Slide 1

Network Sharing Issues Lecture 15 Aditya Akella

Slide 2

Is this the biggest problem in cloud resource allocation? Why? Why not? How does the problem differ wrt allocating other resources? FairCloud: Sharing the Network in Cloud Computing, Sigcomm 2012 What are the assumptions? Drawbacks?

Slide 3

Motivation Network?

Slide 4

Context Networks are more difficult to share than other resources X

Slide 5

Context Several proposals that share network differently, e.g.: proportional to # source VMs (Seawall [NSDI11]) statically reserve bandwidth (Oktopus [Sigcomm12]) Provide specific types of sharing policies Characterize solution space and relate policies to each other?

Slide 6

FairCloud Paper 1.Framework for understanding network sharing in cloud computing Goals, tradeoffs, properties 2.Solutions for sharing the network Existing policies in this framework New policies representing different points in the design space

Slide 7

Goals 1.Minimum Bandwidth Guarantees Provides predictable performance Example: file transfer finishes within time limit A1A1 A2A2 Time max = Size / B min B min

Slide 8

Goals 1.Minimum Bandwidth Guarantees 2.High Utilization Do not leave useful resources unutilized Requires both work-conservation and proper incentives A BBB Both tenants activeNon work-conservingWork-conserving

Slide 9

Goals 1.Minimum Bandwidth Guarantees 2.High Utilization 3.Network Proportionality As with other services, network should be shared proportional to payment Currently, tenants pay a flat rate per VM network share should be proportional to #VMs (assuming identical VMs)

Slide 10

Goals 1.Minimum Bandwidth Guarantees 2.High Utilization 3.Network Proportionality Example: A has 2 VMs, B has 3 VMs A1A1 A2A2 Bw A B1B1 B3B3 B2B2 Bw B Bw A = 2 3 When exact sharing is not possible use max-min

Slide 11

Goals 1.Minimum Bandwidth Guarantees 2.High Utilization 3.Network Proportionality Not all goals are achievable simultaneously!

Slide 12

Tradeoffs Min Guarantee High Utilization Network Proportionality Not all goals are achievable simultaneously!

Slide 13

Tradeoffs Min Guarantee Network Proportionality

Slide 14

Tradeoffs Bw B Bw A AB Access Link L Capacity C Bw B = 11/13 C Bw A = 2/13 C 10 VMs Network Proportionality Bw A C/N T 0 #VMs in the network Bw B = 1/2 C Bw A = 1/2 C Minimum Guarantee Min Guarantee Network Proportionality

Slide 15

Tradeoffs High Utilization Network Proportionality

Slide 16

Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 High Utilization Network Proportionality

Slide 17

Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 Bw B = 1/2 C Bw A = 1/2 C Network Proportionality High Utilization Network Proportionality

Slide 18

Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 P High Utilization Network Proportionality Uncongested path

Slide 19

Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 Bw B Bw A < Network Proportionality L L Tenants can be disincentivized to use free resources If A values A 1 A 2 or A 3 A 4 more than A 1 A 3 Bw A +Bw A = Bw B L L P P High Utilization Network Proportionality Uncongested path

Slide 20

Network Proportionality Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 P Network proportionality applied only for flows traversing congested links shared by multiple tenants High Utilization Uncongested path Congestion Proportionality

Slide 21

Network Proportionality Tradeoffs L B1B1 B3B3 B2B2 B4B4 A1A1 A3A3 A2A2 A4A4 Uncongested path P Bw B Bw A = Congestion Proportionality L L High Utilization Congestion Proportionality

Slide 22

Network Proportionality Tradeoffs Still conflicts with high utilization High Utilization Congestion Proportionality

Slide 23

Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 L2L2 C 1 = C 2 = C Network Proportionality High Utilization L1L1 Congestion Proportionality

Slide 24

Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 L1L1 L2L2 Bw B Bw A = Congestion Proportionality L1 Bw B Bw A = L2 Network Proportionality High Utilization C 1 = C 2 = C Congestion Proportionality

Slide 25

Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 L1L1 L2L2 Demand drops to Network Proportionality High Utilization C 1 = C 2 = C Congestion Proportionality

Slide 26

Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 C - Tenants incentivized to not fully utilize resources Network Proportionality High Utilization C 1 = C 2 = C Congestion Proportionality L1L1 L2L2

Slide 27

Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 C - 2 Uncongested C - L1L1 L2L2 C 1 = C 2 = C Network Proportionality High Utilization Tenants incentivized to not fully utilize resources Congestion Proportionality

Slide 28

L2L2 Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 C - 2 C/2 Uncongested C 1 = C 2 = C Network Proportionality High Utilization Congestion Proportionality Tenants incentivized to not fully utilize resources L1L1

Slide 29

Tradeoffs Proportionality applied to each link independently Congestion Proportionality Link Proportionality Network Proportionality High Utilization L2L2 B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 L1L1

Slide 30

L2L2 Tradeoffs B1B1 B3B3 A1A1 A3A3 B2B2 B4B4 A2A2 A4A4 Full incentives for high utilization Congestion Proportionality Link Proportionality Network Proportionality High Utilization L1L1

Slide 31

Goals and Tradeoffs Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality

Slide 32

Guiding Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Break down goals into lower-level necessary properties

Slide 33

Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality

Slide 34

Work Conservation Bottleneck links are fully utilized Static reservations do not have this property

Slide 35

Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation

Slide 36

Utilization Incentives Tenants are not incentivized to lie about demand to leave links underutilized Network and congestion proportionality do not have this property Allocating links independently provides this property

Slide 37

Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives

Slide 38

Communication-pattern Independence Allocation does not depend on communication pattern Per flow allocation does not have this property (per flow = give equal shares to each flow) Same Bw

Slide 39

Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence

Slide 40

Symmetry Swapping demand directions preserves allocation Per source allocation lacks this property (per source = give equal shares to each source) Same Bw

Slide 41

Goals, Tradeoffs, Properties Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 42

Outline 1.Framework for understanding network sharing in cloud computing Goals, tradeoffs, properties 2.Solutions for sharing the network Existing policies in this framework New policies representing different points in the design space

Slide 43

Per Flow (e.g. today) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 44

Per Source (e.g., Seawall [NSDI11]) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 45

Static Reservation (e.g., Oktopus [Sigcomm11]) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 46

New Allocation Policies 3 new allocation policies that take different stands on tradeoffs

Slide 47

Proportional Sharing at Link-level (PS-L) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 48

Proportional Sharing at Link-level (PS-L) Per tenant WFQ where weight = # tenants VMs on link A B WQ A = #VMs A on L Bw B Bw A = #VMs A on L #VMs B on L Can easily be extended to use heterogeneous VMs (by using VM weights)

Slide 49

Proportional Sharing at Network-level (PS-N) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 50

Proportional Sharing at Network-level (PS-N) Congestion proportionality in severely restricted context Per source-destination WFQ, total tenant weight = # VMs

Slide 51

Proportional Sharing at Network-level (PS-N) WQ A1 A2 = 1/N A1 + 1/N A2 A1A1 A2A2 N A2 N A1 Total WQ A = #VMs A Congestion proportionality in severely restricted context Per source-destination WFQ, total tenant weight = # VMs

Slide 52

Proportional Sharing at Network-level (PS-N) WQ B WQ A = #VMs A #VMs B Congestion proportionality in severely restricted context Per source-destination WFQ, total tenant weight = # VMs

Slide 53

Proportional Sharing on Proximate Links (PS-P) Min Guarantee Link Proportionality High Utilization Congestion Proportionality Network Proportionality Work Conservation Utilization Incentives Comm-Pattern Independence Symmetry

Slide 54

Assumes a tree-based topology: traditional, fat-tree, VL2 (currently working on removing this assumption) Proportional Sharing on Proximate Links (PS-P)

Slide 55

Assumes a tree-based topology: traditional, fat-tree, VL2 (currently working on removing this assumption) Min guarantees Hose model Admission control Proportional Sharing on Proximate Links (PS-P) A1A1 Bw A1 A2A2 Bw A2 AnAn Bw An

Slide 56

Assumes a tree-based topology: traditional, fat-tree, VL2 (currently working on removing this assumption) Min guarantees Hose model Admission control High Utilization Per source fair sharing towards tree root Proportional Sharing on Proximate Links (PS-P)

Slide 57

Assumes a tree-based topology: traditional, fat-tree, VL2 (currently working on removing this assumption) Min guarantees Hose model Admission control High Utilization Per source fair sharing towards tree root Per destination fair sharing from tree root Proportional Sharing on Proximate Links (PS-P)

Slide 58

Deploying PS-L, PS-N and PS-P Full Switch Support All allocations can use hardware queues (per tenant, per VM or per source-destination) Partial Switch Support PS-N and PS-P can be deployed using CSFQ [Sigcomm98] No Switch Support PS-N can be deployed using only hypervisors PS-P could be deployed using only hypervisors, we are currently working on it

Slide 59

Evaluation Small Testbed + Click Modular Router 15 servers, 1Gbps links Simulation + Real Traces 3200 nodes, flow level simulator, Facebook MapReduce traces

Slide 60

Many to one Bw B Bw A AB N One link, testbed PS-P offers guarantees Bw A N

Slide 61

MapReduce One link, testbed Bw B Bw A 5 R5 M M+R = 10 M Bw B (Mbps) PS-L offers link proportionality

Slide 62

MapReduce Network, simulation, Facebook trace

Slide 63

MapReduce Network, simulation, Facebook trace PS-N is close to network proportionality

Slide 64

MapReduce Network, simulation, Facebook trace PS-N and PS-P reduce shuffle time of small jobs by 10-15X

Slide 65

Summary Sharing cloud networks is not trivial First step towards a framework to analyze network sharing in cloud computing Key goals (min guarantees, high utilization and proportionality), tradeoffs and properties New allocation policies, superset properties from past work PS-L: link proportionality + high utilization PS-N: restricted network proportional PS-P: min guarantees + high utilization What are the assumptions, drawbacks?