5 key technologies for cloud architectures

I confess to preferring simplicity, especially when it comes to computing infrastructure. It’s easier for startups and smaller companies to run their businesses with a mix of software as a service (SaaS) with applications and data hosted on one public cloud. Many organizations run on hybrid clouds with applications and data split between private clouds, legacy data center infrastructure, and a selected public cloud.

But many medium and larger enterprises choose to run on multiple public clouds or end up in that situation to support innovation, regulations, service levels, pricing negotiations, or acquisitions. There are choices and trade-offs to cloud strategies. If you work in IT, it’s important to know about the architecture options and better understand your organization’s IT cloud governance model.

I recently wrote a checklist for multicloud readiness with tips on strategy, cloud readiness, and devops practices. Now I’ll follow up with some of the enabling technologies for hybrid and multicloud architectures.

Hyperconverged infrastructure simplifies scaling private clouds

Data centers used to be filled with network equipment, servers, and storage from different manufacturers and management tools. Supporting this infrastructure required specialists, and scaling up and down computing capabilities wasn’t easy.

Today, many companies deploy hyperconverged infrastructure that combines network, compute, and storage in a building block appliance. In addition, management tools enable administrators to virtualize computing clusters and assign resources based on an application’s computing needs.

Tiwan Nicholson, director of IT service operations at UNOS, uses Nutanix and explains the benefits. “We’re seeing 30% improvement on laborious workloads compared with the previous infrastructure, where some of our big data and deep analytics jobs took days. Now, our generalist infrastructure engineers can manage the hardware layer and virtualization layer without the specialized skills needed for a dedicated big data database or blade server and all those different layers of the stack that we used to have to manage.”

In addition to Nutanix, Cisco, Dell, HPE, VMware, and others have hyperconverged infrastructure products.

Multicloud management tools enable self-service and governed provisioning

One of the main benefits of public and private clouds is self-service provisioning capabilities for developers, data scientists, and business analysts. Instead of waiting weeks for infrastructure, IT implements computing options, automation, governance, and chargeback pricing while cloud management tools provide portals and reporting.

Organizations optimizing self-service provisioning on one cloud can use the vendor’s self-service tools, but those seeking multicloud capabilities require management platforms that work across AWS, Azure, Google Cloud Platform, VMware, and other clouds. For example, BMC, IBM, Micro Focus, and others offer multicloud management platforms with key values that include self-service tools, security controls, devops tool integrations, templates, and automations.

Improve resiliency with cloud data management

Multicloud security has several challenges, and best practices require architecting multicloud identity and access management (IAM), network architectures, and encryption standards. These security considerations are important in single and hybrid clouds, but may be more challenging to implement in multicloud architectures.

The one security consideration that’s become critical for all organizations is cloud data protection. Ransomware is an enormous problem, with attacks impacting schools, hospitals, and other vital institutions.

Protecting the data on all clouds is a top concern for security and IT leaders, and cloud data management platforms offer solutions to encrypt, replicate, archive, and restore data while monitoring storage systems for ransomware-like attacks.

Amir Kioumars, data systems supervisor at Novato Unified School District, uses Rubrik to protect mission-critical data. Kioumars shares his thinking, “At the district where I worked previously, we were hit twice by ransomware and were down for 10 days. I was afraid the same thing could happen at Novato at any time.”

Other cloud data management platforms include Veeam, Dell, Cohesity Commvault, and others.

Avoid cloud silos with low-code integration

Organizations that have addressed multicloud management and security have the opportunity to accelerate strategic initiatives such as developing microservices, enhancing customer-facing applications, scaling machine learning initiatives, and re-engineering enterprise workflows. The challenge is to avoid silos where applications on one cloud can’t easily integrate with microservices on a second cloud, machine learning models on a third, or other SaaS platforms. 

Organizations targeting multicloud strategies should consider integration and integration platform as a service (IpaaS) to enable connecting data, microservices, and APIs across clouds, SaaS, and enterprise systems. Low-code integration platforms simplify connecting to common sources, automating transformations, cleansing data, and providing API management functions.

Tara Gambill, senior director of enterprise systems at MOD Pizza, uses Boomi to connect SaaS, data, and applications. Gambill states, “For our 100% SaaS business, having a cloud-native integration platform was table stakes. It’s not just the ability to enable fast, seamless employee onboarding. We’re also getting increased data accuracy and reliability. That allows our staff to spend more time on more impactful work.”

Other IPaaS platforms include Informatica, Workato, and SAP, and they compete on ease of development, out-of-the-box integrations, data management capabilities, and operational functions.

Automate and monitor multiclouds with AIops

IT departments that manage multiclouds require automation, monitoring, and incident management tools that process large-volume, real-time data sets and offer versatile automation capabilities. AIops, or applying machine learning and automation in IT operations, is an emerging capability that should be required to support multiclouds.

Capabilities of AIops platforms vary, but most start with aggregating alerts, monitoring data, observability data, and systems configuration and correlate the information to support incident management. Top platforms also include automation tools, discovery and dependency mapping capabilities, and analytics for site reliability engineers to manage service-level objectives.

Scott Johnson of Equifax uses BigPanda and shares the realities of operating across multiple clouds. Johnson acknowledges, “Running an always-on cloud-native paradigm as well as running on-prem is an extremely tough environment to be in. Troubleshooting, event correlation. Did a change to something you did on the on-prem side blow up something in the cloud? Being able to manage in that hybrid state is tough.”

Other AIops platforms include Moogsoft, OpsRamp, and Resolve Systems, and they compete on multicloud and AI-driven automation, monitoring, alerting, dependency mapping, and other capabilities.

Is the future of cloud multicloud?

Will more organizations adopt multicloud strategies, or will the costs and complexity of operating multiple clouds efficiently, securely, and reliably outweigh the benefits? Will public cloud providers outpace with innovation, or will competing technology platforms offer single-pane-of-glass platforms that enable IT to manage multiclouds efficiently?

These questions are worth pondering, but the better questions might be, how is your organization investing in configuring, managing, securing, integrating, and monitoring public or private clouds? To what extent do selected platforms enable multicloud hosting options?