Showing papers by "Georgios Zervas published in 2018"

Space-division multiplexing in data center networks: on multi-core fiber solutions and crosstalk-suppressed resource allocation

Abstract: The rapid growth of traffic inside data centers caused by the increasing adoption of cloud services necessitates a scalable and cost-efficient networking infrastructure. Space-division multiplexing (SDM) is considered as a promising solution to overcome the optical network capacity crunch and support cost-effective network capacity scaling. Multi-core fiber (MCF) is regarded as the most feasible and efficient way to realize SDM networks, and its deployment inside data centers seems very likely as the issue of inter-core crosstalk (XT) is not severe over short link spans (<1 km) compared to that in long-haul transmission. However, XT can still have a considerable effect in MCF over short distances,which can limit the transmission reach and in turn the data center’s size. XT can be further reduced by bi-directional transmission of optical signals in adjacent MCF cores. This paper evaluates the benefits of MCF-based SDM solutions in terms of maximizing the capacity and spatial efficiency of data center networks. To this end, we present an analytical model for XT in bi-directional normal step-index and trench-assisted MCFs and propose corresponding XT-aware core prioritization schemes. We further develop XT-aware spectrum resource allocation strategies aimed at relieving the complexity of online XT computation. These strategies divide the available spectrum into disjoint bands and incrementally add them to the pool of accessible resources based on the network conditions. Several combinations of core mapping and spectrum resource allocation algorithms are investigated for eight types of homogeneous MCFs comprising 7–61 cores, three different multiplexing schemes, and three data center network topologies with two traffic scenarios. Extensive simulation results showthat combining bi-directional transmission in dense core fibers with tailored resource allocation schemes significantly increases the network capacity. Moreover, a multiplexing scheme that combines SDM and WDM can achieve up to 33 times higher link spatial efficiency and up to 300 times greater capacity compared to a WDM solution.

Optically disaggregated data centers with minimal remote memory latency: Technologies, architectures, and resource allocation [Invited]

Abstract: Disaggregated rack-scale data centers have been proposed as the only promising avenue to break the barrier of the fixed CPU-to-memory proportionality caused by main-tray direct-attached conventional/traditional server-centric systems However, memory disaggregation has stringent network requirements in terms of latency, energy efficiency, bandwidth, and bandwidth density This paper identifies all the requirements and key performance indicators of a network to disaggregate IT resources while summarizing the progress and importance of optical interconnects Crucially, it proposes a rack-and-cluster scale architecture, which supports the disaggregation of CPU, memory, storage, and/or accelerator blocks Optical circuit switching forms the core of this architecture, whereas the end-points (IT resources) are equipped with on-chip programmable hybrid electrical packet/circuit switches This architecture offers dynamically reconfigurable physical topology to form virtual ones, each embedded with a set of functions It analyzes the latency overhead of disaggregated DDR4 (parallel) and the proposed hybrid memory cube (serial) memory elements on the conventional and the proposed architecture A set of resource allocation algorithms are introduced to (1) optimally select disaggregated IT resources with the lowest possible latency, (2) pool them together by means of a virtual network interconnect, and (3) compose virtual disaggregated servers Simulation findings show up to a 34% resource utilization increase over traditional data centers while highlighting the importance of the placement and locality among compute, memory, and storage resources In particular, the network-aware locality-based resource allocation algorithm achieves as low as 15 ns, 95 ns, and 315 ns memory transaction round-trip latency on 63%, 22%, and 15% of the allocated virtual machines (VMs) accordingly while utilizing 100% of the CPU resources Furthermore, a formulation to parameterize and evaluate the additional financial costs endured by disaggregation is reported It is shown that the more diverse the VM requests are, the higher the net financial gain is Finally, an experiment was carried out using silicon photonic midboard optics and an optical circuit switch, which demonstrates forward error correction free 10−12 bit error rate performance on up to five-tier scale-out networks

You get what you give: theory and evidence of reciprocity in the sharing economy

Abstract: We develop an analytical framework of peer interaction in the sharing economy that incorporates reciprocity, the tendency to increase (decrease) effort in response to others’ increased (decreased) effort. In our model, buyers (sellers) can induce sellers (buyers) to exert more effort by behaving well themselves. We demonstrate that this joint increased effort can improve the utility of both parties and influence the market equilibrium. We also show that bilateral reputation systems, which allow both buyers and sellers to review each other, are more responsive to reciprocity than unilateral reputation systems. By rewarding reciprocal behavior, bilateral reputation systems generate trust among strangers and informally regulate their behavior. We test the predictions of our model using data from Airbnb, a popular peer-to-peer accommodation platform. We show that Airbnb hosts that are more reciprocal receive higher ratings and that higher rated hosts can increase their prices. Therefore, reciprocity affects equilibrium prices on Airbnb through its impact on ratings, as predicted by our analytical framework.

Sub-Nanosecond Clock and Data Recovery in an Optically-Switched Data Centre Network

Abstract: We demonstrate a clock and data recovery technique that achieves <625ps locking time for 25.6Gb/s-OOK and show its robustness under worst-case data centre temperature variation. The locking time was improved by 12×, making nanosecond optical switching viable in data centres.

dReDBox: Materializing a full-stack rack-scale system prototype of a next-generation disaggregated datacenter

Abstract: Current datacenters are based on server machines, whose mainboard and hardware components form the baseline, monolithic building block that the rest of the system software, middleware and application stack are built upon. This leads to the following limitations: (a) resource proportionality of a multi-tray system is bounded by the basic building block (mainboard), (b) resource allocation to processes or virtual machines (VMs) is bounded by the available resources within the boundary of the mainboard, leading to spare resource fragmentation and inefficiencies, and (c) upgrades must be applied to each and every server even when only a specific component needs to be upgraded. The dRedBox project (Disaggregated Recursive Datacentre-in-a-Box) addresses the above limitations, and proposes the next generation, low-power, across form-factor datacenters, departing from the paradigm of the mainboard-as-a-unit and enabling the creation of function-block-as-a-unit. Hardware-level disaggregation and software-defined wiring of resources is supported by a full-fledged Type-1 hypervisor that can execute commodity virtual machines, which communicate over a low-latency and high-throughput software-defined optical network. To evaluate its novel approach, dRedBox will demonstrate application execution in the domains of network functions virtualization, infrastructure analytics, and real-time video surveillance.

dRedDbox: Demonstrating Disaggregated Memory in an Optical Data Centre

Abstract: This paper showcases the first experimental demonstration of disaggregated memory using the dRedDbox optical Data Centre architecture. Experimental results demonstrate the 4-tier network scalability and performance of the system at the physical and application layer.

Disaggregated Optical Data Center in a Box Network using Parallel OCS Topologies

Abstract: Two parallel OCS topologies are proposed that deliver 95 nsec round-trip latency on disaggregated optical data center in a box system. They offer 40% cost and 68% power consumption efficiency at maximum IT resource utilization.

In compute/memory dynamic packet/circuit switch placement for optically disaggregated data centers

Abstract: Network function services on conventional hybrid data center (DC) architectures such as HELIOS arehard-wiredanddedicatedto specificnetworkresources. This limits flexibility and performance to handle diverse traffic patterns. Furthermore, disaggregation of server resources has shown promising potential to improve resource utilization, which has been a limitation of conventional server-centric DCs. This paper presents a reconfigurable hybrid disaggregated DC (dRedBox) architecture that combines the concept of server resource disaggregation with cutting-edge software and electronic and optical technologies. ThedRedBoxarchitecture provides a remarkable amount of flexibility and connectivity through hardware-based multilayer network function service programmability. This allows for multilayer network services to be dynamically deployed at runtime to network resources and, in turn, handle diverse traffic patterns. Furthermore, this study proposes algorithms and strategies for selecting and deploying electronic packet switching and optical circuit switching function services to implement virtual machine network requests across dRedBox and conventional hybrid disaggregated architectures under different traffic patterns. Finally, the performance of the various strategies on the dRedBox and conventional hybrid disaggregated DC architectures is evaluated in terms of blocking probability, energy efficiency, network utilization, and cost. Extensive results show that, at 10% blocking probability, dRedBox architecture achieves 100% gain on VM placement and 92% energy savings compared with conventional hybrid disaggregated architectures.

Comparison of SDM-WDM based Data Center Networks with equal/unequal core pitch Multi-Core Fibers

Abstract: A wavelength-dependent crosstalk calculation formula for bi-directional MCF with unequal core pitches is derived. SDM-WDM based DCN simulations indicate that MCF core density and layout play a significant role on optimizing various scales of DCNs.

Parallel Distributed Schedulers for Scalable Photonic Integrated Packet Switching

Abstract: As data centre traffic dynamics are changing, optical networking is becoming increasingly important for low-latency, high bandwidth intra-data centre communication. Nanosecond-reconfigurable, scalable photonic switch fabrics and advances in photonic integration are key enablers for optical packet switching. However, the control plane and in particular the switch scheduler is believed to be a critical factor on packet latency and scalability. To that end, we report a low-latency scheduler for Clos-network switch fabrics based on a fixed path assignment scheme and parallel and distributed path arbitration. Cycle-accurate network emulation results show nanosecond average latency at input port loads up to 60% of capacity for a 256x256 switch size. In comparison to previous work, the scheduler can now control a switch 8 times the size at double the input port saturation load for the same average latency. Scaling the switch from 16 to 256 ports shows only a small drop in saturation load from 70% to 60%. Also fairness on a per flow basis is demonstrated.

LIGHTNESS: All-Optical SDN-enabled Intra-DCN with Optical Circuit and Packet Switching

Abstract: The proposed LIGHTNESS DCN architecture aims at a flat all-optical interconnection design by combining optical circuit switching (OCS) and optical packet switching (OPS) technologies. LIGHTNESS brings innovation into current data center networks (DCNs) through the introduction of a hybrid OCS/OPS data plane that drives nowadays’ multitier hierarchical architectures toward a flatter scheme for enhanced scalability, throughput, and latency. The design of the OPS node targets medium-port count (4 × 4 or 16 × 16) and low latency, which is operated for switching short-lived packet flows (also known as mice flows). On the other hand, the OCS node mainly handles long-lived data flows (known as elephant flows) and has much higher reconfiguration time than the OPS one. Transparent optical switching elements also bypass the use of expensive and power-hungry optical-electrical-optical (O/E/O) conversions and massive electrical cabling, improving the energy efficiency and cost efficiency of present EPS-only solutions. The introduced software-defined networking (SDN) control plane enables the virtualization of computing and network resources creating a virtual data center (VDC) and virtual network functions (VNFs) on top of the data plane.

Reconfigurable Network Stream Processing on Virtualized FPGA Resources

Abstract: The software defined network and network function virtualization are proposed to address the network ossification issue in current Internet infrastructure. Network functions and services are implemented as software applications to increase the programmability of network. However, involving general purpose processors in data plane restricts the bandwidth of network services. Therefore, to keep both the bandwidth and flexibility, a FPGA platform is suggested as a reconfigurable platform to deliver high bandwidth virtual network functions on data plane. In this paper, the FPGA resource has been virtualized by interconnecting partial reconfigurable regions to deliver high bandwidth reconfigurable processing on network streams. With the help of partial reconfiguration technology, network functions on our platform can be configured without affecting other functions on the same FPGA device. The on-chip interconnect system is further evaluated by comparing with existing network-on-chip system. A reconfiguration process is also proposed and demonstrated that it can be performed on our platform. The process can happen in the real time of network services and it is able to keep the original function working during the download of partial bitstream.

Demonstration of NFV for Mobile Edge Computing on an Optically Disaggregated Datacentre in a Box

Abstract: This demonstrator showcases the hardware and software integration achieved by the dReDBox project [1] towards realization of a novel architecture using dynamically-reconfigurable optical interconnects to create a flexible, scalable and efficient disaggregated datacentre infrastructure.