Technology has the power to define the success of any organisation, and IT leaders are feeling the weight of their every decision. These choices extend far beyond managing servers or balancing budgets and can shape the future of an organisation.
For years, IT teams have debated whether it’s better to manage infrastructure on-premises or rely on a colocation provider.
For many organisations, maintaining on-premises IT infrastructure may seem like the most efficient option. Having complete control over servers, data storage, and applications can be highly appealing. “And for some businesses, keeping their infrastructure on-premise does make sense, but there are hidden costs that are often not considered when IT leaders are making the decisions about where to host their infrastructure,” says Warren Schooling, Head of Sales Digital Parks Africa (DPA).
According to Schooling, the key question that needs to be considered when making the decision, is: “How much?”.
“How much downtime is acceptable to you, how much are you willing to pay to mitigate that downtime, how much will that downtime cost you, and how much effort and resources are required,” he says.
Understanding that the cost of on-premises infrastructure is not just about a rack or server, but also all the other requirements that go into managing equipment and ensuring the availability, alternative power and cooling, and the skills to manage not just the IT loads, but also the periphery requirements of maintaining local infrastructure.
Key considerations
- Redundant Power: For most South African businesses, the first key consideration is the power that is required to operate the infrastructure. Redundant power is a stark reality in the South African context with unreliable utilities. That means ensuring your on-premises infrastructure is protected by not only UPSes, but also generators. “The assurance that the UPS can carry the IT load as well as the auxiliary load for cooling, monitoring and support systems,” says Schooling. Battery capacity should be matched to the total load and allow sufficient time for the generator to synchronise and take load.
- Cooling: Overheating of sensitive IT equipment due to cooling units being out of service during a power outage poses significant risks. Organisations must consider the capacity requirements to take the load, as well as comply with the regulations required for storing fuel reserves on their premises. “As irregular and extended power outages continue across the country, businesses are forced to incur additional costs as they must now invest in and maintain in-house generators,” says Schooling.
- Specialised skills: Another important hidden cost that is often overlooked by businesses making this decision. Contracting the skills to manage the redundancy services are specialised and costly. Calculating loads, understanding structural requirements, maintaining generators, UPSes, and cooling systems are jobs for specialist engineers. “In entrepreneurial organisations, these functions are often left to IT support, who are expected to manage not only IT matters, but also a generator,” says Schooling. For most organisations, contracting a team of skilled electromechanical engineers at this level is challenging, costly and not feasible.
- Capacity planning: Taking into account the importance of future-proofing your infrastructure resilience, “What if your business needs to grow? Will you have enough space or capacity to meet the growing demands and infrastructure requirements?” asks Schooling.
- Security and risk management: “When you’re running a mission critical data centre on premises, you have to ramp up physical security and consider factors like emergency maintenance access control,” Schooling says.
These costs add up quickly, requiring significant upfront investment in infrastructure and skilled labour,” says Schooling. “This is why I believe the important question for companies must be ‘how much?’. If the total cost of ownership is too much, then a colocation provider may prove to be the best solution for your business,” he says.
How colocation can mitigate these costs
Colocation providers offer all the required services with a service level agreement (SLA), from redundant power, sufficient cooling, accredited engineering, networking and security. Capitalising on the benefits of shifting hosting expenses from CAPEX to OPEX under an SLA provides companies with confidence that their infrastructure is safe and resilient in contrast to what many believe they have when it is located on-premises.
“At Digital Parks Africa, we provide all the possible requirements, needs and solutions to support customer infrastructure under SLA. This includes electrical, structural and IT engineers overseeing your infrastructure and our data centre operations, the availability of backup power, provisioning of efficient cooling, access control and white space for your business to grow.”
“On-premises infrastructure may be sufficient for some organisations, but when availability, scalability and flexibility are mission critical, colocation becomes the smarter choice. Ultimately, the decision should align with your company’s goals, service priorities – and most importantly – be evaluated against how much you are willing to invest in this area of your business,” he says.