kentoh - Fotolia
Hyper-convergence promises an off-the-shelf product that consolidates resources, simplifies implementations and streamlines IT operations. The installation should streamline data center operations and free up IT resources without adding undue overhead.
Unfortunately, hyper-convergence doesn't automatically guarantee seamless integration with other IT systems, such as monitoring tools or cloud computing technology, resulting in a data center infrastructure that is more of a hindrance than a help.
One way to assess a hyper-converged infrastructure (HCI) option is with three categories: compatibility with the existing data center, support for existing capabilities and features, and the infrastructure's ability to meet current UX expectations.
Compatibility with the existing data center
When choosing an HCI option, organizations can opt for a preconfigured appliance or do-it-yourself (DIY) software approach. In either case, decision-makers should verify with IT admins that the selection's hardware is compatible with existing data center hardware.
An HCI architecture requires the right network and management ports to connect to the physical network; if an IT team is taking the DIY approach and plans to use in-house hardware, both managers and admins should confirm that the HCI software will run properly on current servers and networking appliances.
Decision-makers must check with IT admins that the HCI is compatible with the software that drives the cloud and account for hypervisors, OSes and network protocols. Organizations that use OpenStack to deploy their cloud architecture might consider an HCI built with OpenStack technology.
On the other hand, if IT admins rely on VMware products in their data centers, choosing an HCI that's preintegrated with VMware vSphere makes setup easier.
Decision-makers should also evaluate the HCI's compatibility with the software tools used to manage, secure, orchestrate and provision computing resources. These applications must seamlessly work across all system boundaries to effectively manage cloud and hardware resources. A highly proprietary HCI offering can make it difficult to use the tools or, worse, render them useless.
The available HCIs take different approaches to integrating with other systems and software. One of the most common is to expose a set of standards-based REST APIs that provide access to management capabilities. Offerings such as Hewlett Packard Enterprise's (HPE) SimpliVity, Dell EMC's VxRack Flex and Cisco's HyperFlex all expose management APIs for integration with other products and services. Some options also come preintegrated with third-party systems, such as Dell EMC VxRail, which has integration capabilities for VMware tools.
Support for existing software capabilities
When an organization introduces hyper-convergence into its cloud-supporting data centers, decision-makers must ensure it will reinforce the users' day-to-day workflows. Organizations might require an HCI to run business applications, support special technologies or integrate with third-party services and service-level contracts.
Organizations use a wide range of cloud-based business applications, such as ERP or customer relationship management software. If those programs must run or connect to the hyper-converged setup, decision-makers should verify those capabilities with IT managers before selecting an HCI offering.
Decision-makers should also evaluate whether the HCI can support any required back-end systems that host business applications. If an IT team wants to run SAP HANA in a hyper-converged environment, decision-makers might choose an SAP-certified HCI, such as those available from Dell, Fujitsu, HPE or Lenovo.
Just as important are any special technologies that support different types of workloads, such as containers. In this situation, IT might turn to an option such as NetApp HCI, which is validated for Red Hat OpenShift containers, or Nutanix, which supports Kubernetes.
Organizations rely on multiple cloud services to conduct business. Even if they run their own cloud environments, they might use third-party services to support some of their operations, such as disaster recovery or identity and access management (IAM). Decision-makers must ensure that the HCI they choose can integrate with these other services, without having to jump through multiple hoops to get there.
Ability to meet UX expectations
An HCI setup must meet performance and reliability requirements set out by user expectations, without cloud service-level agreements being compromised by integration issues. If an HCI is incompatible with cloud services, IT might need a software layer that abstracts resources to properly integrate systems, which can affect performance and UX.
Decision-makers should also ensure that the HCI option meets all security and privacy requirements and works with security systems, such as directory services, IAM measures, malware protections or single sign-on protocols.
The HCI should not degrade the end-user experience in any way, such as requiring multiple logins. Diminished UX can quickly translate to an unsuccessful HCI implementation, which can decrease ROI.
Hyper-convergence can introduce its own security complexities and open up the organization to new threats. Decision-makers should evaluate an HCI's potential effect on security and what it takes to keep systems online and secure.
The bottom line is that an HCI setup must deliver the same level -- or better -- of services that users expect from cloud and connected applications, without putting sensitive or personal data at risk. The HCI must be compatible with the underlying cloud infrastructure and support existing capabilities and services to achieve this.