Join us! We’re assembling a world class panel on the impact of digitalization and cloud on the 21st century enterprise. Join us at Future in Review in Park City Sep 27-30.
Join us! We’re assembling a world class panel on the impact of digitalization and cloud on the 21st century enterprise. Join us at Future in Review in Park City Sep 27-30.
I spent five minutes with Marc last week discussing recent system-wide outages at Delta, Southwest and Salesforce. We talked about the challenges facing enterprises with cobbled infrastructures under increased digitalization pressures. As said in the New Normal blog, we’re likely to see more of these types of outages in the future, until eneterprises invest in more advanced, scalable and secure infrastructures designed for the digital age we’re entering.
At CloudVelox we took a disaster recovery best practices survey more than a year ago and findings suggested that most IT departments didn’t test their secondary sites frequently enough to protect their workloads. Testing frequency is part of the problem. Reslience is another. Both are hard to address in infrastructures of cobble.
Delta, Southwest and Salesforce Infrastructures Mark New IT Reality
From the premier IT infrastructure of Salesforce.com to the cobbled infrastructures of major airlines, this year has been a watershed for system-wide outages. Why would we think that next year will be any different?
Digitalization pressures are escalating as customers and employees expect more convenience. And IT teams are adding new layers of cobble to old ones in an effort to keep up.
A recent article in Bloomberg nails the issue:
The failure of Delta Air Lines Inc.’s worldwide computer network this week spotlights the vulnerability of the information systems sustaining the biggest U.S. carriers, each of which has contended with major disruptions during the last year.
Complex networks cobbled together over the decades need major overhauls requiring significant new investments, said Bob Edwards, a former chief information officer for United Continental Holdings Inc. Recent flaws in computer systems quickly escalated into corporate black eyes that exacted costs in both money and reputation.
“I don’t believe the flight ops, maintenance, passenger service systems, crew and dispatch applications are engineered with the level of redundancy needed,” Edwards, who retired in 2014 under pressure after several service disruptions at United, said by telephone. More disruptions are a near certainty: “Mistakes will happen, devices will malfunction.”
We took a disaster recovery best practices survey more than a year ago and it demonstrated how broken “IT as usual” has become under the pressures of digitalization. IT pros didn’t have the time and/or resources to test their redundant systems in case there were failures, like those recently experienced by Southwest and Delta.
For more background read: DR is Broken but Don’t Blame IT.
The question becomes: Now that layers of cobbles are crumbling under the pressures of digitalization what does IT do next? The answer, ironically for the airline industry and others, is in the clouds. Otherwise, systemwide outages become a new normal.
If you’re interested in understanding what the future of IT will look like forget trying to forecast how many trillions in IT spending will be influenced by the cloud. We already know the number will be big, very big. Instead, leave that exercise to the hardware-bound vendors trying to predict how much time they have left as standalone companies.
A far more profound and meaningful question would address the impact that cloud will have on how enterprises operate, even in the very new future. Last fall I moderated a Future in Review panel on the cloud and new operating models. I managed to find it hidden on YouTube. At about 10min 30sec in we kicked into gear and started talking about how the definition of a company has changed and the increasingly fluid relationship between enterprise and customers.
How your company is defined and its customer relationships are much more important cloud considerations than the impacts of cloud spending on IT. Example: Uber
The term “uberfication” has already been used to describe a deep understanding of convenience, especially from a consumer perspective. Yet that definition is too narrow. Yes, we can watch a taxi automagically appear because Uber has built a transformational big data app that allows it to stay connected with drivers, fares and passengers in real-time. The real significance of Uber is in how it has redefined the taxi industry and established remarkably fluid relationships with customers.
Uber’s software, data and relationship with customers proved to be more valuable than ownership of fleets of cars and big yellow pages ads. The same could be said for a host of new app-centric firms including Airbnb.
In the last few minutes of the Future in Review cloud panel Azure’s Staten mentions a dairy farm in Israel using a big data app to track cows with bands (instead of bells) to boost milk productivity. The dairy farm is defined by its ability to optimize its cows for production (and reproduction). In Staten’s example the farm is using advanced supply chain capabilities to obtain competitive advantage.
The Israeli dairy farm didn’t need racks and stacks of servers. It used the cloud. Instead of rack maintenance it focused its resources on farming, applications and services.
Cisco’s Rajendran spoke about a case study recently presented at an industry tradeshow. Philips Healthcare used the cloud to manage unprecedented growth in patient medical records. Their infrastructure before the cloud could not scale to support the needs of their customers. So rather than give up and settle for business as usual they grew with their customer’s demands and improved patient care.
The successful enterprise of the 21st century will have a powerful application or service that differentiates it from those with massive, increasingly complex infrastructures requiring escalating maintenance costs. They’ll be agile and innovative at levels most organizations today don’t understand. And they’ll be able to scale with demand.
So the next time you see a massive cloud adoption statistic, think less about the spend trend and instead of how your team can shift to focusing on innovation and transformation versus traditional information and technology (a tip of the hat to Staten’s comment).
The cloud is about the unprecedented scale and distribution of computing power. It will create a vast new wave of wealth, market caps and business models.
Cloud, digitalization and mobile are changing the game by accelerating the pace of change, and tech vendors tied to legacy premise hardware will likely face years of unprecedented margin and growth pressures. Some will likely not survive in their current form while others will make the shift.
This announcement was pivotal because it defied conventional cloud wisdom. Most analysts described SaaS as the ultimate evolution from premise-bound computing, with IaaS as a necessary middle step for many existing apps. That all changed when leading SaaS player Salesforce.com announced that they would move onto AWS, mere days after a major Salesforce outage. Apparently the move to AWS caused some consternation within Salesforce’s high profile IT department.
==> A JP Morgan report in April cited the cloud as a strategic consideration for the future of dozens of tech companies
The report was only made available to JP Morgan’s clients, but you can read about it here and here. I put the report within the greater context of digitalization and cloud in: “Who will be crushed by Digitalization and Cloud?”
==> An Oppenheimer Report called AWS a massive efficiency flywheel for IT
Larry Dignan at ZDnet put the report in perspective here: “Amazon Web Services is likely to have 1.3 million servers that are more than three times more efficient than enterprise systems, data centers that use space better and generate better returns of about 20 percent.”
Since January I’ve published a pair of articles at Seeking Alpha on what I consider to be early signs of a massive transformation of enterprise IT. Who will be crushed by Digitalization and the Cloud and Two Recent M&A Moves both addressed the new CIO chasm of digitalization and cloud.
Employees, partners, customers, etc. are accessing that infrastructure through endpoints that have as much software/functionality loaded into them as the early software stores. This represents an unprecedented level of scale, complexity and criticality, even for the largest of IT shops.
The Internet of Things as it’s called is becoming a misnomer. It’s really an Internet of Applications. That may seem like a trivial distinction, but it makes all the difference in how one frames the disruption that has already destroyed the growth projections of most hardware vendors. You’ve seen the forecasts for infrastructure growth. Mark Thiele, for example, recently predicted that the world will need 400 million servers by 2020. There is also an explosion in endpoints as the Internet of Things (IoT) redefines the meaning of “end-user” in the age of machine-to-machine communications. Thiele cited predictions ranging from 20 billion (Gartner) to 50 billion (Cisco) endpoints in the near future.
A recent survey found that the average Android user has 95 apps installed on their smartphone, supporting about 100 daily interactions via an average of 35 apps. Yet the explosion in apps has not translated into enhanced growth and margins for many of the hardware vendors. Software is continuing to drive hardware.
The growth in software that enables enhanced usability, agility, scalability, availability, manageability and security. Commoditization of the “things” is already underway, as software and apps become as strategic to endpoints as they became to PCs.
We’ve Seen This Before
The branch office boom and the enterprise web had a similar dynamic. They created pressures for more application layer (referencing the OSI layers, especially 4-7) innovation to support increasingly diverse, specialized networking challenges. Companies like F5 and Riverbed focused on addressing these demands and companies like Cisco, Juniper and Citrix augmented their capabilities with the acquisitions of FineGround, Redline Networks/Peribit and NetScaler, respectively. The server load balancing market cratered as the demands shifted toward traffic management first, and then evolved toward secure application delivery.
Application delivery challenges today are more complex and exponentially more critical than they were ten years ago, when digitalization and smartphones were in their infancy (in terms of numbers shipped and, more importantly, installed applications). Today, there are more apps in a single smartphone interacting 24/7 with an enterprise server and other apps than there were interacting from a branch office to HQ during business hours a decade ago.
The apps trigger the orders and the queries. They determine who is ordering what from whom. They update key supply chain and analytics.
Don’t Forget the Cloud
In terms of app and service delivery the cloud becomes more than a necessity; it becomes an eventuality. Enterprises cannot cost effectively manage the 24/7 demands, including scale and complexity of this new world of digitalization. Even Salesforce, perhaps one of the most advanced and well-managed SaaS offerings has capitulated and switched from their own cloud to a public IaaS cloud.
The cloud will drive an even larger app delivery disruption than the branch office and enterprise web boom because the stakes are even higher and the complexities greater. Instead of another appliance (a cloud front end – lol) it will be software: cloud automation and orchestration software. This emerging category will include cloud management and optimization, cloud recovery and Dev/Test, cloud migration, cloud security and cloud discovery solutions. These technologies will drive IT into new levels of resilience and productivity as these new digitalization and cloud demands emerge.
We will likely see the emergence of innovative cloud “brokerages” that will help CIOs address these exploding complexities. New approaches will be needed in much the same way as exploding volume and complexity forced the financial services industries to evolve from 1990’s era infrastructures, applications, policies, and governance structures. Comparing Wall Street’s transaction volumes of two decades ago against what it does today, you can see that digitalization and cloud will have an even greater impact on a broader range of corporate operating models.
Enterprises are still trying to tackle these changes with yesterday’s tools. For example, they continue to throw scripts and manual processes at the problems they are facing. The challenges will mushroom as enterprises try to fulfill CIO directives without automation. See The Trillion Dollar Cloud Question. And that is just the beginning.
The Resilient Enterprise
Gartner in 2015 determined that 80% of outages are caused by either operation errors or app failures. The remaining 20% included hardware, OS, security, power and disaster events. Put that within the context of this greater mesh of the Internet of Applications (robust endpoints with dozens of apps interacting more frequently, including micro-transactions).
CIOs will have to arbitrate the growing conflict between static, manually scripted actions needed to address increasingly dynamic demands within ever complex app environments in hybrid clouds. Software is the best answer.
It’s no wonder that Gartner is already tracking business resilience and increased spending, especially on second and third tier apps. The first critical mission for the cloud for existing apps, therefore, may be to automate manual processes related to resilience, as IT today already cannot keep up with accelerating demands. See, for example, DR is Broken, Just Don’t Blame IT.
This also extends to security. Enterprise security teams must shift from further futile investments in legacy prevention technologies to more real-time, software-centric automated behavioral analytics and incident response technologies for greater resilience to recover from attacks they cannot stop.
It is natural for traditional IT to think initially of scripts and manual processes to tackle the problem. Yet as delays and issues manifest themselves automation will become a necessity. The alternative is CIO blindness during perhaps the greatest transformation in IT since the advent of the network. Investing in key capabilities will also help CIOs take their seat at the strategy table by finally aligning with and enabling the business rather than being a mere cost center.
A special congratulations to the Exar team for accomplishing what many thought impossible. They migrated their Oracle stack onto AWS in days. Read more at InformationWeek:
The provisioning of servers needed by business users was speeded up by a factor of 4-5X and the speed of projects relying on the Oracle applications was also faster, he concluded.
“IT will look like a profit center instead of a cost center,” he claimed.
To do this in a short timeframe, he turned to a third-party knowledgeable in the migration of legacy systems, CloudVelox. The firm was able to “migrate us very fast, over two days,” he said. The move included the movement of 2TB of data between Fremont and the AWS West facility in San Jose. That transfer was accomplished over the high speed networking available to companies in the Silicon, not by trucking disks, Siljeg told InformationWeek in an interview.
A recent JP Morgan report predicted a massive uptake in enterprise cloud adoption in the next five years based on a recent CIO survey. Within a few weeks a May 2016 Forrester report indicated that migration into the public cloud involves considerable manual processes:
By Forrester’s estimate, the cost of public cloud is relatively small compared to the much larger cost of labor involved, which accounts for over 50% of total migration costs. As detailed in the brief, “[L]abor costs dwarf infrastructure and platform services costs in most of the migration projects we’ve reviewed.”
Adding more icing on the cloud migration cake, Salesforce recently announced that it is moving to AWS. In addition to enterprises moving workloads to the cloud, AWS has landed one of the leading SaaS players. That’s a big deal.
For years conventional wisdom has advised that SaaS represents a progression of app delivery from IaaS. That is, workloads would eventually move from IaaS to SaaS delivery. Now it looks like SaaS providers will be using IaaS to deliver SaaS. This has huge implications for the growth of IaaS and the cloud migration market.
While it may not make much difference for SaaS customers (other than improved resilience perhaps- sorry, I couldn’t resist), this has considerable implications for AWS and Azure and a host of secondary IaaS players.
IaaS gets closer to looking like the future pillar of IT, which poses a question: Are there enough IT pros in the world to migrate the workloads that CIOs in the US want to migrate in the next five years?
This deep conflict between CIO demands and trained labor supply casts the spotlight on recent news about Exar Semiconductor’s automated cloud migration project as demand for automated cloud migration escalates:
It turns out migration wasn’t so painful after all. “We were surprised at how CloudVelox was able to mimic what we have on premise in the cloud,” Seljig said. “They figured it out in one day.”
Things should get interesting. Very interesting. The AWS partner ecosystem has been dominated by manual process body shops. That has to change if CIOs will be able to accomplish what they want, especially in an era of careful IT spending.