data center
Recently Published Documents

At present, a co-location data center often applies an identical and low temperature setpoint for its all server rooms. Although increasing the temperature setpoint is a rule-of-thumb approach to reducing the cooling energy usage, the tenants may have different mentalities and technical constraints in accepting higher temperature setpoints. Thus, supporting distinct temperature setpoints is desirable for a co-location data center in pursuing higher energy efficiency. This calls for a new cooling power attribution scheme to address the inter-room heat transfers that can be up to 9% of server load as shown in our real experiments. This article describes our approaches to estimating the inter-room heat transfers, using the estimates to rectify the metered power usages of the rooms’ air handling units, and fairly attributing the power usage of the shared cooling infrastructure (i.e., chiller and cooling tower) to server rooms by following the Shapley value principle. Extensive numeric experiments based on a widely accepted cooling system model are conducted to evaluate the effectiveness of the proposed cooling power attribution scheme. A case study suggests that the proposed scheme incentivizes rational tenants to adopt their highest acceptable temperature setpoints under a non-cooperative game setting. Further analysis considering distinct relative humidity setpoints shows that our proposed scheme also properly and inherently addresses the attribution of humidity control power.

Many studies focusing on improving Transmission Control Protocol (TCP) flow control realize a more effective use of bandwidth in data center networks. They are excellent ways to more effectively use the bandwidth between clients and back-end servers. However, these schemes cannot achieve the total optimization of bandwidth use for data center networks as they do not take into account the path design of TCP flows against a hierarchical complex structure of data center networks. To address this issue, this paper proposes a TCP flow management scheme specified a hierarchical complex data center network for effective bandwidth use. The proposed scheme dynamically controls the paths of TCP flows by reinforcement learning based on a hierarchical feedback model, which obtains an optimal TCP flow establishment policy even if both the network topology and link states are more complicated. In evaluation, the proposed scheme achieved more effective bandwidth use and reduced the probability of TCP incast up to 30% than the conventional TCP flow management schemes: Variant Load Balancing (VLB), Equal Cost Multi Path (ECMP), and Intelligent Forwarding Strategy Based on Reinforcement Learning (IFS-RL) in the complex data center network.

Abstract Modern data center topologies often take the form of a multi-rooted tree with rich parallel paths to provide high bandwidth. However, various path diversities caused by traffic dynamics, link failures and heterogeneous switching equipments widely exist in production data center network. Therefore, the multi-path load balancer in data center should be robust to these diversities. Although prior fine-grained schemes such as RPS and Presto make full use of available paths, they are prone to experi-ence packet reordering problem under asymmetric topology. The coarse-grained solutions such as ECMP and LetFlow effectively avoid packet reordering, but easily lead to under-utilization of multiple paths. To cope with these inefficiencies, we propose a load balancing mechanism called PDLB, which adaptively adjusts flowcell granularity according to path diversity. PDLB increases flowcell granularity to alleviate packet reordering under large degrees of topology asymmetry, while reducing flowcell granularity to obtain high link utilization under small degrees of topology asymmetry. PDLB is only deployed on the sender without any modification on switch. We evaluate PDLB through large-scale NS2 simulations. The experimental results show that PDLB reduces the average flow completion time by up to ∼11-53% over the state-of-the-art load balancing schemes.

This paper studies the problems of load balancing and flow control in data center network, and analyzes several common flow control schemes in data center intelligent network and their existing problems. On this basis, the network traffic control problem is modeled with the goal of deep reinforcement learning strategy optimization, and an intelligent network traffic control method based on deep reinforcement learning is proposed. At the same time, for the flow control order problem in deep reinforcement learning algorithm, a flow scheduling priority algorithm is proposed innovatively. According to the decision output, the corresponding flow control and control are carried out, so as to realize the load balance of the network. Finally, experiments show, the network traffic bandwidth loss rate of the proposed intelligent network traffic control method is low. Under the condition of random 60 traffic density, the average bisection bandwidth obtained by the proposed intelligent network traffic control method is 4.0mbps and the control error rate is 2.25%. The intelligent network traffic control method based on deep reinforcement learning has high practicability in the practical application process, and fully meets the research requirements.

Abstract. We report on updated trends using different merged zonal mean total ozone datasets from satellite and ground-based observations for the period from 1979 to 2020. This work is an update from the trends reported in Weber et al. (2018) using the same datasets up to 2016. Merged datasets used in this study include NASA MOD v8.7 and NOAA Cohesive Data (COH) v8.6, both based on data from the series of Solar Backscatter UltraViolet (SBUV), SBUV-2, and Ozone Mapping and Profiler Suite (OMPS) satellite instruments (1978–present) as well as the Global Ozone Monitoring Experiment (GOME)-type Total Ozone (GTO-ECV) and GOME-SCIAMACHY-GOME-2 (GSG) merged datasets (both 1995–present), mainly comprising satellite data from GOME, SCIAMACHY, OMI, GOME-2A, -2B, and TROPOMI. The fifth dataset consists of the annual mean zonal mean data from ground-based measurements collected at the World Ozone and UV Radiation Data Center (WOUDC). Trends were determined by applying a multiple linear regression (MLR) to annual mean zonal mean data. The addition of four more years consolidated the fact that total ozone is indeed on slowly recovering in both hemispheres as a result of phasing out ozone depleting substances (ODS) as mandated by the Montreal Protocol. The near global ozone trend of the median of all datasets after 1996 was 0.5 ± 0.2 (2σ) %/decade, which is in absolute numbers roughly a third of the decreasing rate of 1.4 ± 0.6 %/decade from 1978 until 1996. The ratio of decline and increase is nearly identical to that of the EESC (equivalent effective stratospheric chlorine or stratospheric halogen) change rates before and after 1996 which confirms the success of the Montreal Protocol. The observed trends are also in very good agreement with the median of 17 chemistry climate models from CCMI (Chemistry Climate Model Initiative) with current ODS and GHG (greenhouse gas) scenarios. The positive ODS related trends in the NH after 1996 are only obtained with a sufficient number of terms in the MLR accounting properly for dynamical ozone changes (Brewer-Dobson circulation, AO, AAO). A standard MLR (limited to solar, QBO, volcanic, and ENSO) leads to zero trends showing that the small positive ODS related trends have been balanced by negative trend contributions from atmospheric dynamics resulting in nearly constant total ozone levels since 2000.