Carlos Felix

Kratos: Automated Management of Cooling Capacity in Data Centers With Adaptive Vent Tiles

In data centers with raised floor architecture, the floor tiles are typically perforated, delivering the cold air from the plenum to the inlets of equipment located in racks. The environment of these data centers is dynamic in that the workload and power dissipation fluctuate considerably over both short-term and long-term time scales. As such, airflow requirements vary continuously. However, due to labor costs and lack of expertise, the tiles are adjusted infrequently, and many data centers are grossly over provisioned for airflow in general and/or lack sufficient airflow delivery in certain local areas. This wastes energy and reduces data center thermal capacity. We have previously introduced Kratos, an Adaptive Vent Tile (AVT) technology that addresses this problem by automatically adjusting mechanical louvers mounted to the tiles in response to the needs of nearby IT equipment. Our initial results were limited to a 3-tile test bed that allowed us to prove concept but did not provide for scalability. This paper extends the previous work by expanding the size of the test bed to 28 tiles and 29 racks located in multiple thermal zones. We present experimental modeling results on the MIMO (Multi-Input Multi-Output) system and provide insights on the external behavior of the system through CFD (Computational Fluid Dynamic) analysis. We develop an MPC (Model-based Predictive Control) controller to maintain the temperatures of racks below the thresholds through vent tile tuning. Experimental results show that the controller can maintain the temperature below the thresholds while reducing overall cooling air requirements.Copyright

Integrated management of cooling resources in air-cooled data centers

To accommodate the dynamic environment within raised floor data centers, cooling capacity is tuned during operation through zonal control means, e.g., active management of air conditioning resources. However, due to the spatial variance of cooling efficiency and time-varying cooling demand within zones, zonal adjustments alone are not able to maximize the thermal capacity of data centers. Without making local adjustments to the physical structure, such as altering vent tile openings, a data center can suffer significant reduction in thermal capacity and cooling efficiency, and such that facility lifespan. In this paper, we present active cooling technologies using both local and zonal actuators that improve overall cooling efficiency. Experimental evaluation in a data center shows that the integrated controller can adapt to changes to the system under control, significantly improve the controllability of the temperatures and reduce the energy consumption of the cooling facility.

Optimizing data center energy efficiency via ambient-aware IT workload scheduling

Energy consumption in data centers has increased over the years. A significant percentage of the total energy required by data centers may be consumed by cooling equipment. To decrease cooling cost, incorporation of air-side economization, or “free” cooling, which utilizes locally available outside air to augment or replace the cooling capacity provided by traditional data center cooling infrastructure, has been explored in recent years. However, little work has been done on exploiting the diurnal nature of outside air conditions by integrating the active management of cooling and IT workload scheduling. In this paper, we estimate the savings obtained by using this concept of integrated management in a containerized data center that implements a micro-grid of cooling resources consisting of outside air, conditioned air using DX (Direct Expansion) units, and a combination of both. For such a data center, a previously developed energy model is utilized to estimate the annual savings obtained by scheduling workload according to the variable cost of cooling resource generation and delivery. The impact of data center location and allowable IT inlet air temperature range on savings is also studied. The results show that the integrated management can decrease the cooling cost by up to 50% depending on location of the data center, the types of IT workload and total IT resources available in the data center.

Estimating data center thermal correlation indices from historical data

In order to better manage the cooling infrastructure in a data center with multiple computer room air conditioning (CRAC) units, the relationship between CRAC settings and temperature at various locations in the data center needs to be accurately and reliably determined. Usually this is done via a commissioning process which is both time consuming and disruptive. In this paper, we describe a machine learning based technique to model rack inlet temperature sensors in a data center as a function of CRAC settings. These models can then be used to automatically estimate thermal correlation indices (TCI) at any particular CRAC settings. We have implemented a prototype of our methodology in a real data center with eight CRACs and several hundred sensors. The temperature sensor models developed have high accuracy (mean RMSE error is 0.2°C). The results are validated using manual commissioning, demonstrating the effectiveness of our techniques in estimating TCI and in determining thermal zones or regions of influence of the CRACs.

A Data Center Environmental Advisory System

We present Avatar, a data center environmental advisory system for raised floor data centers. Using limited information such as inlet and reference temperatures for IT equipment and basic floor plan geometry, Avatar produces recommendation to adjust the operation of computer room air conditioners (CRACs) and the configuration of vent tiles in a data center so as to reducing excess provisioning of cooling and to remove hot spots. Avatar reduces operating expenses by cooling the same load with less energy. Avatar reduces capital expenses by recovering stranded cooling capacity that would otherwise have to be replaced.Copyright

Cooling capacity and cost metrics for energy efficient workload placement in datacenters

Energy usage in the data centers has been increasing drastically in the recent years attracting many studies in power and thermal management in data centers. However, research has been focused mainly at power management of IT equipment and efficiently delivering cooling resources for a given IT load distribution. To further increase the energy efficiency of data centers, an integrated approach to preferentially placing IT load at locations of highest cooling efficiency is necessary. Such cooling aware workload placement can bring substantial savings when coupled with the migration of workload away from less cooling-efficient locations and the powering down of machines that are not in active use. Along with cooling-aware migration, it is critical to quantify the available cooling capacity and the cooling cost to support additional heat load at different locations in a datacenter to affect placement of new workload and, from a longer term point of view, for IT planning purposes. This paper presents novel thermal metrics which can be used to estimate available cooling capacities for different thermal zones in an operational datacenter. The thermal metrics also show the local distribution of the available cooling capacity at different racks within a thermal zone. The paper will also present another new metric called localized cooling cost which is utilized to rank locations with respect to costs incurred for cooling additional workload. We utilize CFD simulations for a case study showing application of the metrics.

Exploring cyber physical data streams using Radial Pixel Visualizations

Cyber physical systems (CPS), such as smart buildings and data centers, are richly instrumented systems composed of tightly coupled computational and physical elements that generate large amounts of data. To explore CPS data and obtain actionable insights, we construct a Radial Pixel Visualization (RPV) system, which uses multiple concentric rings to show the data in a compact circular layout of small polygons (pixel cells), each of which represents an individual data value. RPV provides an effective visual representation of locality and periodicity of the high volume, multivariate data streams, and seamlessly combines them with the results of an automated analysis. In the outermost ring the results of correlation analysis and peak point detection are highlighted. Our explorations demonstrates how RPV can help administrators to identify periodic thermal hot spots, understand data center energy consumption, and optimize IT workload.

Visual analytics of cyber physical data streams using spatio-temporal radial pixel visualization

Cyber physical systems (CPS), such as smart buildings and data centers, are richly instrumented systems composed of tightly coupled computational and physical elements that generate large amounts of data. To explore CPS data and obtain actionable insights, we present a new approach called Radial Pixel Visualization (RPV); which uses multiple concentric rings to show the data in a compact circular layout of pixel cells, each ring containing the values for a specific variable over time and each pixel cell representing an individual data value at a specific time. RPV provides an effective visual representation of locality and periodicity of the high volume, multivariate data streams. RPVs may have an additional analysis ring for highlighting the results of correlation analysis or peak point detection. Our real-world applications demonstrate the effectiveness of this approach. The application examples show how RPV can help CPS administrators to identify periodic thermal hot spots, find root-causes of the cooling problems, understand building energy consumption, and optimize IT-services workloads.