With so many products and services being delivered through online networks, today’s companies are heavily reliant upon data center infrastructure to deliver high levels of system uptime. Even though data center equipment is more sophisticated than ever before, a few moments of system downtime can undermine everything a company has worked so hard to build. That’s why data center testing is so vital for all types of data centers in today’s hyperconnected world.
The Importance of SLA Uptime
Few things are more important for a data center than it’s SLA uptime guarantee. A service level agreement (SLA) is a legal document that lays out the terms and obligations of data center services. An SLA uptime guarantee specifically identifies the percentage of time the data center promises to keep colocation services up and running. Should uptime service fall below that percentage, colocation customers are entitled to remuneration to compensate them for any potential losses due to system downtime.
It’s important to note that an SLA uptime guarantee is distinct from an uptime reliability record. The former establishes the expectations of service, while the latter actually tracks how well the provider lives up to those expectations. So if a facility has a 100% SLA uptime guarantee, but a 99.99% uptime reliability record, customers are being compensated for the four minutes and nineteen seconds their systems are unavailable every month. That’s why companies should accept nothing less than 99.99999% uptime reliability when assessing colocation providers.
Types of Data Center
Even if a company is colocating assets with a single data center, chances are that their services rely on additional data centers, all of which have their own SLA uptime guarantee. There are a few different types of data centers that could impact services:
Edge Data Centers
Often used to provide services to local customers in smaller but growing markets, these facilities tend to have a smaller footprint than typical enterprise data centers. They are popular colocation sites because they allow companies to locate servers closer to end users, which allows them to use edge computing architecture to reduce latency. If these data centers do not deliver consistently high levels of uptime, they can disrupt services on the network edge, creating problems for Internet of Things (IoT) devices that use them for processing data.
Micro Data Centers
Similar in function to edge data centers, micro data centers are even smaller, often containing no more than a few servers in a modular or transportable infrastructure. They are often used to enhance network performance in high traffic areas (such as cities), so their downtime may not be as impactful as an edge facility’s downtime. Nevertheless, a loss of service could severely impact a company’s ability to deliver services.
Hyperscale Data Centers
The biggest of the big, hyperscale facilities are truly massive, often featuring several hundred thousand square feet worth of data floor space. Most enterprise-level public cloud services are powered by servers located in these facilities. Many of the largest hyperscale data centers are operated by established tech companies like Google, Apple, Facebook, and Microsoft. When they suffer an outage, millions of customers are potentially impacted. Even though the average colocation customer would never place their equipment in one of these facilities, their hybrid cloud network may rely upon the services that hyperscale data centers provide.
Data Center Equipment
Having the right data center equipment operating in top condition is essential to delivering on SLA uptime guarantees. It all starts with power and cooling. Power systems should ideally have more than one feed coming into the facility. In the event of a complete electrical outage, uninterrupted power supply (UPS) battery systems should spring into action until the facility’s backup generator can be brought online. A good UPS system incorporates redundancy to ensure that no single point of failure will endanger services. Cooling systems should also have fault-tolerant redundancies in place to keep the environmental controls up and running in the event of a disaster situation or unexpected equipment failure.
Maintaining and regularly cleaning equipment is important to keeping it running at peak efficiency. Accumulations of dust and other materials not only restrict airflow and limit cooling effectiveness, but they can also build up static electricity that could discharge suddenly and damage critical systems. Poor cable management can also create problems for equipment.
The Value of Data Center Testing
Having all of these components in place won’t do much to deliver on SLA uptime guarantees without regular data center testing. Can the facility withstand power surges without losing service? How long will backup systems run when there’s a power outage? The only way to know how well redundant systems will perform is to test them regularly.
Data center testing can expose potential faults and vulnerabilities in the facility’s infrastructure, any of which could contribute to server downtime or a complete system outage throughout the data center. Leaving key backup systems to sit idle for a long period of time increases the likelihood that vital components will deteriorate, causing these redundancies to fail in critical moments. In addition to regular inspections, both through intelligent monitoring and visual evaluation, data centers should conduct regular load bank testing to put their infrastructure through the paces.
Of course, uptime reliability is about more than just infrastructure. Frequent data center testing also helps to prepare data center personnel for potential problems and disaster scenarios. Reading about what to do in the event of a service outage is one thing; undergoing a live drill that simulates an outage is quite another. By conducting regular data center testing fire drills, facilities can better equip technicians and staff to respond effectively and calmly to emergency situations.
Data center SLA uptime guarantees represent an important promise made to colocation customers. Without consistently available data center services, companies will have difficulty building their business. No matter what types of data center an organization utilizes, knowing that the facility conducts regular data center testing to deliver on its SLA uptime is vitally important.
About Alan Seal
Alan Seal is the VP of Engineering at vXchnge. Alan is responsible for managing teams in IT support and infrastructure, app development, QA, and ERP business systems.