Dynamic resource allocation in hybrid optical-electrical datacenter networks

Joint problem of bandwidth allocation and VM-placement in optical-electrical DC.Challenges: (i) limited reachability due to the size of the optical switch in the datacenter netework, (ii) dynamic creation of optical paths lead to the creation of constrained optical network topology.The problem is formulated as nonlinear optimization problem, that is NP-hard in nature.Input requests from users are abstracted as virtual networks, where a single node in a virtual network is a cluster of VMs, and an edge connecting two nodes is the bandwidth demand between the two clusters of VMs.Three algorithms are developed for embedding input virtual networks on hybrid datacenter network, and their performances are evaluated using simulations. A promising development in the design of datacenters is the hybrid network architecture consisting of both optical and electrical elements, in which end-to-end traffic can be routed through either an electrical path or an optical path. The core optical switch is used to dynamically create optical paths between pairs of electrical edge-switches in such a datacenter network. In this context, the joint problem of bandwidth allocation and VM-placement poses new and different challenges not addressed yet in hybrid datacenter. In particular, we foresee two issues: (i) the number of edge-switches that can be simultaneously reached using optical paths from an edge-switch is limited by the size of the optical switch, (ii) the dynamic creation of optical paths can potentially establish a constrained optical network topology leading to poor performance. In this work, we abstract the requests of tenants as virtual networks, and study the problem of embedding virtual networks on a hybrid datacenter. We formulate the problem as a non-linear optimization problem and analyze its complexity. We develop and analyse three algorithms for embedding dynamically arriving virtual network demands on a hybrid optical-electrical datacenter. Through simulations, we demonstrate the effectiveness of not only exploiting the already established optical paths, but also of using electrical network in embedding requests of virtual networks.

[1]  V. Deineko,et al.  The Quadratic Assignment Problem: Theory and Algorithms , 1998 .

[2]  Dinil Mon Divakaran,et al.  Bandwidth allocation with differential pricing for flexible demands in data center networks , 2014, Comput. Networks.

[3]  Harold N. Gabow,et al.  An efficient reduction technique for degree-constrained subgraph and bidirected network flow problems , 1983, STOC.

[4]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[5]  Konstantina Papagiannaki,et al.  c-Through: part-time optics in data centers , 2010, SIGCOMM 2010.

[6]  Dinil Mon Divakaran,et al.  Dynamic embedding of virtual networks in hybrid optical-electrical datacenters , 2014, 2014 23rd International Conference on Computer Communication and Networks (ICCCN).

[7]  Dinil Mon Divakaran,et al.  Virtual Network Embedding in Hybrid Datacenters with Dynamic Wavelength Grouping , 2014, 2014 IEEE 6th International Conference on Cloud Computing Technology and Science.

[8]  Dorgival O. Guedes,et al.  Gatekeeper: Supporting Bandwidth Guarantees for Multi-tenant Datacenter Networks , 2011, WIOV.

[9]  R. Nejabati,et al.  Enabling multi-tenancy in hybrid optical packet/circuit switched data center networks , 2014, 2014 The European Conference on Optical Communication (ECOC).

[10]  Amin Vahdat,et al.  Helios: a hybrid electrical/optical switch architecture for modular data centers , 2010, SIGCOMM '10.

[11]  Mei-Hui Su,et al.  Characterization of scientific workflows , 2008, 2008 Third Workshop on Workflows in Support of Large-Scale Science.

[12]  Tara Javidi,et al.  End-to-end scheduling for all-optical data centers , 2015, 2015 IEEE Conference on Computer Communications (INFOCOM).

[13]  William J. Cook,et al.  Combinatorial optimization , 1997 .

[14]  Sujata Banerjee,et al.  Application-driven bandwidth guarantees in datacenters , 2014, SIGCOMM.

[15]  Kostas Katrinis,et al.  Performance evaluation of a hybrid optical/electrical interconnect , 2015, IEEE/OSA Journal of Optical Communications and Networking.

[16]  Tajana Simunic,et al.  Dynamic optical switching for latency sensitive applications , 2015, 2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS).

[17]  Hitesh Ballani,et al.  Towards predictable datacenter networks , 2011, SIGCOMM 2011.

[18]  Dinil Mon Divakaran,et al.  An integrated resource allocation scheme for multi-tenant data-center , 2012, 37th Annual IEEE Conference on Local Computer Networks.

[19]  Dinil Mon Divakaran,et al.  Towards Flexible Guarantees in Clouds: Adaptive Bandwidth Allocation and Pricing , 2015, IEEE Transactions on Parallel and Distributed Systems.

[20]  Vasileios Pappas,et al.  Improving the Scalability of Data Center Networks with Traffic-aware Virtual Machine Placement , 2010, 2010 Proceedings IEEE INFOCOM.

[21]  Helen J. Wang,et al.  SecondNet: a data center network virtualization architecture with bandwidth guarantees , 2010, CoNEXT.

[22]  Minghua Chen,et al.  Joint VM placement and routing for data center traffic engineering , 2012, 2012 Proceedings IEEE INFOCOM.

[23]  Ankit Singla,et al.  OSA: An Optical Switching Architecture for Data Center Networks With Unprecedented Flexibility , 2012, IEEE/ACM Transactions on Networking.

[24]  Dinil Mon Divakaran,et al.  An Online Integrated Resource Allocator for Guaranteed Performance in Data Centers , 2014, IEEE Transactions on Parallel and Distributed Systems.

[25]  Sujata Banerjee,et al.  CloudMirror: Application-Aware Bandwidth Reservations in the Cloud , 2013, HotCloud.

[26]  Albert G. Greenberg,et al.  Sharing the Data Center Network , 2011, NSDI.