Posted by Donovan Stewart on Dec 18, 2018 10:00:00 AM
If you’re struggling to meet today’s demanding business requirements with your traditional compute architecture, it may be time to move to a converged architecture. Why? Let’s begin with a discussion of the high-level differences between traditional and converged architectures. We’ll then explore the pros and cons of each approach.
High-Level Compute Architecture Overviews
In most cases, the traditional architecture consists of different types of hardware that use standalone servers backed by a traditional storage array that uses iSCSI, Fiber Channel, and NFS protocols. You may have many top-of-rack switches that add connection points, causing multiple north- and south-bound hops that allow for server-to-server communications to take place. You may have older and newer hardware components that co-exist. Some may be under support, while others may be nearing end-of-life or already past end-of-life dates. Large organizations will generally require more skilled IT team members to manage and maintain these traditional environments.
Converged architecture has many validated reference architectures to base the design from, which creates a more supportable environment that can enhance performance and day-to-day operations. In the converged architecture approach, the networking and fabric are combined to minimize the number of hops and allow for more east/west server-to-server communications. Converged architecture reduces the amount of hardware required in the data center, which in turn reduces the amount of time spent managing and maintaining the equipment. Oftentimes, the storage array will be a shared array, as it is in a traditional architecture, but it will have evolved into a hybrid or all-flash array that allows for improved performance and potential data reduction capabilities. This creates an optimal and more efficient storage footprint in the data center.
Quick Hit Pros + Cons
As long as the environment is properly maintained and the hardware is refreshed often, a traditional architecture can perform well and service a large number of workloads. In many cases, it will require more IT team members who have a stronger skill base, and it can cause silos in management and day-to-day operations. In larger enterprise environments, this is typically not an issue.
A downside to this environment is that many layers of management systems or consoles are required to understand how the system is performing and stay ahead of performance curves. The more difficult provisioning process may cause users to have a negative perception of the IT team due to slower response times. The procurement process and support contracts may be a burden on the IT and procurement teams, which can cause devices or software licensing to fall out of support. One of the more critical downsides is that by running in a traditional environment, IT loses the ability to easily automate tasks, decrease response time to requests, or expand the infrastructure components as needed. When an issue or failure arises, the high level of complexity may require more team members and vendors to be involved to isolate or troubleshoot the issue.
A converged approach will typically require less resources to manage and maintain for day-to-day operations, as some of the traditional silos can be removed. Converged architectures have many validated reference designs, which act as starting points to build upon and result in complete solutions that operate in a more seamless fashion. This increases overall performance and improves operational efficiency for IT staff members, enabling them to be more productive. Automation is easier to achieve at all levels, as APIs have been tested and, in some cases, specifically developed for converged solutions that follow the validated reference architectures. Supportability is greatly simplified, as most vendors will have a better understanding of the configuration of the environment. A single vendor can own the entire support case or provide warm handoffs to other vendors as issues are identified. Either way, resolving issues becomes a much simpler process.
A downside is that IT team members may need to undergo training to feel comfortable working in the environment. In larger environments where responsibilities are split by technology, processes may need to be adjusted as the lines come together and overlap with converged architecture. However, this will ultimately lead to much less work for day-to-day operations and maintenance.
Requirement Short List for Converged Architecture
- Highly virtualized environments
- Multi-protocol environments
- Bare metal system requirements (applications or database systems cannot be or are not virtualized)
- Way to improve day-to-day operation efficiency/supportability and decrease data center footprint
[ GUIDE ]
EXPLORING THE MODERN DATA CENTER
Is your data center keeping up with your digital transformation demands?
ConvergeOne's Data Center Experts have written a guide that provides valuable insights about how you can strategically design your data center infrastructure to power the technology of tomorrow. The following areas are explored:
Active/Active Data Centers
Application Centric Infrastructure (ACI)
Topics: Data Center