Omdia Universe: Container Management Solutions, 2022–23

 

Download Omdia Universe Container Management Solutions 2022 2023 - 4.59mb | pdf

 

Summary

Catalyst

The growth in cloud-native applications(built to execute in the cloud independent of the underlying infrastructure) is synonymous with container use. However, container technology introduces specific challenges for CIOs, the not least of which is the lack of skills in managing large-scale container environments. It is important to understand the key challenges with adopting containers at scale from a management perspective and to discover some of the different vendor approaches to mitigating these challenges.

Figure 1: The Omdia Universe for container management solutions Source: Omdia

All vendors included in the report were approached and provided input to this study.

Omdia view

We are shifting from a predominantly VM-based world to a more mixed cloud-native and VM-based reality. Over the next five years, VMs’ dominance will erode as more workloads are developed and deployed in container-based environments. While Omdia does not believe that all workloads will become container-based, we expect this to represent a significant proportion of workloads by 2023. The developer community has been quick to see the potential of containers and has embraced the technology. However, this shift requires an equally significant cultural and skills shift to occur within operations and security teams, and this is one of the forces holding back the faster adoption of containers.

Software vendors and cloud service providers are responding to this resistance by designing solutions and services to simplify the management of a container environment. The important elements of any container management solution differ significantly from those of a traditional VM management solution. Customers are looking to build high-scale services consisting of microservices. This differs from a traditional VM management solution because containers are application-based, not machine-based. Even though container management is more application-based, the high-level capabilities required in any management solution are the same, but the detailed approach taken by container management solutions will differ. Container management solutions take into account that environments are distributed and that applications move and are ephemeral.

Analyzing the container management solutions universe

Market definition

Omdia believes 10 capabilities are key to any container management solution: developer experience, runtime and registries, scale and flexibility, security operations, operational management, pricing and TCO, observability, support and training, open source credentials, and automation.

Developer experience

Developers drive the adoption of cloud-native applications, and any container management solution must recognize their needs and make the developer experience as productive and intuitive as possible. 

Runtime and registries

One of the aspects of a container management solution that differs from the existing VM management solutions is that it must accommodate the developers and operational teams. The runtime and registries are a critical component of the cloud-native ecosystem, and responsibility for managing these must be an integral part of any solution.

Scale and flexibility

The big challenge is for organizations to manage container environments irrespective of where the containers are executing (on-premises or in a cloud). The other significant challenge is managing the sheer number of container images, which in a microservices architecture are likely to be an order of magnitude greater than the VMs currently managed.

Security operations

The security requirements in a container environment are different from those in a VM environment. Any solution must deal with the specific security requirements of applications as well as more traditional infrastructure security concerns.

Operational management

Managing a container differs from managing a VM. A container requires the workload to be managed separately from the host VM. This means that in a container-based environment, the containers are more portable and significantly greater in number. Therefore, resource and service management must operate at the very granular resource level, the core infrastructure level (Kubernetes elements such as Pods, worker, and master nodes), and the service level (the combination of many different containers).

Pricing and TCO

This capability compares suppliers on the cost and total cost of ownership (TCO) for several scenarios. Each vendor has a different set of offerings made up of different features and elements (for instance, are the infrastructure resources included, is the offering a fully managed service or a partially managed service, etc.).

Observability

One of the key management activities in a cloud-native environment is the ability to move beyond monitoring and perform observability. Monitoring is tooling or a technical solution that allows teams to watch and understand the state of their systems. Observability is tooling or a technical solution that allows teams to actively debug their systems.

Support and training

One of the main reasons for the popularity of container management solutions is a skills shortage and the need for existing IT staff to manage containers with as much familiarity as possible. This is particularly important from a developer perspective because the correct solution will empower the developers and not get in the way.

Open source credentials

The containers market is still an evolving market and, as such, has a number of competing technology standards. The key element is that any platform can manage the most popular technologies and support industry-recognized standards.

Automation

Managing any container deployment at scale requires the underlying nodes to be managed correctly. The nodes must also remain up to date with current patch levels. Automation plays a significant role in simplifying management tasks—whether they are for the customer or being delivered as a managed service—and it is key to supporting containers at scale operationally.

Market dynamics

Containerization dates back to the 1970s, when the technology was initially used solely for isolating application code. The technology initially lacked the usability and portability benefits users have come to expect since the launch of Docker and the introduction of the container engine in 2013. The containerization movement then formed around the open source project, which addresses the common pain point of “dependency hell,” the challenge of making an ever-increasing set of applications, language(s), frameworks, and more interact properly with an ever-increasing set of hardware environments. Kubernetes has its roots in work done by Google, and it launched at DockerCon in 2014. Google had been running containerized workloads in production for more than a decade, and through an internal project named Borg, it developed an orchestration engine. Kubernetes traces its lineage directly from Borg. Many of the developers at Google working on Kubernetes were formerly developers on the Borg project. However, other approaches were also being developed, most notably Service Fabric, which is Microsoft’s container orchestrator solution for deploying microservices across a cluster of machines.

The container market is evolving rapidly, and therefore the container management solutions market is equally dynamic. The market is exhibiting great momentum for containers and equal momentum to drive standardization to accelerate adoption and industry growth. Kubernetes is a ubiquitous solution while AWS ECS and HashiCorp Nomad are widely used alternatives. Some approaches, notably Swarm and Mesos, are open source projects that did not survive. However, Kubernetes and other orchestrators and concepts are often alien to IT administrators, so the direct use of Kubernetes to orchestrate a large deployment can be challenging. The rise of container management solutions has helped IT administrators remove some of the complexities of using Kubernetes and bridge the new and existing operational activities, enabling organizations to transition at a pace that fits business demand and IT capability.

The evolution of the container management market will almost certainly be led by solutions that augment the use of new technologies in the storage, security, and network spaces into a single management offering. Current solutions mostly take the Kubernetes orchestration solution and wrap other management functions around it, creating more of an enterprise solution. However, as the use of these solutions expands, more gaps in capability will be discovered, and the solutions must be flexible enough to accommodate these changes. Many of the solutions are designed to be “open” and extensible to support vendor-specific hardware and new operating models. Omdia expects the market to rapidly expand, with many operational management, cloud service provider, and monitoring vendors producing solutions. Ultimately, the market will coalesce on a smaller number of solutions, almost exclusively built using “open source” as the design criteria.

Figure 2: Vendor rankings in the container management solutions Universe Source: Omdia

Market overview

While the Omdia Universe classifies all the vendors into three distinct groups (Figure 1) based on their scores, the separation between the best- and worst-performing vendors was only 8% in terms of the overall weighted total Universe score; this gap was 5% between the leaders and the challengers. These results demonstrate that most container management solutions share a similar core set of capabilities, and the differences are more about how the solutions add value and are used.

Market leaders

The leaders (Figure 2) all recorded a weighted total Universe score (average of capability, customer feedback, and market impact) of over 62%, and they also scored an average weighted capability score (average of all the capabilities measured and the calculated measures) of over 64%. The leader category consists of two distinct, tightly formed groups. AWS, Red Hat, and VMware recorded an average weighted capability score of over 66% and were clear leaders. However, Alibaba, Huawei, IBM, and Oracle recorded an average weighted capability score of over 64%. The formation of clusters shows that the leader’s market has some differentiation between offerings, but the differences are more nuanced than fundamental in nature. The leaders were also responsible for 70% of the sub-category leading scores, further underlining their difference from the challengers. The leaders did not demonstrate significant strength compared to any other classification in terms of customer feedback scores. However, the leaders’ willingness to recommend scores were consistent, with all recording scores between 83% and 88%.

Market challengers

The challengers (Canonical, Mirantis, Rafay, SUSE, and Tencent) are closely grouped as they all recorded a weighted total Universe score of 59% and a weighted capability score of 62%. The challengers were responsible for only 30% of the top sub-category scores. Again, it is evident that although the leaders and challengers show areas of clear separation, these were not in any fundamental capabilities but were more related to the way some challengers’ capabilities are solely based on the upstream Kubernetes capabilities whereas the leaders demonstrated they have augmented these as enhancements. The challengers were not inferior to the leaders in terms of customer feedback: a challenger achieved the top category score for customer feedback, and challengers recorded two top sub-category scores in customer experience out of the three measured. However, the challengers were less consistent as a group, with wider variation between the highest and lowest scorers.

Market prospects

The prospects (HashiCorp) recorded a weighted total Universe score of less than 56% and a weighted capability score of less than 59%. This category shows some areas where there is clear differentiation between the leaders and challengers in terms of capabilities. However, much of the prospect’s scoring was due to an unwillingness/inability to provide answers to some critical questions.

Opportunities

The market for container management is growing rapidly as more organizations adopt the technology. However, in these early days of adoption, many organizations are not mature enough to have experienced all the challenges of managing containers at scale. Offering solutions that allow the customers to grow and use the most appropriate ways to manage their container estate is a key vendor capability. Omdia believes that vendor offerings that match customer needs will enable vendors to retain customers as they navigate and expand on their container journey.

Threats

The current market has many different vendor solutions that offer similar capabilities. However, customers are increasingly looking to adopt solutions for managing and executing containers in multiple environments that cross public cloud providers and on-premises and off-premises environments. According to Omdia’s research, the edge will be predominantly based on containers, representing a highly distributed and diverse environment. Omdia believes that any solution that does not address the need of the edge and include interoperability with different technology stacks will struggle to win and retain users. The other main threat to the market is forcing customers to only use a solution that has its control plane hosted in a single vendor’s environment. While this approach (central control plane) delivers the simplicity customers seek, it fails the flexibility test. Therefore, any container management solution must offer a choice of how the solution is managed (self-managed or managed service) and where any control plane is hosted (in a single vendor’s environment or in any environment).

Market outlook

The software forecast for container management solutions/platforms shows a CAGR of nearly 25% 2020 to 2025, and the market will be worth nearly $7.5 billion in 2025. Unlike some other technologies (i.e., the hybrid and multicloud management market), the container management market shows greater variation between industry verticals and geographies. The industry vertical with the slowest growth and smallest market worth by 2025 is the energy sector, with a CAGR of 19% and a market worth of $177 million in 2025. By contrast, the fastest-growing vertical with a CAGR of just over 28% is media and entertainment, and the largest market vertical by value in 2025 will be retail at $727 million (nearly 10% of the total market). Omdia’s IoT, Cloud, 5G, and Digital Workplace – IT Enterprise Insights 2022 survey data confirms this variation among industries. The percentage of workloads running cloud-native environments in 2021 for the retail sector was just over 35%, and the energy sector reported less than 30% of workloads running in cloud-native environments.

Indonesia is the fastest growing geography for container management solutions with a 2020–25 CAGR of 41%, and Japan is the slowest with a CAGR of just below 21%. Again, looking at cloud-native workloads, this time by geography, Indonesia has 35% of workloads running as cloud-native in 2021 compared to less than 33% in Japan. This variation in workload percentages appears too small compared to the difference in CAGR. However, Japan expects to show a minimal increase to just less than 34% of workloads in cloud-native environments in 2023, compared to Indonesia, where nearly 37% of workloads are expected to be in cloud-native environments, indicating faster growth in Indonesia. The other factor is the market size in these two countries: by 2025, Japan is expected to be worth $255 million ($99 million in 2020) and Indonesia $125 million ($22 million in 2020).

When it comes to market value, the US is the largest market with a forecast of $2.4 billion by 2025 (and a CAGR of just 22%), but China is the second-largest market with a forecast of nearly $1 billion by 2025 (actual figure $939 million) but a CAGR of 39%. The US expects 34% of workloads to be running in a cloud-native environment in 2023, up from 32% in 2021. In comparison, China in 2021 reported that 33% of workloads were running in cloud-native environments, which is expected to increase to 36% in 2023. In both cases, this shift is at the expense of virtualized workloads, with legacy non-virtualized workloads remaining fairly consistent at approximately 19%.

Vendor analysis

Alibaba (Omdia recommendation: Leader)

Alibaba should appear on your shortlist if you are looking for a consistent solution with robust capabilities in all sub-categories.

Overview

Alibaba is a leader in this report, with an overall weighted average capability score of 64.4% and a weighted average total Omdia Universe score of 62.3%. Alibaba was a consistent performer with scoring across all capabilities varying by only 15 percentage points—better than the 23 percentage point average variation for the cohort—and with the joint second-lowest variance overall. Alibaba was also one of only five vendors to score above 50% for all sub-categories in the capability questions, further demonstrating its consistency of performance. Another measure of Alibaba’s consistency is that it only recorded three sub-category scores below the average, which was in line with the leaders and better than the challenger average of six. The other observation from this degree of consistency is that Alibaba did not record any leading sub-category scores for capability or customer feedback. Alibaba’s overall score for customer experience was in line with the average, but it was above average in the recommendations score. Alibaba’s market impact score was above average as it has more than 20,000 customers globally and records more than 2 billion image pulls per month.

Figure 3: Omdia Universe ratings—Alibaba Source: Omdia

Strengths

Alibaba was strongest in the scale and flexibility sub-category, scoring 66%. The vendor has a few capabilities in this area that contributed to its score. Alibaba enables high availability by making sure the control plane of its managed Kubernetes service has multiple replicas across multiple availability zones. It also ensures that the data is protected with hot backup and cross-regional disaster recovery backup. Alibaba’s other noteworthy strength in this sub-category is the range of environments it supports, including various node operating systems: Alibaba Cloud Linux 2.x/3.x, CentOS 7.x, Windows Server 2019, and Windows Server version 1809 and later. It can also support existing nodes with customized images. The flexibility of Alibaba’s ACK ONE solution impressed Omdia, particularly in relation to providing consistent security management across the environments. Two noteworthy aspects are ACK’s ability to automate policy enforcement and perform security inspection. However, the real value is in ACK’s native container support, where its container security radar visualizes the real-time network topology and cluster assets, and the container firewall provides a granular, declarative policy for the container workload. 

Alibaba’s second strongest sub-category was run time and registries, where it recorded a score of 65%, the third-highest sub-category score and behind only the top category score of 67%. One of Alibaba’s strengths is its range of supported container runtimes and registries; the vendor supports containers, Docker, and other open container initiative (OCI) containers. Although it is not the most comprehensive list in the cohort, it is above the average and standard offering. Alibaba also provides the operational management capabilities needed to ensure the containers are secure, including image scanning, vulnerability analysis, vulnerabilities repair, and image signing. Omdia also likes Alibaba’s approach to workflow, with its cloud-native application delivery chain feature. This feature allows developers and administrators to freely combine tasks such as image building, image security scanning, geo-replication, image distribution, and image deployment in a single delivery chain. The cloud-native delivery chain can be fully observable, traceable, and secure.

Alibaba’s third strongest sub-category was open source credentials, where it scored 63%, an impressive 8% above the average and 21% better than the lowest score from competitors in the sub-category. Alibaba is a platinum member of the Cloud Native Computing Foundation (CNCF) and a general member of OCI. The vendor has an impressive list of open source projects it has been active in over the past 12 months, such as Apache RocketMQ, Dragonfly, and KubeVela to name but a few.

Limitations

Alibaba’s weakest sub-category was automation, where it scored just over 51%, which was broadly in line with the average score of just over 54%. Alibaba provides some capabilities that are above the average, such as its Yida low code platform that helps users build simple automation applications. However, most of Alibaba’s automation capabilities were based on Kubernetes standard capabilities, which was a common response from most vendors, and hence why Alibaba was in-line with the average.

AWS (Omdia recommendation: Leader)

AWS should appear on your shortlist if you are looking for a vendor with a long history of managing containers at scale and an innovator in delivering solutions that solve customer challenges.

Overview

It should be no surprise that AWS is a leader in the Omdia Universe, as AWS has long been the environment where the majority of containers in production are executed. AWS has built an expertise in managing containers at scale, which it has translated into a set of solutions for its customers. AWS has faced more competition to its dominance in the execution market (where containers are being run) as a greater choice of environments now support containers. However, its management experience has enabled the vendor to develop solutions that meet the needs of customers who are new to managing fleets of containers. AWS is a leader in the container management solutions market with a weighted total capability score of 66.1% and a weighted total Universe score of 63% (the average of capability, customer feedback, and market impact). AWS was the most prolific in terms of top sub-category scores, with 6 out of the 15 recorded top sub-category leading scores. AWS has a range of solutions for its customers that provide choice, from a fully managed service to a DIY approach. The vendor also developed capabilities such as EKS Anywhere, which allows customers to run and manage containers anywhere from an on-premises location to a public cloud. AWS scored well in customer feedback for willingness to recommend, where it was third with 86%; overall, customers rated AWS as in line with the cohort average.

Figure 4: Omdia Universe ratings—AWS Source: Omdia

Strengths

AWS had sub-category leading scores of 73% for both support and training and operational management. In operational management, AWS provides a number of services with different options that offer higher degrees of opinionated control or management for customers. In general, AWS aims to remove undifferentiated heavy lifting, which is where most customers realize the greatest benefit. Omdia believes that AWS’s value add in this space is its ability to let customers choose how much is managed (or not) based on their own specific needs. AWS categorizes its services into a three-tiered hierarchy (capacity, orchestration, and provisioning) under a holistic offering.

• Capacity: This is where the resources needed to run the container are managed; this includes AWS-managed services such as EC2 and AWS Fargate. AWS also supports customer-managed capacity, which can be characterized as the “DIY” option.
• Orchestration: Amazon EKS, ECS, and Red Hat OpenShift on AWS (ROSA) manage control planes for customers, which let them run complex applications with a high degree of control and without worrying about management of infrastructure.
• Provisioning: This is a set of products and tools that interface with the orchestrators and hide (or do not hide) the complexities of the orchestration layer. This layer includes services such as Docker Composer, AWS App Runner, AWS Copilot, AWS IoT Greengrass, AWS Elastic Beanstalk, AWS Amplify, Amazon Lightsail, and AWS Batch.

The other strongest sub-category was support and training, where AWS is the clear leader by almost 10% above the second-place vendor and over 15% above the cohort average score. AWS has developed a series of offerings that meet enterprise customers’ support and quality training needs. Services covering migration, training, implementation support, and professional services for development are offered through AWS Professional Services. AWS Professional Services created the AWS Cloud Adoption Framework (AWS CAF) to help organizations develop and execute efficient and effective plans for their cloud adoption journeys. The guidance and best practices provided by the framework help customers build a comprehensive approach to cloud computing across their organizations and throughout the IT lifecycle. Using the AWS CAF helps customers realize measurable business benefits from cloud adoption faster and with less risk. AWS Digital Training offers free, on-demand courses developed by the experts at AWS. With its new learning center, AWS Skill Builder, customers can explore learning plans and more than 500 digital courses to help develop individual careers. In fact, millions of learners have already accessed these free digital training courses, which are available in 16 different languages.

AWS’s third strongest sub-category was security operation, where it had another sub-category leading score of 71%. One of AWS’s key strengths is the approach it has taken to separate the control plane from the data plane by using two VPCs, which provides added security compared to other approaches where they are on a single network. This separation remedies most common security issues or misconfigurations encountered in self-managed systems. Internal control plane hardening is conducted under the guidelines and review of the AWS Security team and is consistent with the procedures used by other AWS services. Another key strength is there is no direct access to the control plane nodes or etcd from the internet or the customer network. The only service exposed is the API server, and that is gated by authentication (AuthN) and authorization (AuthZ). Using a defense-in-depth approach, AWS Config enables customers to assess, audit, and evaluate the configurations of AWS resources, including AWS container services. This demonstrates AWS’s commitment to security management and is why it was the sub-category leader.

Limitations

AWS’s weakest sub-category was open source credentials, where it scored 42%, which was affected by AWS not being able to answer all of the questions. In fact, AWS could or did not answer 50% of the questions, so this must be taken into account when considering this as a weakness.

Canonical (Omdia recommendation: Challenger)

Canonical should appear on your shortlist if you are looking for a vendor with a flexible, innovative solution that delivers highly in terms of price and TCO.

Overview

Canonical is classified as a challenger in the Omdia Universe; the vendor recorded a weighted total capability score of 62% and a weighted total Universe score of 59%. Canonical was the clear customer choice, with the top customer feedback average score of 90% and a second highest category score for pricing and TCO of 75%. However, Canonical was too inconsistent overall, with 10 scores in line or just below the average score among the 15 different classifications, 3 scores well above the average, and 2 scores well below the average.

Figure 5: Omdia Universe ratings—Canonical Source: Omdia

Strengths

Canonical’s strongest sub-categories were in pricing and TCO and scale and flexibility, where it scored 75% in each. Canonical had the top sub-category score for pricing and TCO and was 9% ahead of the second-place vendor in this category and 14% above the average. One of Canonical’s strengths is the approach it takes to pricing. Canonical Kubernetes is freely available for all users to download and use. It is only for companies that need support for Kubernetes clusters that involve a cost to the customer. Canonical provides a security maintenance and support subscription called Ubuntu Advantage. The pricing is based on physical or virtual node count; for example, if a customer wants 24/7 support for a Kubernetes cluster running on 12 virtual machines, they would need to pay 12 x $500 = $6000. However, cost is only one aspect of TCO. In terms of value, Canonical leverages the Juju Operator Lifecycle Management tool to automate application and infrastructure lifecycle operations, improving the efficiency of managing cloud-native estates at scale. Canonical support and managed services teams use the same tool, which is why it can offer a competitive price-to-performance ratio to its customers.

Canonical also scored 75% for scale and flexibility, but this was the third highest score at only 5% behind the sub-category leader. Omdia believes Canonical’s strength is its approach to APIs. The Charmed Kubernetes platform is deployed using Juju and interacts with the underlying API(s) provided by the target platform. When deployed on private clouds and public clouds, scaling with Juju is possible because it can create and destroy machines and deploy services onto the machines with all the relationships and configuration information. A simple example exemplifies the user-friendly commands: in etcd, if a customer wants to add an extra node, they can use the command juju add unit etcd.

The default Charmed Kubernetes architecture is not fixed, and scaling can be achieved at the Pod level for horizontal and vertical autoscaling using Kubernetes.

Canonicals’ third strongest category was operational management, where it scored 62%, which was in line with the average. Most of Canonical’s strength is due to the approach Canonical Kubernetes takes to Day-2 operations. The Juju Operator Lifecycle Manager (OLM) handles these activities. Charmed operators—typically written in Python—drive individual software components. The Charmed operators’ code defines specific actions and configurations and is designed to automate operational activities such as backup/restore, change configuration, etc. Omdia likes that the Charmed operators are all delivered as part of the solution.

Limitations

Canonical’s weakest sub-category was observability, where it recorded a score of 49%. However, there are two key reasons why Canonical scored poorly: it does not integrate with third-party AIOps solutions, and it currently does not use AI/ML techniques. However, Omdia did like that Canonical Kubernetes is deployed and integrated with the Canonical Observability Stack (COS). Support for the Kubernetes cluster and COS components are covered under the same subscription.

HashiCorp (Omdia recommendation: Prospect)

HashiCorp should appear on your shortlist if you are looking for an alternative to the Kubernetes-based solutions that form the majority of the market.

Overview

HashiCorp is classified in this Universe as a Prospect, which reflects the fact that the questions asked focused on the Kubernetes ecosystem, and HashiCorp is an open source project that offers an alternative to Kubernetes. HashiCorp could not answer 33 of the 92 questions, which is the main reason it only scored a weighted total capability score of 58% and a weighted Universe total score of 56%. Considering that for nearly a third of the questions, HashiCorp scored zero or a nominal partial capability score (based on Omdia’s secondary research), its performance was very good. HashiCorp scored above average for two categories; one of these was customer feedback, which is a significant indicator of how its users rate the vendor. The other area was developer experience, where HashiCorp also achieved a top sub-category score. Another key factor to consider is that Nomad is focused only on cluster management and scheduling, and although Nomad is directly affixed to Kubernetes, Nomad supports container network interface (CNI), container storage interface (CSI), and many other leading open source projects for container management.

Figure 6: Omdia Universe ratings—HashiCorp Source: Omdia

Strengths

HashiCorp’s strongest capability was developer experience, where it scored 75%, the top sub-category score. From a developer experience perspective, the fact that HashiCorp Nomad integrates seamlessly with HashiCorp Consul for service discovery and HashiCorp Vault for secrets management was a strength, but Nomad’s ability to use other service discovery tools made it so developer friendly. In fact, Nomad has an extensible plug-in system for task drivers. This allows users to use Docker, ContainerD, or Podman for their container runtime or not to use containers altogether. Tasks can be run using virtual machines, BSD Jails, LXC, raw executables, remote runtimes like AWS Lambda, or any runtime of choice with a custom driver. Nomad also integrates with leading APMs, including Prometheus, Datadog, and Splunk. This flexibility is a real strength for developers, and Omdia considers HashiCorp to be noteworthy in its approach to GitOps. For example, Nomad exposes the API for the CLI and HTTP(s), so it is not really limited in any way, no matter the tools used. Nomad Pack is currently in tech preview, but it is a package manager and templating tool that can also support CI/CD pipelines by having the CI tool invoke the pack on merge and then reconcile the new job spec. 

HashiCorp’s second strongest capability was scale and flexibility, where it scored in line with the average at 68%. Omdia liked its approach to edge and Nomad’s ability to enable the use of a simple topology with a single cluster across multiple regions and data centers, or multiple clusters respectively, without the need for an overarching control plane. Nomad’s lack of a built-in service mesh allows it to fit into most networking topologies used with edge compute. Nomad also supports device plug-ins that enable it to take advantage of hardware running on specialized devices.

Limitations

HashiCorp’s difficulty in answering the questions meant its weakest category was open source credentials, where it was unable to answer nearly all of the questions. This was because the wording of the questions focused on the Kubernetes open source projects as examples, and HashiCorp does not participate in these groups. This was the major contributing factor why HashiCorp recorded a score of 34%. However, Omdia does not believe that this truly reflects HashiCorp’s open source credentials—it is primarily an open source company, just not part of the Kubernetes open source projects.

Huawei (Omdia recommendation: Leader)

Huawei should appear on your shortlist if you are looking for a solution that supports a wide range of environments and a vendor that is active in the open source community.

Overview

Huawei is classified as a leader in the Omdia Universe with a weighted capability score of 64% and a weighted total Universe score of 62%. Overall, Huawei scored consistently across the 13 sub-categories in the capability section, with six above average, six in line with the average, and only one below-average score. Huawei did not record any leading sub-category scores but did have three second-highest sub-category scores and one third-highest score. In terms of the customer experience, Huawei scored an average of 81% for all categories rated by the customers, and its scores were consistent, with less than 5% variation across the different sub-categories. It is also important to note that the customer scores were from a mixture of customers in different countries. Analysis of the key reasons customers selected Huawei showed that product quality was overwhelmingly the most important selection criterion, with 42% of customers citing it as top reason for selection. The second most important selection criterion was features and functions with 21% of customers, and support and maintenance was third with 15%.

Figure 7: Omdia Universe ratings—Huawei Source: Omdia

Strengths

Huawei’s strongest sub-category was scale and flexibility, where it scored 70%. It can support any cloud platform or private cloud environment and has a list of certified environments, including Huawei Cloud ECS, FusionCompute, OpenStack, VMWare, QingCloud, and Alibaba Cloud Apsara Stack. Omdia considered the vendor’s ability to manage from the edge to the central core to be noteworthy. Huawei Cloud uses a distributed, cloud-native approach to manage distributed Kubernetes clusters and edge nodes with different specifications. Its solutions align with the capabilities needed in a hybrid and multi-environment world. For example, Ubiquitous Cloud-Native Service (UCS) centrally manages clusters that run on Huawei Cloud, private clouds, third-party clouds, and on-premises. This approach enables customers to centrally manage Kubernetes clusters in different locations on one management console. Huawei also supports multiple network access modes, such as public network, private line, and VPNs. In fact, Huawei unifies many of the required management capabilities such as permissions, security controls, and audit compliance. Its Operator Service Center (OSC) supports Helm and its own operator frameworks. Using OSC customers can interconnect with each cluster in the distributed cloud and deploy third-party cloud-native services in the clusters.

Huawei’s second strongest sub-category was open source credentials, where it scored 69%, which was the second-highest score, only behind the sub-category leading score of 70%. Huawei has taken the open source market seriously and boasts an impressive list of credentials, including a CNCF governing board seat, a CNCF ambassador seat, a CNCF TAG runtime chair seat, a CNCF TOC contributor seat, SIG chair seats, and maintainer seats in Kubernetes; it was also a founder of the CNCF KubeEdge Project, the CNCF Volcano Project, and the CNCF Karmada Project. Huawei benefits from its large customer base and reports an impressive 2.6 million developers using its cloud, which makes it the largest among vendors that provided an answer. Huawei was also one of the largest in terms of cluster size under management, with 10,000 nodes, which was jointly second behind the largest cluster size under management of 13,000. Huawei also reported that it was the number-one contributor as measured by accumulated committed code in Asia and Oceania and number 24 for Kubernetes globally.

Huawei’s third strongest sub-category was developer experience, where it scored 68%. Omdia particularly liked Huawei’s approach to development for multiple environments. Huawei Cloud DevCloud (full-code platform) and AppCube (low-code platform) provide public cloud and hybrid cloud delivery modes. DevCloud’s application development and deployment capabilities are multi-modal and use various technologies to reduce the burden on the developer, such as the ability to deploy applications on physical machines, VMs, containers, or functions. AppCube is Huawei’s low-code platform that enables capabilities, processes, and modules to be built quickly without the user needing to write any code. AppCube also has a low-code capability for more complex applications.

Limitations

Huawei’s weakest capability was automation, where it scored 51%. Huawei’s score was in line with the average and therefore is not a significant weakness. Huawei lacked some of the advanced features of competitors, but its automation capabilities demonstrate a solid, pragmatic approach delivered through its low- and no-code capabilities.

IBM (Omdia recommendation: Leader)

IBM should appear on your shortlist if you are looking for a solution that emphasizes the security of the management process by integrating with DevSecOps.