Running large scale experimentation on Content-Centric Networking via the Grid’5000 platformMassimo GALLO (Bell Labs, Alcatel - Lucent)

Joint work with:Luca Muscariello (Orange)Giovanna Carofiglio (Bell Labs, Alcatel - Lucent)

Agenda

ICN

Lurch

Experiments

Conclusions and future works

ICN

Running large scale experimentation on Content-Centric Networking via the Grid’5000 platform

ICN

Today’s Internet

ICN advantages

imdb.com

www.imdb.com/title/tt12242

www.imdb.com/title/tt12242

Ever-growing amount of digital info Point-to-point dissemination Mobility issues Waste of resources in content

replication

Simplified management Traffic reduction and localization Seamless, ubiquitous

connectivity Congestion reduction Effective Security

named packets( )

Names not addresses Name-based routing/ forwarding In-network storage Pull-based transport

ICN properties

LURCH

Running large scale experimentation on Content-Centric Networking via the Grid’5000 platform

Lurch

A newly designed protocol need to be tested Event driven simulation:

limited in the number of events (hence topology size) computation is hard to parallelize

Large scale experiments: Complex to manage

We needed a test orchestrator

From protocol design to large scale experimentation

Lurch

Lurch is a test orchestrator for CCNx1 Simplify and automate ICN’s protocol testing over a

list of interconnected servers (i.e. G5K). Lurch run on a separate machine and control the

test

Controller

Lurch

Application

Control Plane

Virtualized Data Plane

Managem

ent

CCNx

TCP/UDP

Virtualized IP

IP layer

PHY layer

Data PlaneProtocol stack

Architecture Lurch controller:

Virtualized Data plane Control Plane Application layer

Lurch

Lurch

Create virtual interfaces between nodes (i.e. G5K) Bash configuration file computed remotely by the orchestrator and transfered

to experiment nodes Network iptunnels to build virtualized interfaces One physical interface (eth0), multiple virtual interfaces (tap0,..,)

Topology management

#!/bin/bash sysctl -w net.ipv4.ip_forward=1modprobe ipip

iptunnel add tap0 mode ipip local 172.16.49.50 remote 172.16.49.5ifconfig tap0 10.0.0.2 netmask 255.255.255.255 uproute add 10.0.0.1 tap0

iptunnel add tap1 mode ipip local 172.16.49.50 remote 172.16.49.51ifconfig tap1 10.0.0.3 netmask 255.255.255.255 uproute add 10.0.0.4 tap1

1.2.3.4.5.

6.7.

8.

9.10.

tap0

tap1

eth0 eth0 eth0

172.16.9.50 172.16.49.5172.16.49.51

10.0.0.2

10.0.0.3

tap010.0.0.1

tap010.0.0.4

Controller

Virt

ual

Phy

sica

l

Lurch

Remotely assign network resources to nodes preserving physical bandwidth constraints

Bash configuration file computed remotely by the orchestrator and transferred to experiment nodes

Traffic Control Linux tool to limit bandwidth, add delay, packet loss, etc..

Resource management

#!/bin/bash tc qdisc del dev eth0 | cut -d " " -f 1) roottc qdisc add dev eth0 | cut -d " " -f 1) root handle 1: htb default 1

tc class add dev eth0 | cut -d " " -f 1) parent 1: classid 1:1 htb rate 100.0mbit ceil 10.0mbittc filter add dev eth0 | cut -d " " -f 1) parent 1: prio 1 protocol ip u32 match ip dst 172.16.49.5 flowid 1:1

tc class add dev eth0 | cut -d " " -f 1) parent 1: classid 1:2 htb rate 100.0mbit ceil 50.0mbit

tc filter add dev eth0 | cut -d " " -f 1) parent 1: prio 1 protocol ip u32 match ip dst 172.16.49.51 flowid 1:2

1.2.3.

4.5.

6.

7.

8.9.

10Mbps

Controller

Virt

ual

Phy

sica

l

50Mbps

1Gbps

Lurch

Remotely control name-based forwarding tables Bash configuration file computed remotely by the orchestrator and transferred

to experiment nodes CCNx’s FIB control command ccndc

Name-based control plane

#!/bin/bash

ccndc add ccnx:/music UDP 10.0.0.1

ccndc add ccnx:/video UDP 10.0.0.4

1.2.3.4.5.

Name prefix face

ccnx:/music 0

ccnx:/video 1

FIB

ccnx:/music

Controller

Virt

ual

Phy

sica

l

ccnx:/video

Lurch

Remotely control experiment workload File download application started according experiment’s needs

Arrival process: Poisson,CBR File popularity: Zipf, Weibull, et..

Application Workload

Two ways: Centralize workload generation at the

controller Delegated workload generation to clients

for performance improvement

tap0

tap1

eth0 eth0 eth0

172.16.9.50 172.16.49.5172.16.49.51

10.0.0.2

10.0.0.3

tap010.0.0.1

tap010.0.0.4

Controller

Virt

ual

Phy

sica

l

Lurch

Remotely control experiment statistic’s Bash start/stop commands sent remotely

CCNx’s statistics (e.g. caching, forwarding) through logs top / vmstat monitoring active processes CPU usage (e.g. ccnd) Ifstat monitoring link rate

Measurements

At the end of the experiment statistics are collected and transferred to the user

tap0

tap1

eth0 eth0 eth0

172.16.9.50 172.16.49.5172.16.49.51

10.0.0.2

10.0.0.3

tap010.0.0.1

tap010.0.0.4

Virt

ual

Phy

sica

l

Controller

EXPERIMENTS

Running large scale experimentation on Content-Centric Networking via the Grid’5000 platform

Experiments

20 different, simultaneous content requests (flows)

1 name prefix in all the FIBs

4

615M

5M

5M

10M

10M

20M

Link i,j Measured/optimal Rate [Mbps]

0 -> 4 4.7 / 5

1 -> 4 9.2 / 10

2 -> 5 2.4 / 2.5

3 -> 5 2.4 / 2.5

4 -> 6 13.9 / 15

5 -> 6 4.8 / 5

5

ccnx:/

ccnx:/

ccnx:/

ccnx:/

ccnx:/

ccnx:/

Experiments

Large topologies Up to 100 physical nodes More than 200 links

Realistic scenarios Mobile Backhaul

CONCLUSIONS AND FUTURE WORKS

Running large scale experimentation on Content-Centric Networking via the Grid’5000 platform

Conclusions and future works

With Lurch, we tested multiple ICN’s mechanisms in a real big test-bed: Forwarding, caching strategies, Congestion control

Ongoing: Project started in the Orange – Bell Labs collaboration and is now under the

SystemX Architecture de Resaux Future open source release

Future works: Extend site based experiments to grid experiments Exploit the power of the servers offered by grid using two or more virtual machines

per server Adapt the tool to run different ICN prototypes (e.g. NDNx)