Virtual Network Embedding with Collocation Benefits and Limitations of Pre-Clustering urst 1 , Stefan Schmid 2 , Anja Feldmann 1 Carlo F¨ 1: TU Berlin 2: TU Berlin & Telekom Innovation Laboratories November 12, 2013 Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 1 / 21
Today’s Datacenters... Multi-tenant virtualized Tenants typically pay for host resources Connectivity is guaranteed Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 2 / 21
Today’s Datacenters... Multi-tenant virtualized Tenants typically pay for host resources Connectivity is guaranteed Problem [Ballani’11]: Studies have shown that the intra-cloud bandwidth can vary by an order of magnitude. ⇒ Unpredictable application performance Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 2 / 21
Remove the uncertainty ? ? Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 3 / 21
Remove the uncertainty Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 3 / 21
Outline Explain model and problem Identify the impact of the collocation option on embedding algorithms Introduce Pre-Clustering - a technique to enable any existing algorithm to generate collocated embeddings Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 4 / 21
Virtual Network Embedding Problem Physical Machine Physical Link Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
Virtual Network Embedding Problem Physical Machine - Abstract aggregated "Compute Resource" Physical Link - Bandwidth Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
Virtual Network Embedding Problem Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
Virtual Network Embedding Problem Virtual Link Virtual Node Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
Virtual Network Embedding Problem Virtual Link - Requested Compute Units Virtual Node - Requested Bandwidth Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
Virtual Network Embedding Problem Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 5 / 21
What is a ‘good’ mapping? Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 6 / 21
What is a ‘good’ mapping? Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 6 / 21
Existing Solutions Many existing mapping algorithms ViNE [CHOWDHURY, Infocom 2009] SecondNet [GUO, Co-NEXT 2010] Oktopus [BALLANI, Sigcomm 2011] Isomorphism Detection [LISCHKA, Sigcomm 2009] Various Mixed-Integer-Programs . . . Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 7 / 21
Existing Solutions Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 7 / 21
Existing Solutions Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 7 / 21
Collocated Mappings Physical Machine with capacity 2 Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 8 / 21
Collocated Mappings Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 8 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Benchmarking Algorithm: LoCo Backtrack on failure Backtrack only over possible start nodes Graph exploration is directed by node / link resource requests Avoid Backtracking by forward checking Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 9 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Add Requests Until: Sum of requested node resources = Sum of substrate node resources Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Measure node utilization Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Increase time until a Request expires Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Add Requests Until: ... Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Substrate Topologies FatTree Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Evaluation Setup ADD REQ1 ADD REQ2 ADD REQ3 REM REQ1 ADD REQ4 STATE Request ... sequence Substrate Topologies FatTree Unmodified Requests Embed. SNet LoCo Algorithm Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 10 / 21
Impact of the collocation option Slight Impact 1.0 ● ● ● ● ● ● 0.8 ● Node Utilization ● 0.6 0.4 0.2 0.0 Loco SecondNet Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 11 / 21
Impact of the collocation option Slight Impact Strong Impact 1.0 1.0 ● ● ● ● ● ● 0.8 ● 0.8 Node Utilization ● 0.6 0.6 0.4 0.4 0.2 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.0 0.0 Loco SecondNet Loco SecondNet Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 11 / 21
Impact of the collocation option Slight Impact Strong Impact Average Impact 1.0 1.0 1.0 ● ● ● ● ● ● 0.8 ● 0.8 0.8 Node Utilization ● ● ● ● 0.6 0.6 0.6 ● 0.4 0.4 0.4 ● ● 0.2 0.2 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.0 0.0 0.0 Loco SecondNet Loco SecondNet Loco SecondNet Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 11 / 21
Can we leverage the benefits of collocation with the existing algorithms?
Pre-Clustering Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 13 / 21
Pre-Clustering Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 13 / 21
Pre-Clustering 4 4 3 1 Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 13 / 21
Pre-Clustering We use: Farhat LoCo OptCut (runtime optimized MIP) Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 13 / 21
LoCo Preclustering Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 14 / 21
LoCo Preclustering Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 14 / 21
LoCo Preclustering 2 Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 14 / 21
OptCut Generates an optimal (w.r.t. the amount of link resources between the merged nodes) Pre-Clustering Substrate is represented by two numbers: ◮ MAX V : The estimated host resources of a node ◮ MAX E : The estimated link resources attached to a node ⇒ run time independent of substrate size and topology Removes symmetry from the problem to speed up the solution process Carlo F¨ urst (TU Berlin) Virtual Network Embedding with Collocation November 12, 2013 15 / 21
Recommend
More recommend