Analyst(s):Matthew Cheung, Laurie F. Wurster, Dennis Smith
Cloud management platforms are positioned to facilitate multicloud usage. Technology strategic planners of infrastructure software vendors need to integrate or partner with complementary cloud management tool providers to extend capabilities to manage multiclouds effectively.
Demand for CMP tools that effectively manage across clouds will grow gradually at a 6.6% CAGR through 2021 as adoption of public and multicloud usage increases.
Market consolidation is likely to continue since the CMP market is very fragmented with more than 20 vendors, most of them small in size, with varying levels of functionality that generate small amounts of revenue.
Competition will increase within the infrastructure software marketplace and from fabric-based infrastructure vendors and managed service providers as vendors integrate CMP functionality into product and service offerings.
To grow agile infrastructure opportunities, technology strategic planners of infrastructure software vendors:
Differentiate CMP products beyond tight integration to an individual stack by developing multicloud capabilities.
Improve core CMP functionality and gradually extend cloud management functionality to complementary areas by making building/buying/partner decisions based on completeness of your CMP and capabilities.
Deliver new functions rapidly, and compete with HCIS and service providers that already offer CMP as a service, by providing CMP through SaaS along with your on-premises offering.
Gartner reported that the public cloud infrastructure as a service (IaaS) market grew 31.4% in 2016 (see "Market Share: Public Cloud Services, Worldwide, 2016" ), resulting in a $22.6 billion market in 2016. While Amazon Web Services (AWS), Microsoft Azure and Alibaba are the leading cloud IaaS providers, the rest account for 46% revenue share in 2016. All this means that, as the "supply-side" market becomes busier and end-user organizations increasingly use more cloud IaaS services from different providers, cloud management becomes a challenge.
An increased number of Gartner clients have shown interest in "cloud management" in recent years. Gartner internal analytic statistics show that inquiries about cloud management grew 30% to 40% in the past 24 months, reaching more than 10,000 Gartner inquiries. The key driver of growing interest in cloud management platform (CMP) tools is adoption of a multicloud strategy. CMP and peripheral cloud management tools serve a higher abstraction layer to unify management of multiple clouds.
Gartner estimates the CMP market will amount to $346 million in 2017 and increase at an average growth rate of 6.6% during the next five years as cloud computing becomes a mainstream deployment model.
To capture this opportunity, infrastructure software vendors need to make key decisions (that is, buy/build/partner) about CMP go-to-market strategy. This document targets infrastructure software vendors that have already provided some or incomplete CMP functions or are considering entering the CMP market to complement their software portfolio (for example, platform as a service [PaaS] or IT operations management [ITOM] vendors), as well as pure-play point CMP solution vendors. This report also helps other vendors listed in the ecosystem understand the market trends and positions of different players. Decision factors to consider when filling gaps in the technology portfolio include:
Build CMP technology internally as a differentiator.
Integrate open-source components as building blocks to insure faster time to market.
Acquire a CMP point solution provider technology to fill product gaps. Partner with a CMP provider to expand market reach and fill product gaps when acquisition is not possible.
Package CMP products in modular forms so they can go thin to leverage the ecosystem (through partnership) or go heavy to provide rich functions by adding optional tools.
CMP is a growing market, with more than 20 players currently. Gartner does not see a clear leader in the market, but rather, a mix of a small number of larger vendors (for example, VMware, Microsoft, Cisco and Red Hat) and a long tail of smaller vendors (see sample vendors in the "Vendors to Watch" section). Some infrastructure software vendors have made inroads into the CMP space. However, increasingly, hardware vendors and service providers (xSPs) are adding more CMP functions to their product and service portfolio through build-and-buy or acquire approaches.
Gartner defines five core CMP functions and six adjacent cloud management tools/functions (see Figure 1 and the Definition of CMP section). However, not all vendors provide all functionalities, and in many cases, emphasize different capabilities of their products. Many vendors offer CMP through software as a service (SaaS) models (for example, RightScale and ServiceNow) or try to extend existing on-premises CMP offerings to SaaS. Also, CMP as a service can also allow quick deployment and continuous delivery of new CMP functions, with an aim to match native and unique public cloud management tools.
Note: Blue = Core CMP; Orange = adjacent cloud management tools
DR = disaster recovery
Source: Gartner (October 2017)
The major driver of CMP interest and adoption is the growing acceptance of public cloud and multicloud cloud usage. Gartner's forecast shows that public cloud IaaS will grow at a 28.6% compound annual growth rate (CAGR) through 2021 (see "Forecast: Public Cloud Services, Worldwide, 2015-2021, 2Q17 Update" ). Also, Gartner believes that about 70% of cloud service segment revenue will be dominated by the top 10 public cloud providers by 2021 — it's very likely that end-user organizations will use more than one (public) cloud in this backdrop.
With public cloud management the major driver for CMP today, multicloud will gradually compound the need as enterprises use more cloud services from multiple vendors, and attempt to manage them with CMP tools. CMP providers, whether they are infrastructure software vendors or adjacent market providers (hyperconverged infrastructure system [HCIS], PaaS and ITOM, for example), should differentiate CMP tools beyond tight integration with their own stack, and focus more on multicloud usage for both core and extended functionality.
As already mentioned, the CMP market is highly fragmented, an indicator that market consolidation will continue. In the past 24 months, the following mergers and acquisitions in the CMP space have taken place:
In August 2017, NetApp announced plans to acquire GreenQloud, an Icelandic CMP software offering Qstack infrastructure management software.
Micro Focus acquired cloud service automation through a Hewlett Packard Enterprise (HPE) Software division "spin-merger" (completed in September 2017).
In June 2017, Microsoft acquired Cloudyn, a cloud startup that provides multicloud billing and management.
In April 2016, Cisco acquired CliQr, a CMP provider offering multicloud management, to complement its pure-software CMP Enterprise Cloud Suite.
In January 2017, HPE acquired Cloud Cruiser, which provides cloud-consumption analytics software for the HPE Flexible Capacity business.
Moving forward, additional hardware and larger software vendors will enter the CMP market or strengthen gaps in existing CMP solutions through acquisitions. These acquisitions are designed to differentiate current offerings and complement their cloud-aspired systems, such as, integrated systems and HCIS with multicloud strategy. Managed service providers that offer managed and professional services on third-party cloud IaaS providers also build and buy their own CMP solutions.
As has been the case with ITOM tools in general, multiple CMP tools are necessary to provide the functionality required by end-user organizations (see "Market Guide for Cloud Management Platforms" and Figure 1) to effectively manage cross-platform (public/private) cloud solutions. Although most enterprise IT leaders indicate that a single solution is not optimal due to the potential of vendor lock-in and lack of functionality, requiring multiple solutions increases complexity and cost. End users currently need four to six different cloud management tools to gain complete functionality.
In general, Gartner sees two approaches to cater to cloud management needs among enterprises today:
Multiple CMP tools: Some organizations need best-of-breed tools in key areas (for example, cost management and configuration management) and want these specific tools to combine with other tools. Many organizations of this type turn away from having a CMP that has average capabilities across all areas.
One "good-enough" CMP, with a limited range of functionality: To reduce complexity of the IT environment, both on-premises and in the cloud, end users are demanding more comprehensive solutions with good-enough functionality to effectively manage their multicloud environments.
To meet these vastly different demands, CMP tool providers need to enhance functionality either through development or acquisition, and also package their CMP in a more modular fashion; thus, the consumption of these tools can be "mix and match."
Public cloud service providers offer their native cloud management tool, and continue to innovate their cloud services. In a multicloud management perspective, CMP providers need to provide open APIs (for example, Apache Libcloud) to connect to growing cloud services, and expose or integrate with public-cloud-native tooling. The aim is to provide nonrestricted and frictionless CMP for end users to enjoy new cloud services provided by individual cloud service providers without sacrificing governance of these services with the CMP layer.
As end users continue to deploy multicloud environments, they are evaluating solutions that address management in a cohesive and integrative way. The CMP market (see Figure 2) comprises vendors from a wide variety of other market segments:
Traditional ITOM vendors
Infrastructure software "stack" vendors
Point solution vendors
Fabric-based infrastructure vendors
Service providers (xSPs)
Source: Gartner (October 2017)
Large traditional ITOM suite providers' primary focus for management has been targeted at traditional physical and virtual infrastructures. Their management portfolios address a variety of key management disciplines: monitoring, automation, configuration/provisioning, IT service and support management, and software asset management. The CMP solutions that these vendors provide are often a combination of tools from an existing portfolio, supplemented with acquisitions that have specific cloud or virtualization capability, tied together as an entire suite. Additionally, because they focus on traditional ITOM management, they have the ability to offer a way to bridge management in traditional computing environments and cloud computing environments, and for some vendors, there is also shared technology between the traditional and cloud solutions. However, since these vendors focus more on traditional infrastructure and operations (I&O) processes, they struggle with DevOps teams' requirements in CMP functionality and multicloud usage.
Infrastructure software "stack" vendor segment of the CMP market represent providers of virtual infrastructure resources (such as Microsoft, Red Hat and VMware), in essence, the hypervisor and virtualization management. While some of these vendors offer some multiplatform (hypervisors or OS) capability, their predominant focus and product development will be for their own platform. Their differentiation lies in their ability to build greater instrumentation into the virtualization platform, as well as integrating management of the other layers of the virtual infrastructure. This category also includes PaaS vendors that provide a vertically integrated stack and aspire to provide multicloud support for their application infrastructure platform.
Point solution vendors include mostly smaller CMP providers that, potentially, can bring innovation in the CMP area since no legacy products have to be integrated into a CMP solution. Their solutions are mainly an integrated CMP stack. These vendors come from various software backgrounds — from open-source (for example, Scalr and HashiCorp) to pure-play CMP offerings (for example, Abiquo and Embotics). These providers would have a simpler product, with a limited range of functionality, but are quick to deploy. Moreover, they also try to address the DevOps requirements that traditional ITOM vendors are short of.
Fabric-based infrastructure vendors include mostly hardware infrastructure vendors offering cloud management software enabling them to sell private and hybrid cloud "solutions" and not just the features and benefits of their hardware (for example, Dell EMC Enterprise Hybrid Cloud and Native Hybrid Cloud offerings with VMware). Among them, HCIS providers and integrated system vendors are beginning to integrate CMP functions into their offerings. For example, Cisco has included its acquired CloudCenter (formerly, CliQr) in its HyperFlex (Cisco's HCIS) as an integrated offering; Nutanix Calm is another key offering of this kind.
This segment is complementary to the software-based CMP market that goes higher in the stack; however, they often compete for the same wallet share. The current problem with fabric-based infrastructure CMP (including HCIS) is that it is hardware-centric and rarely can support heterogeneous hardware — which end users may find inflexible. However, software-based CMP may require more integration effort and other tools (for example, hardware APIs such as OpenStack) to work in heterogeneous hardware environment or adopt a more "software-defined" approach.
Increasingly, we see more xSPs (including communications, managed and IT service providers) offering CMP functions as part of their cloud service portfolio. This could be in a managed style; thus, end customers would not need to cater to integration of individual CMP functions. The xSPs either use a few off-the-shelf CMP tools, or provide their own to bridge the functional gaps for specific needs. However, in most of cases, xSPs have specific CMP requirements (for example, multitenancy); thus, the off-the-shelf CMP targeting enterprises would not be fully functional to them. In this case, some xSPs will create their homegrown CMP solutions and compete with point solution or other infrastructure software providers.
In market conditions where acceptance of cloud slows or public cloud IaaS is monopolized (that is, dominated by one major vendor), and multicloud is not in demand, the growth prospects of third-party CMP vendors will stall. Moreover, if one public cloud vendor dominates the market, users will adopt a native public cloud management tool, rather than third-party CMP to manage multiple public clouds.
Cloud management vendors and technologies have both proliferated and consolidated during the past several years, as a result of interest and investment in public and multicloud adoption by enterprise IT organizations, as well as uptake of the technology by service providers seeking to build new cloud service businesses. The large number of small cloud management providers will increasingly find it difficult to differentiate themselves in a very crowded market, especially with large infrastructure and management vendors competing against them. However, because of their ability to innovate, best-of-breed capabilities will continue to attract visionary enterprises and service providers during the next three to five years.
To meet buyers' demands, CMP vendors need to add features and/or partner with complementary solution vendors to fill gaps in current functionality. Incomplete functionality explains why end-user organizations require multiple CMP tools to gain a fully functional CMP, especially when best of breed" in each category is important.
Sample CMP vendors include Abiquo, Accelerite, BMC, Cisco, CloudBolt Software, Cloudsoft, DivvyCloud, Embotics, NetApp (formerly, GreenQloud), Micro Focus (formerly HPE Software), IBM, Microsoft, Morpheus, Platform9, Red Hat, RightScale, Scalr, ServiceNow, HashiCorp and VMware.
The vendors listed above do not imply an exhaustive list .
We cover 11 cloud management functions, in which five are core CMP functions, and the others are adjacent cloud management tools. The following are detailed definitions:
Service Request Management — This is the self-service interface by which cloud services are consumed. Enterprises often face competing requirements where, on the one hand, some users will require a clearly defined, finite set of services provisioned from a service catalog, while, on the other hand, some users require a service interface that serves as a pass-through to native capabilities within a public cloud service. The former could cater to I&O staff that just wants to provision commoditized IaaS resources, while the latter could cater to developers that want flexibility to use a full suite of public cloud services. Service request management functionality is typically provided by CMPs or IT service support management (ITSSM) tools (see "ITSSM and CMPs Are Colliding at the Service Portal" ).
Provisioning, Orchestration and Automation — This refers to core orchestration/automation/provisioning capabilities that are often provided by CMPs, IT process automation engines within the CMP or native tools used to provision core infrastructure. This functionality includes user service blueprinting. Increasingly, this functionality is evolving into deployments where a CMP, when used, is becoming an orchestrator of a wide mix of underlying orchestrators, where other orchestrators (for example, AWS and Microsoft Azure native cloud orchestration/automation capabilities, OpenStack and/or VMware vSphere) are doing the native provisioning, while the CMP is operating at a higher layer of coordination (see "OpenStack Is Not a Cloud Management Platform" ). In this case, the CMP is often primarily used for applying governance and policy.
Governance and Policy — This is a key capability within CMPs, where this functionality could be detailed policy enforcement and governance (for example, clearly defining what users can do) or lighter tagging (for example, directing developers to a specific public cloud provider, but allowing them to access any service there). This also includes automation policies that allow reactions to environmental conditions (for example, enabling autoscaling).
Monitoring and Metering — This is base-level monitoring (the lower level of the stack with no awareness of the application layer) and metering functionality. It is mostly geared toward monitoring consumption of the cloud resources, possibly with showback for internal users. These basic capabilities are found in many CMPs. Note that additional monitoring functionality can be provided through service-level management, which is discussed below.
Multicloud Brokering — This is connectivity and content that allow the bridging of on-premises and off-premises cloud resources. In some cases, decision support capabilities are also provided (that is, identifying where best to run a specific workload). This functionality is often in CMPs and also tooling that is more geared toward cloud service brokering (see "Market Insight: Top 10 Things 'To Do' to Seize the Cloud Service Brokerage Opportunity" ).
Continuous Configuration Automation (CCA) — This refers to functionality that provides "last mile" configuration management support. This capability will typically integrate with a CMP, where a CMP orchestrates and the CCA tool performs directed automation. Increasingly, the tooling here is also adding orchestration and application release automation capabilities (see "Market Guide for Continuous Configuration Automation Tools" ).
Security and Identity — There are various security-related functionalities needed in a cloud computing environment (for example, risk management, segregation of duties, single sign-on and key management), particularly those involving public cloud services. Some of this functionality can be obtained through use of a cloud access security broker (see "Market Guide for Cloud Access Security Brokers" ).
While a small set of functionality is within CMPs, they mostly interface or federate with existing Active Directory, Security Assertion Markup Language or Lightweight Directory Access Protocol systems. Additionally, scanning and remediating deployed infrastructure for security vulnerabilities is needed where the remediation is performed by the automation functionality discussed in the Provisioning, Orchestration and Automation section above.
Service-Level Management — There is a set of functionality here that is broader and deeper than the functionality referred to in the Monitoring and Metering section above. It includes end-to-end monitoring of the application (for example, user interface, business logic and data persistence), which normally includes introspection within the application itself. Often, the functionality involves reactive, proactive and forensic monitoring of applications and infrastructure. This functionality is often linked to orchestration/automation to allow an elastic infrastructure where automated action is taken based on triggered events (see "Magic Quadrant for Application Performance Monitoring Suites" and "Market Guide for IT Infrastructure Monitoring Tools" ).
Cloud Migration and Disaster Recovery — The functionality here involves the cloud readiness determination, discovery, lifting and shifting of workloads between and/or among on-premises and off-premises environments. The use cases involve both workloads being permanently migrated from one environment to another, and workloads repositioned during DR testing or during an actual disaster. The linkage between these tools and other cloud management tools (for example, CMPs) is the ability of the other tool to manage the workloads that have been migrated by another tool. These tools are often procured through third parties that also offer migration assistance.
Capacity and Resource Planning — This functionality allows for the efficient operational use of the infrastructure footprint. It is often tied to orchestration and automation functionality. This area is increasingly being combined with the next function listed, Cost Transparency and Optimization. This is a logical tethering as capacity and resource planning has definitive cost considerations (see "Innovation Insight for Dynamic Optimization Technology for Infrastructure Resources and Cloud Services" ).
Cost Transparency and Optimization — The functionality involves enabling tracking, budgeting and optimization of the cloud expenses. This functionality is also often tied to orchestration and automation functionality, where action is taken based on usage (see "Innovation Insight for Cloud Service Expense Management Tools" ).