Optimizing Flow Bandwidth Consumption with Traffic-diminishing - - PowerPoint PPT Presentation
Optimizing Flow Bandwidth Consumption with Traffic-diminishing Middlebox Placement Yan Yang Chen en, Jie Wu, and Bo Ji Center for Networked Computing Temple University, USA VNF: Evolution of Network Service l Network Function Virtualization
l Network Function Virtualization (NFV)
¡ Virtualizing network functions into software building blocks
l Virtualized Network Function (VNF) or Middlebox
¡ Software implementation of network functions ¡ Improve performance & enhance security
l Examples l Middlebox Deployment
¡ Deployment location selection on multiple servers
Firewall NAT Proxy
l VNFs may change flow rates in different ways
¡ Citrix CloudBridge WAN accelerator: 20% (diminishing) ¡ BCH(63,48) encoder: 130% (expanding)
[1] Traffic Aware Placement of Interdependent NFV Middleboxes (INFOCOM ’17)
1 1 1 1 1 Data Checksum
¡ Deploy a single type of VNFs with traffic-
¡ Minimize total bandwidth consumption of all flows
¡ Each flow gets processed ¡ Deploy a limited number of the single type of VNFs
l NP-hard l Decrement function
¡ Decrement of total bandwidth consumption compared to no VNFs
l Marginal decrement
¡ Additional bandwidth decrement by deploying on 𝒯 beyond 𝒬
l Decrement function is submodular
¡ More VNFs, less bandwidth consumption ¡ Flow gets processed no later than 𝒬
¡ General Topology Placement (GTP)
¡ Iteratively select v ∈ V with the maximum marginal
¡ O(|V|2 log |V |)
l 𝐺(𝑤, 𝑙)
¡ Minimum total occupied bandwidth of all flows with 𝑙 deployed
¡ All flows get fully processed in Tv
l 𝑄(𝑤, 𝑙, 𝑐)
¡ Same as F(v,k) ¡ When flows with only a total bandwidth 𝑐 processed
l Optimal solution l Time complexity (|V|: #node, 𝑠
!"#: largest flow rate) ¡ 𝑃(|𝑊| (log |𝑊|)!𝑠
"#$)
¡ LCA(v,w): lowest vertex have both v and w as descendants
¡ Deploy one VNF on each leaf vertex ¡ Delete two VNFs on v and w with minimum difference of
¡ Place one VNF on LCA(v,w) ¡ Until total number of deployed VNFs no more than k
¡ Min-heap ¡ Improve time efficiency
¡ O(|V |2 log |V |) ¡ |V|: #vertices
¡ Random
l Randomly deploy k VNFs
¡ Best-effort
l Deploy on the vertex, which can reduce the total
¡ General topo
l Alg. GTP
¡ Tree topo
l Algs. GTP, DP, HAT
l Topology l Middlebox traffic-diminishing ratio
¡ From 0 (e.g., spam filters) to 0.9 (e.g., traffic optimizer) with a
stride of 0.1
¡ Additional simulation on spam filter
l Flow rate distribution
¡ CAIDA data center 1-hour packet trace
l Alg. DP performs
l k = 1, only one
l Traffic-changing
l Random has the
Tree Topology
15 20 25 30 Topology size 0.5 1 1.5 2 2.5 3 Bandwidth consumption 105
Random Best-effort GTP HAT DP
5 10 15 k 0.6 0.8 1 1.2 1.4 1.6 Bandwidth consumption 105
Random Best-effort GTP HAT DP
0.2 0.4 0.6 0.8 Traffic-changing ratio 0.5 1 1.5 2 2.5 Bandwidth consumption 105
Random Best-effort GTP HAT DP
0.3 0.4 0.5 0.6 0.7 0.8 Flow density 0.8 1 1.2 1.4 1.6 1.8 2 Bandwidth consumption 105
Random Best-effort GTP HAT DP
l Alg. GTP always
l Error bars become
l Bandwidth consumption
l When flow density is
General Topology
12 14 16 18 20 22 k 3.5 4 4.5 5 5.5 Bandwidth consumption 105
Random Best-effort GTP
0.2 0.4 0.6 0.8 Traffic-changing ratio 3 3.2 3.4 3.6 3.8 4 Bandwidth consumption 105
Random Best-effort GTP
0.3 0.4 0.5 0.6 0.7 0.8 Flow density 2.0 4.0 6.0 8.0 Bandwidth consumption 105
Random Best-effort GTP
20 30 40 50 Topology size 2.0 4.0 6.0 8.0 Bandwidth consumption 105
Random Best-effort GTP
l Flow density plays a more important role in affecting
l When flow density doubles from 0.3 to 0.6,
Spam Filter (Traffic diminishing ratio: 0)
l Problem
¡ Deploy a limited number of traffic-diminishing VNFs ¡ All flows get processed
l Objective
¡ Minimize total bandwidth consumption
l Solutions
¡ Tree: optimal and greedy ¡ General graph: performance-guaranteed
l Future Work
¡ Traffic-expanding VNFs ¡ Service chain: an ordered set of multiple VNFs