The public cloud has had a major impact on how organisations of all sizes receive IT services. Some organisations use private clouds, either on-premises or outsourced to a service provider. But with public clouds enjoying widespread adoption, when would you use a private cloud, and are there any advantages in doing so? Can the private cloud keep up with the public cloud providers to offer the kinds of services, security and scale you need? Let’s take a look at cloud evolution and considerations for the future.
The first generation of public cloud
Elasticity and self-service were huge benefits of the first-generation public clouds, which emerged in 2006. Developers could spin up a new application by themselves, quickly and with very little expense on the public cloud. If the application drew more demand, they could scale up the service automatically. The first set of workloads developed for the public cloud were consumer applications such as Netflix and Dropbox.
It’s difficult to achieve elasticity and self-service with on-premises data centres. That’s because developers have to submit proposals for compute resources with plans for expansion. IT has to allocate capacity from existing infrastructure or requisition new infrastructure, which means large capital expenditures and up-front costs. The lead times to fulfil requests can often be months. Naturally, these challenges have led to more developers adopting on-demand public cloud services, because scalability and speed mean a greater chance of application success.
The second generation of public cloud
Fear made enterprise customers hesitant to migrate to the first-generation public cloud. The biggest fears were related to security and data sovereignty. Over time, though, enterprise customers who wanted the benefits of lower capital expenses, faster time to value and greater business agility started looking at the public cloud more carefully. With so many years of experience with rights management, governance and systems monitoring, public cloud providers were able to largely alleviate security concerns by 2011.
Organisations have come to realise the sophistication of the security processes, personnel and monitoring tools in place at public cloud providers. As security capabilities have continued to evolve, most CIOs now agree that the public cloud providers can do a better job securing corporate information assets than in-house resources can, and that realisation really marked the beginning of rapid growth in enterprise adoption of public clouds.
During this time many infrastructure vendors attempted to provide self-service, metering and elastic capabilities with tools such as OpenStack and CloudStack. It proved difficult, however, for on-premises, private cloud infrastructure to match the pricing and the vast capabilities of public clouds offering hundreds of services.
An impressive set of cloud services also became popular at this time. They included turnkey, multi-tenant services built for generation 1 and 2 public clouds, such as Azure SQL Database and Amazon EMR. While enterprise interest in the notion of “information technology as a service” accelerated, interest in private clouds marched on too. But many savvy CIOs began to realise that running their own data centres or even outsourcing them to tier 2 service providers would not and could not take their businesses where they needed to go.
The third generation emerges
The third generation of public clouds really emerged in 2016. As even the largest enterprise customers struggled to run two or three data centres, consensus developed around the idea that running an on-premises data centre was not strategic to most companies’ evolution and growth. Not only are data centres connected with expensive 10-100Mbps leased lines, but the personnel required to “keep the lights on” is costly and prevents more strategic use of those valuable IT resources.
Consequently, CIOs began taking a hard look not only at the overall cost involved in managing all that infrastructure, but perhaps more importantly, they saw how on-premises infrastructure actually limits corporate growth by inhibiting agility. Having done this due diligence, most IT leaders began taking their organisations down a different path, and migration to the public cloud accelerated faster than anyone really expected.
Making the case for continuing to maintain on-premises infrastructure becomes pretty tough when you compare the capabilities of an on-premises data centre to Microsoft Azure, for example. With Azure, every organisation has instant access to 50-plus regions around the world, all connected with a 30-40Gbps internet backbone. How is that kind of scale ever going to happen on-premises? We’re talking about a globally distributed architecture that has never existed before, and it offers customers exciting, new opportunities. A private cloud just can’t keep up.
Cloud solutions at planet scale
The first generation of cloud solutions such as Azure SQL database and legacy on-premises solutions such as Oracle database did not take advantage of a globally distributed architecture. That’s why they can’t serve customers well ultimately. They are stuck in the world of old architectures that can’t easily scale and therefore can’t keep up with the third-generation public cloud solutions that are based on modern architectures.
There are solutions available today that leverage this modern, globally distributed architecture. One of the first is the new database service from Microsoft called Azure CosmosDB. It is a planet-scale, globally distributed multi-modal database that features low latency and the ability to scale globally. This is what enterprises need to achieve agility – speed and scalability!
VDI at planet scale
The first generation of VDI solutions, much like the old Oracle database, were on-premises. They were complex to deploy and troubleshoot, expensive to buy and maintain, and delivered disappointing and productivity-limiting performance to users. Equally as important is that these solutions devoured IT resources and stymied corporate growth. Second generation VDI solutions were better but these solutions did not take advantage of the globally distributed architecture of the public cloud.
The third generation of VDI solutions were built from the ground up to be multi-tenant, turnkey and planet-scale. Today, a third-gen VDI customer can deploy virtual desktops and applications in any cloud region in the world, in minutes, and then scale it almost limitlessly. This is a brand-new set of capabilities, uniquely enabled by a globally distributed architecture, that allow organisations to achieve levels of agility that have never been possible. The door is now wide open to exciting new opportunities for enterprise growth.
Amitabh Sinha, co-founder, Workspot
Image Credit: Everything Possible / Shutterstock