Looking at Microservices, Containers, and Kubernetes with 2020 Vision

Some years are easy to predict than others. Stability in a market makes tracing the trend line much easier. 2020 looks to be that kind of year for the migration to microservices: stable with steady progression toward mainstream acceptance.

There is little doubt that IT organizations are moving toward microservices architectures. Microservices, which deconstruct applications into many small parts, removes much of the friction that is common in n-Tier applications when it comes to development velocity. The added resiliency and scalability of microservices in a distributed system are also highly desirable. These attributes promote better business agility, allowing IT to respond to business needs more quickly and with less disruption while helping to ensure that customers have the best experience possible.

Little in this upcoming year seems disruptive or radical; That big changes have already occurred. Instead, this is a year for building out and consolidating; Moving past the “what” and “why” and into the “how” and “do”.

Kubernetes will be top of mind to IT in the coming year. From its roots as a humble container orchestrator – one of many in the market – Kubernetes has evolved into a platform for deploying microservices into container clusters. There is more work to do with Kubernetes, especially to help autoscale clusters, but it is now a solid base on which to build modern applications.

No one should delude themselves into thinking that microservices, containers, and Kubernetes are mainstream yet. The vast majority of applications are still based on n-Tier design deployed to VMs. That’s fine for a lot of applications but businesses know that it’s not enough going forward. We’ve already seen more traditional companies begin to adopt microservices for at least some portion of their applications. This trend will accelerate in the upcoming year. At some point, microservices and containers will become the default architecture for enterprise applications. That’s a few years from now but we’ve already on the path.

From a vendor perspective, all the biggest companies are now in the Kubernetes market with at least a plain vanilla Kubernetes offering. This includes HPE and Cisco in addition to the companies that have been selling Kubernetes all along, especially IBM/Red Hat, Canonical, Google, AWS, VMWare/Pivotal, and Microsoft. The trick for these companies will be to add enough unique value that their offerings don’t appear generic. Leveraging traditional strengths, such as storage for HPE, networking for Cisco, and Java for Red Hat and VMWare/Pivotal, are the key to standing out in the market.

The entry of the giants in the Kubernetes space will pose challenges to the smaller vendors such as Mirantis and Rancher. With more than 30 Kubernetes vendors in the market, consolidation and loss is inevitable. There’s plenty of value in the smaller firms but it will be too easy for them to get trampled underfoot.

Expect M&A activity in the Kubernetes space as bigger companies acquihire or round out their portfolios. Kubernetes is now a big vendor market and the market dynamics favor them.

If there is a big danger sign on the horizon, it’s those traditional n-Tier applications that are still in production. At some point, IT will get around to thinking beyond the shiny new greenfield applications and want to migrate the older ones. Since these apps are based on radically different architectures, that won’t be easy. There just aren’t the tools to do this migration well. In short, it’s going to be a lot of work. It’s a hard sell to say that the only choices are either expensive migration projects (on top of all that digital transformation money that’s already been spent) or continuing to support and update applications that no longer meet business needs. Replatforming, or deploying the old parts to the new container platform, will provide less ROI and less value overall. The industry will need another solution.

This may be an opportunity to use all that fancy AI technology that vendors have been investing in to create software to break down an old app into a container cluster. In any event, the migration issue will be a drag on the market in 2020 as IT waits for solutions to a nearly intractable problem.

2020 is the year of the microservice architecture.

Even if that seems too dramatic, it’s not unreasonable to expect that there will be significant growth and acceleration in the deployment of Kubernetes-based microservices applications. The market has already begun the process of maturation as it adapts to the needs of larger, mainstream, corporations with more stringent requirements. The smart move is to follow that trend line.

From #KubeCon, Three Things Happening with the Kubernetes Market

This year’s KubeCon+CloudNativeCon was, to say the least, an experience. Normally sunny San Diego treated conference-goers to torrential downpours. The unusual weather turned the block party event into a bit of a sog. My shoes are still drying out. The record crowds – this year’s attendance was 12,000 up from last year’s 8000 in Seattle – made navigating the show floor a challenge for many attendees.

Despite the weather and the crowds, this was an exciting KubeCon+CloudNativeCon. On display was the maturation of the Kubernetes and container market. Both the technology and the best practices discussions were less about “what is Kubernetes” and, instead more about “how does this fit into my architecture?” and “how enterprise-ready is this stuff?” This shift from the “what” to the “how” is a sign that Kubernetes is heading quickly to the mainstream. There are other indicators at Kubecon+CloudNativeCon that, to me, show Kubernetes maturing into a real enterprise technology.

First, the makeup of the Kubernetes community is clearly changing. Two years ago, almost every company at KubeCon+CloudNativeCon was some form of digital forward company like Lyft or cloud technology vendor such as Google or Red Hat. Now, there are many more traditional companies on both the IT and vendor side. Vendors such as HPE, Oracle, Intel, and Microsoft, mainstays of technology for the past 30 years, are here in force. Industries like telecommunications (drawn by the promise of edge computing), finance, manufacturing, and retail are much more visible than they were just a short time ago. While microservices and Kubernetes are not yet as widely deployed as more traditional n-Tier architectures and classic middleware, the mainstream is clearly interested.

Another indicator of the changes in the Kubernetes space is the prominence of security in the community. Not only are there more vendors than ever, but we are seeing more keynote time given to security practices. Security is, of course, a major component of making Kubernetes enterprise-ready. Without solid security practices and technology, Kubernetes will never be acceptable to a broad swatch of large to mid-sized businesses. That said, there is still so much more that needs to be done with Kubernetes security. The good news is that the community is working on it.

Finally, there is clearly more attention being paid to operating Kubernetes in a production environment. That’s most evident in the proliferation of tracing and logging technology, from both new and older companies, that were on display on the show floor and mainstage. Policy management was also an important area of discussion at the conference. These are all examples of the type of infrastructure that Operations teams will need to manage Kubernetes at scale and a sign that the community is thinking seriously about what happens after deployment.

It certainly helps that a lot of basic issues with Kubernetes have been solved but there is still more work to do. There are difficult challenges that need attention. How to migrate existing stateful apps originally written in Java and based on n-Tier architectures is still mostly an open question. Storage is another area that needs more innovation, though there’s serious work underway in that space. Despite the need for continued work, the progress seen at KubeCon+CloudNativeCon NA 2019 point to future where Kubernetes is a major platform for enterprise applications.  2020 will be another pivotal year for Kubernetes, containers, and microservices architectures. It may even be the year of mainstream adoption. We’ll be watching.

Understanding Microsoft’s Investment in OpenAI

On July 22, Microsoft announced a $1 billion investment in OpenAI, a lab focused on “artificial general intelligence,” or the goal of creating artificial intelligence with human-like observation and learning capabilities. With this announcement,  Microsoft becomes the “exclusive” cloud computing provider for OpenAI and will have access to productizing OpenAI capabilities as they come to market.

Key Takeaways: Microsoft makes a long-term investment in “general intelligence” to start on the next generation of AIs that will be coming to market in five-to-ten years and will be able to recoup some costs back as OpenAI’s cloud provider and monetizer of OpenAI technologies.

Continue reading “Understanding Microsoft’s Investment in OpenAI”

Data Science and Machine Learning News Roundup, May 2019

On a monthly basis, I will be rounding up key news associated with the Data Science Platforms space for Amalgam Insights. Companies covered will include: Alteryx, Amazon, Anaconda, Cambridge Semantics, Cloudera, Databricks, Dataiku, DataRobot, Datawatch, Domino, Elastic, Google, H2O.ai, IBM, Immuta, Informatica, KNIME, MathWorks, Microsoft, Oracle, Paxata, RapidMiner, SAP, SAS, Tableau, Talend, Teradata, TIBCO, Trifacta, TROVE.

Domino Data Lab Champions Expert Data Scientists While Outpacing Walled-Garden Data Science Platforms

Domino announced key updates to its data science platform at Rev 2, its annual data science leader summit. For data science managers, the new Control Center provides information on what an organization’s data science team members are doing, helping managers address any blocking issues and prioritize projects appropriately. The Experiment Manager’s new Activity Feed supplies data scientists with better organizational and tracking capabilities on their experiments. The Compute Grid and Compute Engine, built on Kubernetes, will make it easier for IT teams to install and administer Domino, even in complex hybrid cloud environments. Finally, the beta Domino Community Forum will allow Domino users to share best practices with each other, as well as submit feature requests and feedback to Domino directly. With governance becoming a top priority across data science practices, Domino’s platform improvements around monitoring and making experiments repeatable will make this important ability easier for its users.

Informatica Unveils AI-Powered Product Innovations and Strengthens Industry Partnerships at Informatica World 2019

At Informatica World, Informatica publicized a number of key partnerships, both new and enhanced. Most of these partnerships involve additional support for cloud services. This includes storage, both data warehouses (Amazon Redshift) and data lakes (Azure, Databricks). Informatica also announced a new Tableau Dashboard Extension that enables Informatica Enterprise Data Catalog from within the Tableau platform. Finally, Informatica and Google Cloud are broadening their existing partnership by making Intelligent Cloud Services available on Google Cloud Platform, and providing increased support for Google BigQuery and Google Cloud Dataproc within Informatica. Amalgam Insights attended Informatica World and provides a deeper assessment of Informatica’s partnerships, as well as CLAIRE-ity on Informatica’s AI initiatives.

Microsoft delivers new advancements in Azure from cloud to edge ahead of Microsoft Build conference

Microsoft announced a number of new Azure Machine Learning and Azure AI capabilities. Azure Machine Learning has been integrated with Azure DevOps to provide “MLOps” capabilities that enable reproducibility, auditability, and automation of the full machine learning lifecycle. This marks a notable increase in making the machine learning model process more governable and compliant with regulatory needs. Azure Machine Learning also has a new visual drag-and-drop interface to facilitate codeless machine learning model creation, making the process of building machine learning models more user-friendly. On the Azure AI side, Azure Cognitive Services launched Personalizer, which provides users with specific recommendations to inform their decision-making process. Personalizer is part of the new “Decisions” category within Azure Cognitive Services; other Decisions services include Content Moderator, an API to assist in moderation and reviewing of text, images, and videos; and Anomaly Detector, an API that ingests time-series data and chooses an appropriate anomaly detection model for that data. Finally, Microsoft added a “cognitive search” capability to Azure Search, which allows customers to apply Cognitive Services algorithms to search results of their structured and unstructured content.

Microsoft and General Assembly launch partnership to close the global AI skills gap

Microsoft also announced a partnership with General Assembly to address the dearth of qualified data workers, with the goal of training 15,000 workers by 2022 for various artificial intelligence and machine learning roles. The two companies will found an AI Standards Board to create standards and credentials for artificial intelligence skills. In addition, Microsoft and General Assembly will develop scalable training solutions for Microsoft customers, and establish an AI Talent network to connect qualified candidates to AI jobs. This continues the trend of major enterprises building internal training programs to bridge the data skills gap.

Data Science and Machine Learning News Roundup, April 2019

On a monthly basis, I will be rounding up key news associated with the Data Science Platforms space for Amalgam Insights. Companies covered will include: Alteryx, Amazon, Anaconda, Cambridge Semantics, Cloudera, Databricks, Dataiku, DataRobot, Datawatch, Domino, Elastic, Google, H2O.ai, IBM, Immuta, Informatica, KNIME, MathWorks, Microsoft, Oracle, Paxata, RapidMiner, SAP, SAS, Tableau, Talend, Teradata, TIBCO, Trifacta, TROVE.

Alteryx Acquires ClearStory Data to Accelerate Innovation in Data Science and Analytics

Alteryx acquired ClearStory Data, an analytics solution for complex and unstructured data with a focus on automating Big Data profiling, discovery, and data modeling.  This acquisition reflects Alteryx’s interest in expanding its native capabilities to include more in-house data visualization tools. ClearStory Data provides a visual focus on data prep, blending, and dashboarding with their Interactive Storyboards that partners with Alteryx’s ongoing augmentation of internal visualization capabilities throughout the workflow such as Visualytics.

Dataiku Announces the Release of Dataiku Lite Edition

Dataiku released two new versions of its machine learning platform, Dataiku Free and Dataiku Lite, targeted towards small and medium businesses. Dataiku Free will allow teams of up to three users to work together simultaneously; it is available both on-prem and on AWS and Azure. Dataiku Lite will provide support for Hadoop and job scheduling beyond the capabilities of Dataiku Free. Since Dataiku already partners with over 1000 small and medium businesses, creating versions of its existing platform more financially accessible to such organizations lowers a significant barrier to entry, and grooms smaller companies to grow their nascent data science practices within the Dataiku family.

DataRobot Celebrates One Billion Models Built on Its Cloud Platform

DataRobot announced that as of mid-April, its customers had built one billion models on its automatic machine learning program. Vice President of Product Management Phil Gurbacki noted that DataRobot customers build more than 2.5 million models per day. Given that the majority of models created are never successfully deployed – a common theme cited this month at both Enterprise Data World and at last week’s Open Data Science Conference – it seems likely that DataRobot customers don’t currently have one billion models operationalized. If the percentage of deployed models is significantly higher than the norm, though, this would certainly boost DataRobot in potential customers’ eyes, and serve to further legitimize AutoML software solutions as plausible options.

Microsoft, SAS, TIBCO Continue Investments in AI and Data Skills Training

Microsoft announced a new partnership with OpenClassrooms to train students for the AI job marketplace via online coursework and projects. Given an estimate that projects 30% of AI and data jobs will go unfilled by 2022, OpenClassrooms’ recruiting 1000 promising candidates seems like just the beginning of a much-needed effort to address the skills gap.

SAS provided more details on the AI education initiatives they announced last month. First, they launched SAS Viya for Learners, which will allow academic institutions to access SAS AI and machine learning tools for free. A new SAS machine learning course and two new Coursera courses will provide access to SAS Viya for Learners to those wanting to learn AI skills without being affiliated with a traditional academic institution. SAS also expanded on the new certifications they plan to offer: three SAS specialist certifications in machine learning, natural language and computer vision, and forecasting and optimization. Classroom and online options for pursuing both of these certifications will be available.

Meanwhile, TIBCO continued expanding its partnerships with educational institutions in Asia to broaden analytics knowledge in the region. Most recently, it has augmented its existing partnership with Singapore Polytechnic to train 1000 students in analytics and IoT skillsets by 2020. Other analytics education partnerships TIBCO has announced in the last year include Yuan Ze University in Taiwan, Asia Pacific University of Technology and Innovation in Malaysia, and BINUS University in Indonesia.

The big picture: existing data science degree programs and machine learning and AI bootcamps are not providing a large enough volume of highly-skilled job candidates quickly enough to fill many of these data-centric positions. Expect to hear more about additional educational efforts forthcoming from data science, machine learning, and AI vendors.

Tom Petrocelli Clarifies How Cloud Foundry and Kubernetes Provide Different Paths to Microservices

DevOps Research Fellow Tom Petrocelli has just published a new report describing the roles that Cloud Foundry Application Runtime and Kubernetes play in supporting microservices. This report explores when each solution is appropriate and provides a set of vendors that provide resources and solutions to support the development of these open source projects.

Organizations and Vendors mentioned include: Cloud Foundry Foundation, Cloud Native Computing Foundation, Pivotal, IBM, Suse, Atos, Red Hat, Canonical, Rancher, Mesosphere, Heptio, Google, Amazon, Oracle, and Microsoft

To download this report, which has been made available at no cost until the end of February, go to https://amalgaminsights.com/product/analyst-insight-cloud-foundry-and-kubernetes-different-paths-to-microservices

Microsoft Loves Linux and FOSS Because of Developers

Tom Petrocelli, Amalgam Insights Research Fellow
For much of the past 30 years, Microsoft was famous for its hostility toward Free and Open Source Software (FOSS). They reserved special disdain for Linux, the Unix-like operating system that first emerged in the 1990s. Linux arrived on the scene just as Microsoft was beginning to batter Unix with Windows NT. The Microsoft leadership at the time, especially Steve Ballmer, viewed Linux as an existential threat. They approached Linux with an “us versus them” mentality that was, at times, rabid.

It’s not news that times have changed and Microsoft with it. Instead of looking to destroy Linux and FOSS, Microsoft CEO Satya Nadella has embraced it.

Microsoft has begun to meld with the FOSS community, creating Linux-Windows combinations that were unthinkable in the Ballmer era.

In just the past few years Microsoft has:

Please register or log into your Amalgam Insights Community account to read more.
Log In Register

Lynne Baer: Clarifying Data Science Platforms for Business

Word cloud of data science software and terms

My name is Lynne Baer, and I’ll be covering the world of data science software for Amalgam Insights. I’ll investigate data science platforms and apps to solve the puzzle of getting the right tools to the right people and organizations.

“Data science” is on the tip of every executive’s tongue right now. The idea that new business initiatives (and improvements to existing ones) can be found in the data a company is already collecting is compelling. Perhaps your organization has already dipped its toes in the data discovery and analysis waters – your employees may be managing your company’s data in Informatica, or performing statistical analysis in Statistica, or experimenting with Tableau to transform data into visualizations.

But what is a Data Science Platform? Right now, if you’re looking to buy software for your company to do data science-related tasks, it’s difficult to know which applications will actually suit your needs. Do you already have a data workflow you’d like to build on, or are you looking to the structure of an end-to-end platform to set your data science initiative up for success? How do you coordinate a team of data scientists to take better advantages of existing resources they’ve already created? Do you have coders in-house already who can work with a platform designed for people writing in Python, R, Scala, Julia? Are there more user-friendly tools out there your company can use if you don’t? What do you do if some of your data requires tighter security protocols around it? Or if some of your data models themselves are proprietary and/or confidential?

All of these questions are part and parcel of the big one: How can companies tell what makes a good data science platform for their needs before investing time and money? Are traditional enterprise software vendors like IBM, Microsoft, SAP, SAS dependable in this space? What about companies like Alteryx, H2O.ai, KNIME, RapidMiner? Other popular platforms under consideration should also include Anaconda, Angoss (recently acquired by Datawatch), Domino, Databricks, Dataiku, MapR, Mathworks, Teradata, TIBCO. And then there’s new startups like Sentenai, focused on streaming sensor data, and slightly more established companies like Cloudera looking to expand from their existing offerings.

Over the next several months, I’ll be digging deeply to answer these questions, speaking with vendors, users, and investors in the data science market. I would love to speak with you, and I look forward to continuing this discussion. And if you’ll be at Alteryx Inspire in June, I’ll see you there.