Pre-re:Invent hype illustrates the growing importance of the AWS ecosystem

Profile picture for user kmarko By Kurt Marko November 27, 2018
Summary:
re:Invent swings into high gear later today. This preview helps you navigate the maze of announcements with pointers as to where you should give your attention.

 

aws re:invent
What started as an event to evangelize, publicize and popularize Amazon’s groundbreaking system of rentable IT infrastructure has rapidly evolved into an influential industry gathering celebrating all things cloud. AWS re:Invent remains the primary vehicle for the cloud pioneer to introduce new products, highlight customers and train users. However it now transcends its sponsor and has become a showcase for a growing cadre of companies peddling cloud-related wares.

Like VMware's VMworld for the enterprise virtualization market and NVIDIA’s GTC for AI infrastructure and software, re:Invent has spawned a technical and business ecosystem that reaches far beyond its founder. By that measure, re:Invent has become a barometer of progress in cloud technology, enterprise adoption, and software evolution, in which a growing menu of on-demand infrastructure, platform and application services are catalyzing significant changes in business.

Re:Invent’s growing importance, along with AWS’s outsized presence means that the many satellite businesses seeking to capitalize on the event's massive publicity risk getting lost in the noise. More and more of them use the weeks preceding re:Invent to preannounce products in an attempt to stand out before the AWS publicity fireworks go off casting a shadow over everything else. The prequel PR is similar to how retailers now attempt to leapfrog each other's Black Friday sales by pre-announcing sales weeks in advance of the big day. Given AWS's parent company, the promotional similarity between retailing and tech marketing shouldn't come as a surprise.

At the risk of overgeneralizing an event with dozens of premium sponsors and more than a hundred exhibitors, the barrage hitting my inbox the past few weeks reveals that the re:Invent enterprise ecosystem consists of three broad categories:

  • Products to enable multi-cloud infrastructure and facilitate connecting legacy, on-premise systems with public cloud resources.
  • Products designed to manage, optimize and simplify the use of cloud services.
  • Products, services and consultancies that utilize (or facilitate the exploitation of) cloud services to improve business operations and enable innovative new products

Here's a quick look at some examples of each.

Bridging enterprise and cloud infrastructure

The trend of enterprises using a diverse array of cloud services and building multi-cloud IT environments isn't new to regular Diginomica readers. Those wanting more details can peruse some of my columns over the past couple of years for the background.

Like most things in IT, what sounds good in theory gets messy in practice as organizations soon discover devil in the details of bridging on-premise networks, data sources, user directories and back-office systems with counterparts in the cloud. While the cloud vendors generally provide excellent documentation and solution guides on how to complete the various tasks facing the multi-cloud IT organization, many find the DIY approach about as appealing as assembling a living room full of Ikea furniture. A growing number of cloud infrastructure software add-ons are filling the void.

An area of robust innovation and competition is multi-cloud network configuration and management in which, as I discussed here, products from VMware (NSX), Aviatrix, Nuage and others enable organizations to stitch multiple cloud VPCs and enterprise LANs into a unified fabric. As I detail here, Aviatrix mainly targets AWS users with an overlay technology encompassing various physical and virtual networks that enable enterprises to complex, multi-region, multi-cloud topologies, while also providing programmable management automation.

Network automation using intent-based semantics is also being extended from enterprise data centers to the cloud with companies like Apstra and Veriflow now supporting popular cloud platforms. For details on Apstra see this column, but the company’s vision is to have a network automation platform that works across an organization’s choice of hardware (Cisco, Juniper, et. al.), switch OS, compute platform (ESX, containers/Kubernetes, Windows) and cloud infrastructure (AWS, Azure).

Another type of cloud-enabling software is designed to manage data that must span cloud and on-premise systems. Many products target data replication for DR to the cloud, however, an increasingly important segment is data integration software to facilitate building hybrid data lakes and warehouses in which data generated in either on-premise or cloud environments is processed and stored in a central repository, which again might be in either location. Most of the traditional database and data integration vendors (Attunity, Informatica, Talend, etc.) now support cloud infrastructure, but they are joined by SaaS/PaaS products like Apigee, Dell Boomi, Jitterbit and Mulesoft.

Managing and optimizing cloud services

Each cloud service provides a management portal and set of APIs that can be used via a CLI or automation script to administer, monitor and debug cloud resources. While these are quite complete and continuously improving, they are understandably designed for one particular platform. As organizations incorporate multiple cloud services into their IT environment, the prospect of using a separate interface for each, adding to an existing enterprise infrastructure management system, is a non-starter.

The market for cloud management platforms (CMP) is still immature and lacking a universally agreed upon set of features. Gartner has probably the most comprehensive taxonomy, outlining 11 types of features, but the most common are:

  • Service provisioning and automation
  • Security and policy enforcement
  • Service monitoring and performance measurement
  • Cost monitoring and resource optimization, often combined with capacity planning
  • Disaster recovery/business continuity, often combined with multi-cloud data management.

A few companies like BMC, Morpheus and ServiceNow attempt to provide a full stack, cloud agnostic platform, but most specialize in a particular area with monitoring and cost management being two particularly active and competitive segments. Indeed, in analyzing the market for AWS tools, I found no less than 15 companies providing cost management and optimization products. While some focus on niche scenarios like identifying workloads suitable for spot instances, or only work with AWS, some like CloudHealth (recently acquired by VMware), ParkMyCloud and RightScale support each of the big-three cloud vendors.

Monitoring cloud-based applications is another area of active product development with products spanning low-level event and network monitoring to application performance management. Infrastructure monitoring software like Splunk and SolarWinds have extended on-premise systems into the cloud or been designed from scratch as multi-cloud products. Others like LogicMonitor, Sumo Logic, ThousandEyes (see my column on the company and its capabilities here) are cloud-native products delivered as SaaS.

As I recently detailed, an emerging trend is the use of data-driven analytics and machine learning, so-called AIOps, to aid in troubleshooting, system configuration, performance optimization and capacity planning. Many of these products, such as AppDynamics (now part of Cisco), Datadog, Lightstep and others work with multiple IaaS and SaaS environments.

Exploiting cloud flexibility and innovation for business

Another significant piece of the AWS ecosystem is less focused software products than on the use of cloud technology in business. While many of the applications are inward-facing, designed to improve the efficiency of IT operations or business processes, the most exciting and significant examples center on new products, services or customer programs that increase customer loyalty and generate revenue, either by augmenting existing streams, increasing margins or creating entirely new revenue sources.

Diginomica is full of use case studies in our "Digital enterprise in the real world" section and expect to see AWS highlight many new examples of companies innovatively using its services to create and improve their products. Like most cloud providers, AWS has a subsite dedicated to customer case studies that is full of example, but re:Invent also has a Builders Sessions track that allow small groups of customers to work with an AWS expert to prototype ideas and work through questions.

My take and advice

AWS is famous for packing re:Invent with so many new and enhanced product announcements that it's impossible to keep track of them all in real time. For example, AWS  publicized 10 products and technologies on the first night. However, don't be so overwhelmed by some of the bombshells AWS throws, like its custom ARM SoC for a new EC2 instance, that you lose track of the broader evolutionary trends of subsidiary cloud products and increasing enterprise adoption.

Whether wondering through re:Invent or following it online:

  • Pay attention to the growing symbiotic cloud ecosystem that augments or improve upon native AWS capabilities, particularly those designed to work across multiple cloud environments.
  • Watch for innovative enterprise uses of both native AWS services and ancillary third-party products.