BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Too many enterprises are mired in the weeds of dated technologies and are not benefiting from modern advancements that have the capacity to reduce costs, generate revenue and improve nonfunctional application requirements, such as performance, security and reliability. For organizations interested in undergoing a digital transformation that will revolutionize their information technology department, adopting cloud computing technologies must be the first stop in their digital transformation roadmap.
Unless the need for speed outweighs every other requirement, organizations should be avoiding bare-metal deployments and should be leveraging cloud-based systems. Since the cloud abstracts away the underlying hardware, cloud computing platforms make it possible to perform various operations, such as the deployment of an application to a distributed subsystem, or the creation of a globally accessible data bucket from a single point of administration.
By abstracting away the underlying hardware, cloud computing technology can ensure a distributed deployment across a cluster of any number of machines, so when the application is invoked, clock cycles can be harvested from any idle CPU in the system. The benefit is failover, simplicity and reliability, with the only discernable drawback being a small performance impact caused by the cloud computing software that makes the separation between the application and the underlying hardware a reality.
Virtualization versus cloud computing
The manner in which cloud computing software abstracts away CPUs, memory and data is similar in many ways to the concept of virtualization that revolutionized IT departments around the turn of the century. However, the key difference between cloud computing technologies and virtualization technologies is the fact that distribution and the clustering of hardware devices is a core function of the cloud, which isn't always so when dealing with virtualization.
From an infrastructure point of view, adding new servers to the cloud when demand is predicted to increase, or reallocating servers to other tasks when demand recedes is as simple as connecting or disconnecting a server. Increasing capacity is a fairly simple task from the data center's perspective, while the end user of the cloud computing network is completely oblivious to any underlying hardware changes.
Gil TeneCTO, Azul
Fortunately, placing cloud computing or virtualization on your digital transformation roadmap doesn't have to be an either-or decision. Organizations can leverage the best of both worlds by running virtualized instances in the cloud, which is exactly what happens when an Amazon Web Services (AWS) user spins up an Elastic Compute Cloud instance.
"In practical terms, most of what is being run on Amazon today is run as virtual instances running on the public cloud," said Gil Tene, CTO at Azul. "And, at the end of the day, they end up looking like normal servers running Linux on an x86 somewhere; it just so happens that they run on Amazon, and they do it very efficiently and very elastically."
Drawbacks to the public cloud
When cloud computing first rose to prominence, public clouds like AWS or Microsoft Azure were the only game in town. Unfortunately, the public nature of these systems caused consternation for organizations that saw adopting cloud computing technologies as a smart first step on their digital transformation roadmap.
The problem with the public cloud is the eponymous fact that it is public. This has always created grave concerns with governments and financial institutions that deal with very stringent guidelines in terms of data privacy and process auditing.
Having an off-site third-party like Amazon or Google processing personal data about customers opens up a floodgate of privacy concerns. And auditing becomes an issue, as well. After all, how can an arms-length auditing firm inspect processes and procedures when those processes and procedures take place on an external server farm that might not only reside somewhere else in the country, but potentially in another country?
Fortunately, these issues no longer disqualify the cloud from being part of an organization's digital transformation roadmap, especially given the insurgence of open source cloud computing software and technologies that make it possible for companies to bring the cloud home. Cloud computing systems are no longer the exclusive domain of the big Bay Area behemoths, as private cloud software has emerged that enables companies to leverage the benefits of a cloud-based architecture right inside a local data center.
A number of cloud computing software projects have arisen within the open source community that enable enterprises to create their own in-house cloud computing environments. This type of configuration makes it possible for organizations to overcome security, auditing and privacy concerns, enabling a private cloud, as opposed to a public cloud, to be part of the organization's digital transformation roadmap.
The key benefit to making a private cloud part of the digital transformation roadmap is that it takes computational control away from third parties and returns it to the local data center. Public clouds are not foolproof, as the horrendous February 2017 AWS outage will attest.
"If you're depending on one provider, like Amazon, for storage through something like S3 [Simple Storage Service], you can imagine that something could happen to that company or to that infrastructure that would render that capability unavailable," said Sinclair Schuller, CEO and co-founder of Apprenda. "Cloud is actually a very centralized thing."
When AWS went down, organizations that leveraged this public cloud had no other remedy than to point their fingers at Amazon and say it was their fault, and then wait idly until the problem was fixed, during which companies lost millions of dollars in business.
Private clouds are not foolproof either, but at least when things do go sideways, managers and C-level executives can assign blame to actual employees, give directives to technical personnel and put in procedures to ensure similar issues do not arise in the future. That level of control does not exist with a public cloud, and that lack of control can be very disconcerting for organizations whose entire business depends on the web.
Leveraging a hybrid cloud
Because both public and private clouds have their own sets of benefits and drawbacks, it's not uncommon for organizations to make both part of their digital transformation roadmap. For mission-critical items where auditing or privacy are serious concerns, resources from the local cloud can be used.
For less critical items that might not involve personal data, the public cloud can be leveraged to provide cheaper clock cycles and additional redundancy. It becomes a win-win situation if the additional management required to run a hybrid cloud environment is justified.
To stay relevant, organizations must recognize when it makes sense for new technologies to enter into their digital fold. For those who are interested in staying technically relevant, making cloud computing part of their digital transformation roadmap is an intelligent first step that makes a lot of sense.
Will the Amazon S3 outage permanently hurt cloud computing adoption?
Here are some of the most common security threats your cloud may expose
Can you really achieve bare-metal performance in the cloud?