt), OpenStack, RHEL OpenShift, & Mesosphere DC/OS (Marathon);
  • LiquidWeb CloudSites.

The retail web hosting sector does not widely support elastic cloud clusters in production for individual clients, although they may use them actively behind the scenes to manage their own data center requirements. Elastic cloud cluster servers include better load balancing through the monitoring and analytics performed on web traffic, data streams, user transactions, etc. Developers use elastic cloud servers with automated disk image deployment to support the most complex web/mobile app code in production.

"Together with microservices & APIs, programmers can create a new generation of web/mobile applications on elastic cloud servers, or use them to modernize legacy apps, databases, & data center operations in business organizations."

AWS EC2 has a major market share, where many software startup companies have sought to build similar solutions for elastic cloud server cluster management based around OpenStack, CoreOS Tectonic, CloudStack, Mesosphere DC/OS, and Kubernetes. Docker has emerged as the leading standard for container virtualization; RancherOS/MicroOS/NanoOS installations of Linux/Windows in containers; Kubernetes for cloud orchestration of cluster web servers & load balancing; with LiquidWeb CloudSites the original small business oriented elastic web server hosting platform developed by Rackspace.

Cloud systems administrators in DevOps can manage their own OpenStack or OpenShift/CoreOS/Docker Swarm + Kubernetes installs on hardware using containers on AWS, Google Cloud, or Microsoft Azure to orchestrate cloud web servers in clusters according to user web traffic demands. The public cloud PaaS/IaaS/CaaS tools are easier to manage, while including advanced network diagnostics, browser traffic request reports, and business intelligence feedback analysis utilities that are helpful for ecommerce websites, internet platform security, web/mobile app traffic monitoring, Agile team knowledge sharing, and AI/ML/DL meme recognition in "big data" applications, such as those built with Apache Hadoop, Cassandra, Kafka, Ignite, Mesos, or NGINX/LiteSpeed alternatives.