Sergey Nivens - Fotolia
As cloud-native computing reaches mainstream enterprises, new Red Hat OpenShift hybrid cloud services extend the platform beyond Kubernetes into hosted applications.
Red Hat OpenShift Managed Cloud Services, OpenShift Platform Plus and Insights for OpenShift, all announced this week, are based on the vendor's Kubernetes platform, which is compatible with most public clouds and private data centers. These new offerings are meant to support distributed applications such as Apache Kafka anywhere a customer chooses to run them, while hiding their operational complexity.
"Customers aren't going to just drop everything they have on premises, but there's a lot of movement to public cloud," in part driven by the COVID-19 pandemic, said Stephanos Bacon, senior director of portfolio strategy for application platforms at Red Hat. "As they're moving to cloud, customers will also want managed services, but not for everything."
The first Red Hat OpenShift Managed Cloud Services will be API Management, Streams for Apache Kafka and Data Science, an AI as a service product. The API Management hybrid cloud service, based on Red Hat's 3scale API gateway, will be generally available on OpenShift Dedicated and Red Hat OpenShift Service on AWS (ROSA) this week, while the other two will be available in developer preview. Red Hat OpenShift Dedicated is a Red Hat managed service that runs on AWS and Google public clouds, and ROSA is a service jointly supported by Red Hat and AWS. Red Hat also offers OpenShift managed services on Microsoft Azure and IBM Cloud.
None of these applications are new territory for Red Hat OpenShift -- 3scale has been integrated with the Kubernetes platform since its acquisition in 2016, and Red Hat already supported self-managed Kafka on OpenShift through its AMQ Streams product. Self-managed OpenShift also previously supported various data science apps and GPU hardware.
The difference with these new hybrid cloud services is that they are fully managed by Red Hat, though users can also deploy and administer them alongside self-managed OpenShift.
"Some customers have [strict] requirements for performance and security configurations where AMQ Streams will be a better fit," Bacon said. "But within large organizations, there are groups that just want to build apps on a tight time frame and don't want to take on the operational burden."
OpenShift Streams targets Kubernetes - Kafka combo
As Kubernetes and container-based microservices architectures gained popularity, other platforms such as Apache Kafka big data streaming also emerged as important infrastructure components, according to one industry analyst.
"Companies that have been able to respond rapidly to changing conditions have done better [during COVID], and Kafka is a core element of managing data between highly distributed applications in that area," said Maureen Fleming, an analyst at IDC.
However, operating those platforms is at least as complex and difficult as managing Kubernetes. While their open source roots allow for some immunity to vendor lock-in, many mainstream organizations will still need a vendor's help to make them consumable, Fleming said.
Red Hat OpenShift Streams for Apache Kafka will be packaged differently in its initial release from its big data and API counterparts, according to Bacon. API Management and Data Science hybrid cloud services will be delivered as add-ons for OpenShift Dedicated and ROSA, but OpenShift Streams will initially run separately on multiple Kubernetes clusters managed by Red Hat.
Users will be able to link the hosted Kafka service into both self-managed and hosted OpenShift clusters using connection strings Red Hat plans to deliver via Red Hat Cloud API, command line interfaces and GUIs.
"Initially, adoption will be smoother [this way], though we anticipate having different [deployment] plans for Kafka," Bacon said. "Running it separately and internally will help us to develop the patterns we need to build additional services."
Stephanos BaconRed Hat
The Red Hat hybrid cloud services roadmap for Kafka will also include a Kafka Connector to directly link it to data science services, Bacon said.
"Kafka is the linchpin for connecting application-level services," he said. "Everything else is building around it."
OpenShift security, IT ops integrations expand
Red Hat also expanded OpenShift integrations with other elements of its IT automation portfolio this week. The Red Hat Insights performance and security observability tool already built into Red Hat Enterprise Linux will now be included with OpenShift and Ansible Automation Platform subscriptions.
These new integrations will offer predictive monitoring and troubleshooting support, security vulnerability management, automated issue remediation and cloud cost management tools for OpenShift and Ansible users. Insights integration with OpenShift ACM is also on the vendor's roadmap.
As with Managed Cloud Services, the theme for these updates to self-managed OpenShift is to more broadly support a full distributed application stack in multi-cloud and hybrid cloud environments, which includes the Linux OS, Kubernetes container platform and IT automation tool set. Red Hat officials said they don't expect Red Hat Insights to displace generalized Kubernetes observability utilities, but it can bolster Red Hat admins' troubleshooting toolkits.
While Red Hat Insights is vendor-specific for Kubernetes and Linux, it's cloud infrastructure-agnostic, which will likely appeal to Red Hat customers that want to avoid being locked into a public cloud provider, according to analysts.
"Our research indicates most [enterprises] use multiple clouds," said Al Gillen, also an analyst at IDC. "They want to avoid overcommitting to any one cloud vendor, and Red Hat can simplify the [multi-cloud] administration burden."
It's true that using Red Hat OpenShift as a central management layer for multiple cloud infrastructures moves the customer's lock-in burden to Red Hat instead, but that's where Kubernetes may act as an equalizer, Gillen said.
"A customer may have to migrate surrounding workloads if they want to move away from Red Hat, but the infrastructure itself theoretically could move easily to another Kubernetes platform," Gillen said, adding, "lock-in is inevitably a part of almost every decision [a user makes]."
Red Hat vs the world in hybrid cloud services
Red Hat has plenty of company in offering enterprises a platform for multi-cloud and hybrid cloud management based on Kubernetes. VMware Tanzu has a similar strategy, and each hyperscale cloud vendor has its own comparable hybrid cloud offering -- Microsoft Azure has Arc, Google has Anthos and AWS has Outposts and EKS Anywhere.
While the major public cloud providers' hybrid cloud services ostensibly support multi-cloud, third-party vendors such as Red Hat and VMware might be seen as more truly neutral centralized management options, according to IDC's Fleming.
Still, it's unlikely any one of these major vendors will capture a dominant share of the market any time soon, IDC's Gillen said.
"Anybody who's in this space has to have this kind of solution," he said. "Red Hat has the advantage of a large install base [for OpenShift], and chances are customers will remain loyal [to what they've already deployed] -- it will be more of an uphill battle to win new customers."
Enterprises may be using cloud computing more and managing deployments among multiple public cloud vendors, but multi-cloud portability for apps that move between different infrastructures is still rare, analysts say.
"While there is interest in using containers and Kubernetes for cross-cloud capabilities, most customers seek those capabilities today across hybrid environments more so than multi-cloud scenarios," said Arun Chandrasekaran, an analyst at Gartner. "Red Hat has an opportunity to address some of these challenges, [but] the near-term opportunity is to enable a common management plane and policy engine for its services deployed across distributed environments."