Implementation of hyperconverged systems requires the installation of two components. First is the hardware, made up of servers and the networking that connects them. Second is the software, made up of the hypervisor that creates a cluster and storage software that provides data and services to the virtual machines the physical servers will host. Hardware-based solutions deliver these components as a single package with the goal of making life easier on the IT staff. But a side effect is loss of flexibility and higher costs.
The alternative to hardware hyperconvergence is software-based hyperconvergence where each component is available separately. In theory, a software-based approach is more difficult than a hardware-based approach since each component needs to be independently installed.
Comparing Green Field Implementations
The fundamental assumption in hardware-based hyperconvergence is that every installation is green field, meaning IT needs to start from scratch and install each component one at a time. If that assumption is correct, then a turnkey approach like hardware-based hyperconvergence should be easier to implement since all that work is already done. But software-based solutions also have partnerships with hardware server vendors. In combination they could provide a turnkey experience to the customer, too. At best green field implementations are a tie. Most importantly, green field implementations are not the norm.
Comparing Brown Field Implementations
Brown field implementations are situations where the server, hypervisor and network is already installed. In these situations, the organization may be looking at hyperconvergence as an alternative to expanding its shared storage infrastructure. In a brown field situation, hardware-based hyperconvergence is at a disadvantage. The vendor providing that solution either has to convince the organization to prematurely replace its existing infrastructure or create a separate cluster and put new applications just on it.
Software-based hyperconverged solutions fit much easier into the brown field situation. All IT has to do is install the storage software. Depending on the available internal capacity of the servers and the performance aspects of that storage, IT may need to also implement additional storage devices in those servers.
In most cases there are available drive bays for additional expansion, especially in a virtualized environment that was previously depending on centralized shared storage. Even if there is not, the software-based solution should support the implementation of “capacity nodes” to fill out any capacity requirements the environment might need.
Certainly there are green field situations where a turnkey approach to implementation may be useful. But green field is not the exclusive opportunity of hardware-based implementations. Software-based hyperconvergence can also be delivered in a turnkey manner but because it is software-based retain its future flexibility.
It is in brown field situations that software-based hyperconverged solutions have the upper hand. Software-based solutions can install on almost any server hardware type and support a mixture of different vendors. It can also support a wide variety of storage devices and does not require each server have the exact same storage configuration. That flexibility also provides hyperconverged solutions with better future proofing as new server and storage technologies are made available.
To learn more about hardware-based and software-based hyperconvergence watch our webinar, “Showdown: Hardware-Based Hyperconvergence vs. Hyperconvergence Software”. All registrants get a copy of Storage Switzerland’s Latest White Paper “Is A Hardware Approach Breaking Hyperconverged Architectures?”