Machine learning-optimised stacks, container-first architectures and DevOps for both software and hardware are among the key trends redefining enterprise IT infrastructure, according to a new report from McKinsey.

The piece, authored by Arul Elumalai, Kara Sprague, Sid Tandon, and Lareina Yee, looks at what is changing and how companies need to fight back.  

Many of these have frequently been covered by this publication; some, like the public cloud going mainstream, are long overdue. Yet there is an interesting titbit here. Given the long-established leadership of Amazon Web Services (AWS), Microsoft, and Google in public cloud, McKinsey argues that their entrenched dominance will mean only organisations with ‘significant capital investment capabilities’ will be able to compete in future. The article offers Alibaba as a potential suitor; the Chinese firm said last month its cloud business “continued to defy gravity”, while in September Gartner placed the company in third place for public cloud IaaS.

Other predictions which readers will have heard before – but are still invaluable – revolve around cybersecurity and increased usage of open source offerings. Examples of how the latter is gaining traction in the enterprise involves TensorFlow, Google’s machine learning system first launched in 2015. The customer list today is impressive, McKinsey notes, from Airbnb, to eBay, and Qualcomm.

It is the emerging technologies, however, which take the honours. According to McKinsey, B2B applications will account for almost 70% of the value coming from the Internet of Things (IoT) in 10 years’ time; and IoT business applications are now ready for adoption. Elsewhere, the article notes how the new DevOps business model is moving beyond app development to integrate operations and IT infrastructure, while artificial intelligence is ‘delivering benefits to companies across industries.’

“The scale of disruption in the technology infrastructure landscape is unprecedented, creating huge opportunities and risks for industry players and their customers,” the report concludes. “Executives at technology infrastructure companies must drive growth by transforming their portfolios and rethinking their go-to-market strategies.

“They should also build the fundamental capabilities needed for long-term success, including those related to digitisation, analytics, and agile development.”

You can read the full piece here.

Agility is one of the key characteristics that distinguishes a successful managed service provider from the rest of the pack. Being agile means being able to respond quickly to onboard new customers, and fulfill new service requests. Unfortunately, this is often in direct conflict with another important MSP goal: minimizing the number of staff required to deliver those services.

One of the smartest ways to resolve that conflict is to implement your own cloud infrastructure, with customer self-service and automation. With the right approach to cloud, you can ease the burden on your technicians, accelerate onboarding, bring services to customers more quickly, and shrink your time to revenue.

What stops you being agile?

Most MSPs are at a significant disadvantage when it comes to onboarding new customers or provisioning new services.  Traditionally, your infrastructure exists in silos where compute, storage and networking are managed as separate functions. To provision a new customer, you first assess the customer’s needs, and then build the necessary infrastructure from scratch – racking specific servers, firewalls and other systems; buying various software licenses; and layering them to deliver the service.

The customer gets a bespoke solution – eventually – but this traditional siloed approach is not conducive to efficient, sustainable and most importantly repeatable growth for you as an MSP. There are three key reasons why:

Complexity delays revenue: Onboarding new customers or spinning up new resources, in this type of environment, involves careful coordination across infrastructure silos, teams, vendors and technologies. It's time-consuming to design, configure and test services that depend on multiple platforms, multiple UIs, and multiple networking and storage technologies.  Provisioning can take several weeks, delaying time-to-revenue.  

Worse still, it's a process you have to repeat for each customer – and it's difficult to manage when customer needs change. The servers provisioned at the beginning of the year may not be adequate at the end of the year.

People cost money: These problems are compounded by the need to adjust staff levels to minimize salary expenses. Most MSPs size their teams for service maintenance rather than service provisioning, and don’t typically have dedicated teams for bringing up new customers. Technicians must fit that into an already busy work schedule, adding even more time between a customer's order and service delivery.

You have too much or too little hardware: In many cases, IT resources are either under- or over-provisioned. No MSP wants to see equipment sitting idle, but when the alternative is waiting days for new kit to arrive – creating further delays for the client – having a stock of unused hardware may be the lesser of two evils.  

The benefits of a cloud management platform

Building your own cloud infrastructure lets you tackle these issues head on. As well as giving you a platform for private and public cloud service delivery, cloud infrastructure brings much greater agility and efficiency to your operations as an MSP.

Provisioning efficiency: Because cloud provisioning is software-driven, it requires minimal amounts of staff to perform the operation. Rather than racking new devices for new customers, an MSP can carve out a section from existing infrastructure and provision resources on the fly.  To offer public and/or private cloud services, you need a cloud platform with the ability to orchestrate across a range of hypervisors – and to achieve peak efficiency, you also need to be able to manage these services centrally.  By being able to see all physical servers, firewalls, storage and Virtual Servers in one place, it’s easier to react to customer needs and issues as they arise.

Administration efficiency:  A cloud management platform should minimize manual effort at every point in the customer lifecycle. With the right cloud management platform, properly-trained personnel, and some consulting from the cloud infrastructure vendor, one or two technicians should be able to provision a new private cloud in hours rather than weeks.

Vital to this is the need to be able to create permission-based user roles and user groups so that, once the cloud is in production, clients can self-serve resources within a secure framework, minimizing the need to interact with your teams. The cloud management platform should also leverage customer profile templates. Once a template is created for one customer, it can be easily modified to onboard a second customer, and so on. Having a central template repository makes provisioning easier and faster for IT administrators and also reduces provisioning errors.

Billing efficiency: Leveraging a solution that also intricately calculates resources for billing by customer is another element that will save hours of manual work, and improve margins quickly.

Resource efficiency: with the ability to treat the entire compute, network, and storage infrastructure as a flexible pool of resources, MSPs can easily assign specific resources to specific clients and bill for them accordingly. The process becomes a software-based provisioning activity that requires fewer technicians and eliminates custom racking and stacking for individual clients. What’s more, the MSP can replicate one customer’s setup for the next customer, and simply tweak the resource allocations or service mix to suit the new customer.

What the cloud translates to for the MSP is: more efficient use of resources, faster time to revenue for new customers, higher revenue from private cloud services and more fluid resource planning for future needs. Moving to a cloud-based infrastructure not only enables new services, it also simplifies and streamlines provisioning to improve service agility while reducing costs.

Most technology leaders, contemporary and from the hardware era, are reshaping their businesses to do software. They hope to capture value from emerging technologies such as IoT, SDN, and AI. Ultimately, irrespective of the vertical, it is about deriving value from independent software applications participating in an ecosystem as one comprehensive solution. In his session at @ThingsExpo, Kausik Sridhar, founder and CTO of Pulzze Systems, discussed how given the magnitude of today’s application ecosystem, tweaking existing software to stitch various components together leads to sub-optimal solutions. This definitely deserves a re-think, and paves the way for a new breed of lightweight application servers that are micro-services and DevOps ready!

read more

In 2011, Marc Andreessen wrote a thought provoking article in the Wall Street Journal that software is eating the world. Today, in 2017, we can say that cloud transformation is happening all around us and cloud is now indeed eating the world. While Cloud services consumption is becoming prolific within the enterprises, it is often challenging to decipher who is using the cloud, how much is it being used and for what purposes ? Further, is the usage of cloud across the enterprise optimal? While migration to cloud is saving money, not utilizing the cloud resources optimally is nullifying some of the monetary gains. In order to better govern the cloud resources, a top notch cloud analytics engine is the need of the hour.

read more

It was a statement inferred during NVIDIA’s most recent earnings call – and now it has been confirmed.

The company’s Tesla V100 GPU has been chosen by every major cloud provider, with the likes of Alibaba Cloud, Amazon Web Services, Microsoft Azure, Oracle and Tencent all announcing Volta-based cloud services.

Alongside this, the company’s financial results revealed record revenues of $2.64 billion, up 32% from this time last year, with data centre revenue of $501m, more than doubled from a year ago, and growth across all platforms. For the first three quarters of 2018, NVIDIA returned to shareholders $909 million in share repurchases and $250m in cash dividends.

Speaking to analysts following the financial results announcement, Colette Kress, executive vice president and CFO, said NVIDIA’s data centre business had an ‘outstanding’ quarter.

“As we have noted before, Volta delivers 10x the deep learning performance of our Pascal architecture, which has been introduced just a year earlier, far outpacing Moore’s Law,” Kress said, as transcribed by Seeking Alpha. “The V100 is being broadly adopted with every major server OEM and cloud provider. We expect support from V100 from other major cloud providers as well.”

Replying to an analyst question for the next couple of quarters on the data centre side, CEO Jenson Huang said: “This ramp is just the first part of supporting the build out of GPU-accelerated service from our company for data centres all over the world as well as cloud service providers all over the world. The applications for these GPU servers has now grown to many markets.”

Huang added that there were five primary segments for its Tesla GPUs, including high performance computing, deep learning training, inference, and putting it all together in the public cloud.

You can take a look at NVIDIA’s financial results here.

For those of you following the SD-WAN market, it comes as no surprise that VMware has announced its intent to acquire VeloCloud for an undisclosed amount. So how does it affect the SD-WAN market, and what does the future hold?

VMware’s latest acquisition: A case of ‘me too’?

VMware’s acquisition of VeloCloud comes on the heels of Cisco’s Viptela acquisition earlier this year. Don’t, however, dismiss VMware’s purchase as a “me too.”

VMware’s announcement is aligned with its commitment to software defined networking (SDN) as an architectural framework and its NSX offering. VeloCloud extends SDN capabilities to the most prevalent use case, software-defined WAN. Many VMware customers have been asking their sales reps about SD-WAN and the purchase of VeloCloud provides the answer.

Cisco, rightly, views this move as a competitive threat. VMware is moving squarely into its territory. It means that Cisco will need to make good on its Viptela acquisition, doubling down on R&D and on integration efforts. Cisco may go so far as to rationalise its portfolio in SD-WAN, which currently spans four offerings.

Rethinking the network

SD-WAN technology is causing IT and network teams to rethink their entire networks.

“If organisations are to harness the power of digital transformation, it’s time for the network to evolve,” according to Zeus Kerravala, industry analyst. “Resiliency, operational efficiency and agile orchestration capabilities are now table stakes and must be part of the network design. However, this can’t be delivered with legacy networks.”

Traditional WANs relied upon error-prone manual configurations, long lead times and costly MPLS protocols. Next generation SD-WANs decouple the control and data plane in order to centrally manage the WAN using software. SD-WANs help organisations benefit from automation, optimised routing, and lower costs while still leveraging their existing hardware and bandwidth investments.

With this type of flexibility and ability to overlay SD-WAN on existing networks, it’s no wonder network managers are taking a serious look at the technology. Research firm Forrester estimates that 90% percent of network managers are looking to evolve their WAN using a software-defined approach.

Factor in the cloud

Cloud-based applications are further stressing the existing network infrastructure as they become the preferred platform. Companies now run 79% of workloads in cloud, with 41% of workloads in public cloud and 38% in private cloud, according to RightScale’s State of the Cloud 2017.

IT organisations traditionally backhauled cloud traffic from remote/branch sites through a central point of access in hub-spoke configurations. These additional hops cause unacceptable application latency, incremental cost and network congestion. Many IT teams have increased the number of break-out points to the cloud but still rely upon manual configuration and updates.

SD-WAN has proven invaluable to companies with large numbers of users spread across many sites who are accessing cloud-based applications. With SD-WAN, these organisations are now able to centrally orchestrate and manage direct connections from geographically dispersed locations to the cloud. They can define and instantly apply policies that govern security and performance across the network using one management console.

The future of SD-WAN

Expectations are high for SD-WAN. Research firm IDC predicts SD-WAN will be a more than $1 billion market this year and growth at 69% to more than $8 billion by 2021. VeloCloud alone boasts 1,000 customers and 50 service providers.

Like compute and storage, networking will move to software-defined architectures to take advantage of agility and cost savings. SD-WAN is the most compelling use case to date; product maturity is accelerating the p­­­ace of deployments and simplifying ongoing management/orchestration. With SD-WAN, error prone, manual configuration becomes a relic of the past.

SimplePay, a financial and insurance platform based in Australia, is a case in point. In explaining why SimplePay made the move to SD-WAN, Rob Gillan, chief technology officer said, “The (network) systems were working fine for us locally in Australia, but I recognised early on that they couldn’t be easily scaled across the globe or easily managed from a reliability and sustainability perspective. We didn’t want to have to drop IT staff in to every new location.”

SD-WAN wish list

Almost all SD-WAN vendors today provide the ability to leverage any combination of different network connections, define network path selection based on business requirements, and provide direct internet access.

Many vendors are now going beyond these core capabilities and are integrating security, public cloud connectivity, and application performance management into the mix. Some also allow administrators to deploy their products without the need to rip and replace the existing network infrastructure.

Organisations evaluating SD-WAN should consider the following:

Security: Many SD-WAN appliances come with a good firewall, which combined with a VPN, supports most use cases. Confidential traffic can be routed for additional protection through more robust firewalls or through a cloud-based security service, such as Zscaler, as needed. In this case, the SD-WAN will abstract away the complexity of connecting to the cloud security service. Business policy rules can be easily set within the management console to authorise which traffic must go to the cloud-based service and which traffic is either backhauled or sent directly through the Internet.

Public cloud integrations: SD-WAN can be deployed as a gateway inside Microsoft Azure or AWS public clouds to manage the network traffic of a customer’s application as it enters or exits the cloud. In this case, the SD-WAN is deployed as a software instance running on an IaaS stack. SD-WAN fees can be passed through to the cloud provider for simplified, usage-based billing.

Digital experience management: IT teams often feel a loss of control as their reliance on cloud-based applications increases. Uncertainty due to visibility gaps between siloed monitoring tools is a common issue. Integrated digital experience management (DEM) tools can provide visibility from the end-user device across the network and into the application, whether it runs on-premises or in the cloud. DEM gives IT the ability to detect and troubleshoot performance problems before they affect users.

Key takeaways

Here are a few key takeaways as you move from legacy to next generation networks:

• SD-WAN technology is now mainstream; many smaller competitors have been absorbed by established vendors with the wherewithal to invest in R&D and public cloud connectivity.
• Cost savings are a given; agility, performance and security are equally important to end users who have come to expect consistently high levels of service.
• In the near term, SD-WAN technologies should allow customers to adopt gradually while continuing to use existing network equipment and protocols.

The ongoing shift to cloud computing continues across the globe. The worldwide public cloud services market grew 28.6 percent year-over-year in the first half of 2017 (1H17) with revenues totaling $63.2 billion, according to the latest market study by International Data Corporation (IDC).

"Public cloud adoption is accelerating in large part as enterprises recognize that the cloud has become the launchpad for virtually every new IT innovation in the last 24 months — including AI, blockchain, quantum computing and more," said Frank Gens, senior vice president and chief analyst at IDC.

Public cloud market development

While stronger than expected growth was seen across all regions, Asia-Pacific saw the highest regional growth at 38.9 percent and this market now represents 11.5 percent of all public cloud services revenues. Strong public cloud spending in China, which saw 55.6 percent year-over-year growth in the first half of 2017, is a key driver.

Among the three primary segments of public cloud services (SaaS, PaaS and IaaS), the SaaS segment, which holds 68.7 percent of overall market share, was the slowest growing segment with a 22.9 percent year-over-year growth rate.

More CIOs and CTOs now think 'cloud first' when it comes to their IT strategy and software footprint, since the benefits of cloud have been demonstrated in most industries. Many companies have picked the low-hanging fruit, in terms of apps that could be easily moved to the cloud, and are now evaluating the potential migration of their next set of larger strategic systems to a SaaS model.

That said, the smallest segment overall was PaaS, with 13.6 percent of the public cloud services market. However, the PaaS market continues to deliver stronger growth than the other two segments at 50.2 percent year-over-year in 1H17.

According to the IDC assessment, the rapid adoption of container technology in the PaaS segment has given developers additional tools to accelerate application development and deployment that is important in the typical enterprise digital transformation journey.

Outlook for private cloud applications

The IaaS segment represented 17.8 percent of the public cloud services market in 1H17, and continues to exhibit strong year-over-year growth at 38.1 percent. Growing interest from enterprises and continued investments by cloud service providers has resulted in enhancements in the IaaS segment across multiple dimensions.

The recent introduction of on-premises offerings into the market also enable easier hybrid IT models, and reduce the barrier to more cloud service adoption for enterprises. Cumulatively, these are paving the way for the next wave of enterprise application deployments on cloud computing IaaS.

Read more: IDC says global public cloud revenues hit $63 billion – with PaaS quickest growing segment

Recent survey done across top 500 fortune companies shows almost 70% of the CIO have either heard about IAC from their infrastructure head or they are on their way to implement IAC. Yet if you look under the hood while some level of automation has been done, most of the infrastructure is still managed in much tradition/legacy way. So, what is Infrastructure as Code? how do you determine if your IT infrastructure is truly automated?

read more

Nordstrom is transforming the way that they do business and the cloud is the key to enabling speed and hyper personalized customer experiences. In his session at 21st Cloud Expo, Ken Schow, VP of Engineering at Nordstrom, discussed some of the key learnings and common pitfalls of large enterprises moving to the cloud. This includes strategies around choosing a cloud provider(s), architecture, and lessons learned. In addition, he covered some of the best practices for structured team migration and discussed ways to control cloud costs.

read more