The cloud computing world is undergoing a massive transformation as technologies evolve, infrastructure becomes virtualized and more efficient, and applications and services increasingly become the stars of the show. This in turn has had a major impact on the way enterprises do business – in fact GE just announced the company is downsizing from eight to a mere three data centers by 2018, fueled by the advantages of working with AWS as a cloud infrastructure partner.
SysAid (News - Alert) Technologies, a global provider of ITSM, is in a unique position when it comes to observing the impact of cloud computing. Since the company works directly with IT professionals at businesses of all sizes, it has seen firsthand just how disruptive and beneficial the cloud has been in the IT world. We got a chance to speak to Sarah Lahav, CEO of SysAid, about the state of cloud computing and where she anticipates its future lies.
Q: What are some of the trends you’ve observed in the cloud computing world over the past several months - both from the perspective of enterprise adoption and how service providers and data centers are handling infrastructure and services?
A: In my opinion, there are a number of different cloud adoption scenarios out there. Many enterprises and service providers are still bogged down in “brownfield mode,” struggling with how to do both the “old” and the “new” IT, with the same shrinking budgets. At best, they see a partial (not all-in), five-year transition to an “infrastructure and services” landscape that is part on-premise and part cloud. This of course varies with organizational size, legacy, and complexity.
Small new companies, which are unburdened with legacy technology, not only choose cloud first – they are also pushed by their investors to go cloud first. Older companies with what can, at best, be called a history of insufficient investment in IT, such as notable high street banks, continue to struggle to adapt to the new technology landscape, with new entrants to their markets threatening to disrupt the establishment because of the different approach to IT, not just the difference in market offerings.
Inside the traditional corporate data center, new vendors are bringing cloud-like capabilities such as hyper-converged infrastructure. These one-click, simple-to-use systems are full of enterprise features and scalable like the cloud; ultimately giving a cloud-like feel. For example, Microsoft (News - Alert) is partnering with HCI (human-computer interaction) providers to give enterprises an Azure-like experience via their hybrid cloud, and VMware is also doing something similar.
Service providers are generally “having their lunch stolen” by the cloud service providers as the gap they filled between enterprises and cloud is shrinking. They are thus moving models to become MSPs (managed service providers) on top of being pure cloud service providers (for instance Rackspace (News - Alert)). They are not just doing infrastructure-based MSP work, they are also being experts in popular business cloud services, such as Microsoft Office 365 and Google Apps for Work. This works well because as the cloud service providers become more complex they become more “difficult” to use, and this complexity invites an amount of ”lock in” due to the cost of changing providers (and moving from one complex user interface to another).
Q: Mobility, BYOD, big data and analytics and a shift toward applications and services on the user side have all been key technology trends this year. What type of impact have they had on cloud adoption and usage?
A: Mobility has made cloud the default for the users of the most-ubiquitous compute device on the planet: the smartphone. And it’s not just limited to consumers (retail customers), this is also penetrating the enterprise through BYOD and advancements in corporate provisioning policies, and it can put “more oil on troubled waters” when end users with consumer-world expectations meet the friction-filled world of enterprise IT. One of the easiest options for corporate IT, in dealing with these new and raised expectations, is to offer cloud-like services including rebadging/white-labeling public cloud services such as Box or Dropbox (News - Alert) for enterprise storage. This means that “consumers” aren’t just growing their own personal use of cloud, they are also driving up enterprise consumption too.
In my opinion, big data and analytics seem to have a more complex cloud interaction. The “brains” for these systems are definitely easier to access in the cloud though, such as machine learning and Hadoop-type systems, because the cloud provider offers them ready-built, as a service. However, data has “gravity” and thus enterprises can be nervous to push large datasets into the cloud, plus sometimes the sheer physics makes it prohibitive. This is why AWS and other providers will let you use “sneakernet” to physically post tapes and disks to them – it can be so much faster than using the Internet. This has resulted in the emergence of hybrid analytics systems that let cloud systems access on-premise datasets via dedicated lines from AWS and Azure.
Q: On the infrastructure side, we’ve seen increased adoption of SDN and NFV to create more efficient network architectures. Do you think cloud computing has been a driver for these changes?
A: The cloud leaders like AWS and Azure say that they couldn’t manage their networks without SDN and NFV technologies. Azure is also taking its learnings and bringing them out of the web-scale cloud and into the service provider and enterprise domains. In this way, cloud has driven these changes.
However, other purveyors of SDN and NFV, such as VMware and Cisco (News - Alert), have a different plan. They want to put the “brains” into the network, and their products, so that they stay central and strategic to enterprises and their IT. Their goals are similar but appeal to different audiences: Cisco wants to help the network guy to be, or to stay, relevant; VMware wants the server guy to be, or to stay, relevant. So different vendors are pushing SDN and NFV from different perspectives, and new entrants like vArmour, Illuminous, and others are adding value on top of SDN and NFV in a cloud context. So in my opinion, cloud will continue to be central to SDN and NFV, as customers try to bridge their on-premise and cloud systems with consistent network controls.
Q: What types of changes do you think we’ll see in the immediate future for the cloud? Over the next few years?
A: The change in the European Safe Harbor agreement might be a lifeline for regional cloud service providers who have used it as an excuse for customers to pick them over global public cloud providers. However, the leading global cloud players are now coming to most regions, and they offer a service that cannot be matched by the smaller regional players. This capability gap might be fixed by the emergence of cloud federations such as HP Cloud28+ and Cisco Intercloud, but this remains to be seen.
Security will change from being “a reason not to do cloud” to being “the reason to do it.” The leading cloud service providers are excellent at security and they offer the same best-in-class security features to everyone at the same price. This utilitarian approach to security, at a time when corporate security breaches are a daily occurrence, means that it will become a “no brainer” for many companies to take a cloud-first approach because of security. There are also numerous start-ups in this area that add even more security-goodness on top of existing cloud service provider security features. The cost of doing things outside of cloud because of security issues is prohibitive and should be questioned by business executives.
Q: What are some of the advantages of using the cloud for processing, storing and analyzing certain types of workloads?
A: Requirements change and experimentation and agility are the “new black” in IT. In business, it is rare to have just one never-changing workload; it is more normal to have many variations, seasonal adjustments, sales and competitive adjustments. The cloud is so agile, in terms of adding/changing features as well as pay-as-you-go pricing, that it gives the business agility and can even make IT look like a hero to executives relative to the traditional lead times for change.
Just like cloud democratizes security, it also democratizes data processing. In the 1990s it would cost companies millions of dollars, and take months, if not years, to “do” data warehousing. Now, using the cloud, a teenager in their bedroom can do machine learning for pennies.
Cloud storage is cheaper and more flexible than anything that can be done on-premise. It also gives the business confidence that they don’t have to pay upfront for storage or processing that they might never end up needing, as well as the confidence that if they’ve underestimated their needs, they can extend their cloud storage consumption very easily.
Q: Are there any other trends/predictions in the cloud space you would like to discuss?
A: Cloud is almost an extraneous term that isn’t needed anymore. Instead of cloud being the thing that is different, the focus is moving to the things that aren’t cloud and questions are being asked about why they are still needed. In a typical enterprise that is a decade or more old, then there is the aforementioned brownfield situation with the (uncomfortable) coexistence of old and new technologies and methodologies. Gartner calls this bi-modal, and in their view (and therefore commonly in their enterprise clients’ views) there is the steady-state Mode 1 and the agile Mode 2.
Mode 2 is also called “cloud native” and it’s a somewhat idealistic growth area for companies like Pivotal, RedHat, Activestate, and Apprenda and their Platform-as-a-Service (PaaS) systems. The idea is that we all stop doing infrastructure, leaving it to the cloud service providers, and that we instead focus on what’s important to the business: profit and revenue won through agility, and winning market performance driven by a focus on apps and data.
The common mantra is that “all companies are software companies” (matched to the phrase, “software is eating the world”), and this is being taken seriously by companies such as GE, JPMC, and others. There is serious merit in this approach because it represents an evolution of “what next after cloud?” and moving up the stack. These PaaS systems make life easier by having built-in standard capabilities that everyone in IT keeps repeating across applications – such as logging, access, security, etc. PaaS systems also democratize new technologies, such as containers and new architectures like microservices, by doing the heavy lifting for you in a plug-and-play way.
Edited by Kyle Piscioniere