The calculation of data center capacity is essential to ensure that the data center operates efficiently and effectively. Understanding the limits of your data center is crucial for long-term growth planning.
This calculation is not a simple operation and requires the use of a Data Center Infrastructure Management (DCIM) tool to cross-reference all the necessary data and adjust it to obtain the most realistic capacity possible.
The concept of capacity in a data center can encompass various aspects, including physical space for equipment, storage capacity, and power capacity.
Historically, four critical capacity metrics have been identified:
- Space: The amount of physical space available in the data center is one of the most critical factors influencing capacity. Sufficient space is required for IT and M&E (Mechanical & Electrical) equipment, with additional space for future growth. Adequate space is also essential for proper airflow and preventing overheating.
- Power: Calculating electrical capacity is possibly the most important of all, as inaccurate predictions can lead to catastrophic failures when the data center exceeds its power limits.
- Cooling: Cooling capacity is critical to ensure the proper functioning of equipment. Calculating cooling capacity is important for planning new cooling needs and optimizing the efficiency of the data center by distributing the load effectively.
- Network Port Capacity: Having an adequate number of network ports to connect all equipment and users is crucial. Ensuring that port speed and bandwidth meet network traffic expectations is also essential.
Of these four metrics that we have just mentioned, the most complex to obtain are undoubtedly those related to energy and cooling since the rest are simply the result of contrasting the total available versus the total consumed. However, in the case of the energy capacity metric, the calculation, which involves comparing what is available with what is consumed, is not as straightforward as, for example, in the case of space, because neither the available energy nor the energy consumed are metrics as obvious as they might seem.
This is where a DCIM comes to the rescue because such tools understand the complexity of this calculation and are capable of accurately formulating this calculation, leaving us only with the responsibility of inputting the necessary values into the system.
These values to be entered are often a source of confusion during the creation of these metrics, and this is due to the existence of two general types of values, nominal values and monitored values.
To perform the calculations for energy and cooling capacity, it's essential to understand the fundamental difference between these two types of values. Nominal values are the theoretical maximum values that different manufacturers claim their equipment will achieve at peak performance. Monitored values, on the other hand, are instantaneous values, a snapshot of their consumption at a specific moment, and only by creating a historical record of these data can we somehow obtain a value similar to the theoretical maximum represented by nominal values, which, as we will see, are the values we really need. Nevertheless, monitored values, as we'll discuss later, also have their importance in the accurate calculation of capacities.
The need to obtain a theoretical maximum value for the calculation of energy capacity is because we must consider "the worst-case scenario," which is the moment when the energy demand of our data center is at its highest possible level. This allows us to plan for our growth accordingly.
Es erróneo plantear el punto de consumo para el cálculo de capacidad basándonos en un consumo medio o puntual de nuestro CPD, ya que no podemos arriesgarnos a que en un momento determinado el consumo de este sea mayor, y pueda provocar un fallo catastrófico por exceso de demanda de energía. Se hace imprescindible plantear ese punto de consumo, como el punto de máximo consumo posible, es decir, la tormenta perfecta, donde nuestro centro de datos se encuentre a su 100%. Este será el punto de partida desde el cual nosotros tendremos que calcular nuestra capacidad.
On the other hand, and no less important, the available power in our data center that we are going to specify for our calculation also needs to adhere to the criterion of 'the worst-case scenario.' In other words, the maximum available power should be determined by the weakest link in the chain, which is typically our Uninterruptible Power Supply (UPS)
Having said all of this, it is important to clarify that the electrical consumption metrics introduced by manufacturers in their technical specifications are generally significantly overestimated, often reaching up to 50% higher than reality. To overcome this deviation and perform a calculation that is as close to reality as possible, thus increasing data center efficiency, monitored values come into play. Monitored values are essential for correcting the previously mentioned deviation by creating a historical relationship of these values, which can help reveal the level of overestimation in the nominal values entered into the tool
In summary, calculating the capacity of a data center is essential to ensure that the data center operates efficiently and effectively, as well as to ensure proper scalability. Metrics are crucial for making precise capacity calculations within a DCIM, with a particular focus on critical aspects such as electrical and cooling capacity.
Let it work for you
