Hyperconverged servers are modular systems designed to evolve over time depending on load and needs. Discover the advantages and disadvantages of these systems.
If you are in the acquisition phase of storage or server platforms, it’s impossible not to trip over the recent offers for hyperconverged systems. Your first questions might be: What are the advantages of hyperconverged systems, and how are they different from convergedsystems? The best questions to ask are: What are converged systems, and how do they differ from traditional infrastructure systems?
How do hyperconverged systems differ from traditional infrastructure systems?
Hyperconverged systems are a natural evolution of the traditional IT infrastructure, which is usually made up of silos according to business and operational needs. This results in groups and separate administrative systems for storage, the servers and the network. The group responsible for storage, for example, manages the purchase, supply, and storage support infrastructure, but is not necessarily concerned with network or enterprise server infrastructure issues. The same situation exists for the servers and the network. The relatively recent concept of hyperconverged systems combines two or more of these infrastructure components as a solution.
At what point should you become interested in a hyperconverged system?
During the evolution of a business’ infrastructure, such as for a data center, there are several pivotal moments that require consideration about its evolution and composition. According to recognized needs, the identified uses and their alignment with the business’ strategy, the path to take will not be the same in all cases. To summarize, there are essentially three options available for a company: expand, consolidate or renew the existing with one or more hyperconverged systems in order to, and this is essential, support and serve the business and its evolution.
First option : expand the existing
The first option is the expansion of existing infrastructure by adding one or more hyperconverged systems, which means retaining all or part of the existing infrastructure. This approach offers the advantage of not questioning the existing infrastructure and simply adding new network and storage resources. However, keep the consistency of the infrastructure in mind. The risk can be an unbalanced architecture at two distant performance levels from each other, making load balancing tricky. There two aspects to take into consideration: the network of a hyperconverged system is often at 10 Gbit/s, where the majority of firms still operate at 1 Gbit/s. The bottleneck is sudden as the entire network operates at only a tenth of the optimal performance of a new generation appliance. On the storage side it’s not much better, although being contained within the same server, hybrid or full flash, can significantly improve the flow rates and response times. Sometimes there are unwelcome surprises due to the heterogeneity and length of storage devices, as demonstrated by the findings of the DataCore survey SDS (see inset). Conversely, the addition of more powerful recent equipment can slow down applications and critical business data, requiring faster access and treatment such as VDI.
- No need to interrupt activity
- Performance gain for storage and applications
- Risk of imbalanced network flows
- Software Licenses to be considered (VMware)
Second option: consolidate the existing
When it comes to replacing all or part of a company’s servers because of obsolescence or performance reasons, one of the possibilities offered by hyperconverged appliances is allowing their virtualization. The advantages are many: replacing a physical server with a hosted virtual machine, providing a new way of administration, more flexible and unified through a hypervisor. The benefits are also physical(space gained) and economic (reduced energy consumption). Taking into account savings on amortization periods facilitates calculations and comparisons against a hyperconverged server. The process of backup and disaster recovery is thereby also simplified, virtual machines have replaced physical servers by being centralized in a hyperconverged appliance. In this case, it enforces regular backups or replication in a disaster recovery scenario (DRP) or Business Continuity (BCP).
- Centralised administration
- Saving floor space
- Energy savings (performance, cooling)
- The hyperconverged server represents a single point of failure
- Implementation of emergency backup and/or replication
Third option: renew the existing
A hyperconverged approach can selectively replace certain application servers depending on whether they are job or business oriented. It’s not uncommon that a server is traditionally dedicated to each application, such as a Microsoft Exchange mail server, for example. P2V Migration (Physical to Virtual) of such a server within a hyperconverged appliance ensures its continuity without requiring reinstallation (the included migration tools abound) while benefiting from the material gain of the appliance. It is therefore possible to gradually replace physical servers by their equivalent in virtual machines. With one caveat, however. If this applies to the vast majority of x86 servers and applications, it does not hold true for “Legacy” servers based on AS/400 hardware, for example.
- Selective P2V Migration spread over time
- Progressive decommission of physical servers
- Intrinsic performance gain
- Some “Legacy” applications will not tolerate virtualized operation (AS/400, etc.)
More around this topic...
© HPC Today 2019 - All rights reserved.
Thank you for reading HPC Today.