pinger: actively measuring the worldwide internet’s end-to-end performance

PingER: Actively measuring the worldwide Internet’s end-to-end performance

Les CottrellSLAC

Workshop at the University of Malaya in Kuala Lumpur , June 24-25, 2013

AgendaUsing PingER measurements going back to 1998 and covering 168 countries, this talk will illustrate, Internet performance worldwide.

•Brief history

•How can the Internet help development?

•How does PingER measure Internet performance?

•What do we measure, what does it tell us?

•What do we find?

•Case studies illustrating PingER

•Managing PingER

History• Story of Ping

• Early PingER

• Extension to Developing Regions

• Extension to Pakistan

• Extension to Malaysia

The start• Ping tool invented by Mike Muuss

– “a little thousand-line hack” during a single evening to troubleshoot “odd behavior” on the computer network at the U.S. Army’s Ballistic Research Laboratory in Maryland.

– sent a small data packet known as an echo request to an IP address, typically a remote server or network node. If the target address was reachable, it echoed back the same data, and the program recorded the time it took for the round-trip journey.

– Reminded Muuss of the percussive sound pulse sonar systems use to detect objects underwater, he named it after that sound—ping

• Now defined by RFC 792

5 Joint Techs: I2 & ESnet,Stanford

Measurement Mechanism: PingER

Internet

10 ping request packets each 30 mins

RemoteHost(typicallyweb server)

>ping remhost

Ping response packets

Measure Round Trip Time & Loss

Monitor Host

6

Early PingER• As the head of networking at SLAC, I set up the

system using ping simply to test connections between the laboratory and several dozen research institutions in about a dozen countries that were collaborating on a physics experiment known as BaBar to study properties of subatomic particles.

• Over the next half-decade, as word of PingER’s value spread, I extended monitoring to hundreds more physics laboratories and science centers across the globe. But the project didn’t take a humanitarian turn until 2001.

•UNIMAS

Workshop, Sarawak, Dec 2012

7

Extension to the Developing regions• In 2001 I visited ICTP in Italy.• Driven by ICTP’s goals of bringing first-class science

and technology to developing countries they wanted to know how well the networks were working.

• The simple PingER project was the perfect tool for the job. Ubiquitous ping so nothing to install at remote targets.

• They offered to help expand the project to those parts of the world that needed it most.

• Within the next year, we began establishing monitoring and target hosts in countries as diverse as Ecuador, Rwanda, Jordan, and Bhutan.

UNIMASWorkshop, Sarawak, Dec 2012

8

Extension to Pakistan• In 2004 set up joint agreement with NUST in Pakistan• Soon got my first real glimpse of just how much of a

difference PingER can make. – Set up a PingER monitoring site in the country to

assess performance on the then year-old Pakistan Educational Research Network (PERN).

– The network’s providers touted its bandwidth of 155 Mbps, impressive at the time. But PingER revealed that the “last mile” links to universities were dreadful. These bottleneck connections funneled data at no more than 1 Mbps, causing long delays and high packet loss.


9

Extension to Pakistan

• He clearly took PingER’s lessons to heart. When construction of PERN2 began in 2009, its plans included extending high-speed, 1-Gbps data links all the way to university data centers


• During a visit to the university, I presented our findings to the chairman of Pakistan’s higher education commission, Atta-ur-Rehman, who was preparing to fund the next major upgrade to PERN.

10

Extension to Malaysia

• Have set up an official signed MoU between SLAC & U of Malaysia in Sarawak (UNIMAS)

• Idea was to replicate the NUST project

• Fortnightly meetings by Skype

• Just getting started, no students yet

• Met with Vice Chancellor (VC) at one meeting

• This workshop is a follow up.


11

Why do measurements of the Internet matter


Why does it matter

• African scientists isolated

• Lack critical mass

• Need network to collaborate but it is terrible

• Brain drain

• Brain gain, tap diaspora

• Blend in distance learning

• Provide leadership, train trainers

12

Internet Users 2002

Cartograms from:www.geog.qmw.ac.uk/gbhgis/conference/cartogram.html

Tertiary Education fromhttp://www.worldmapper.org/

http://www.geog.qmw.ac.uk/gbhgis/conference/cartogram.html

http://www.geog.qmw.ac.uk/gbhgis/conference/cartogram.html

13 eGY Africa 2012Workshop, Nairobi Oct 2012

How does the Internet help• Investment in information technology plays the role of

a "facilitator" that allows other innovations to take place. http://findarticles.com/p/articles/mi_m1093/is_3_45/ai_86517828/

• World Bank / IFC report: for every 10% increase in high-speed Internet connections there is an increase in economic growth of 1.3 percentage points. April 2010. http://www.infodev.org/en/Article.522.html

• Example: Uganda 15% increase in price of maize based on improved farmer bargaining power. www.itu.int/ITU-D/.../S1-01-NG-ICT_Indicators-Tim_Kelly.pptx

14

How does PingER work

• Mechanism

• Coverage


15 Joint Techs: I2 & ESnet,Stanford

Measurement Mechanism: PingER

Internet

10 ping request packets each 30 mins

RemoteHost(typicallyweb server)

>ping remhost

Ping response packets

Measure Round Trip Time & Loss

On

ce a Day

Uses ubiquitous ping

Monitor Host

Repositories

NUST

16

Deployment• Monitors > 90 in 23 countries, 4 in Africa

Summer Joint TechsStanford, July 2012

• Beacons monitored by most monitors (~100)• Remote sites monitored by some monitors (~750)

17

Metrics Available from PingER• UnReachability

• Minimum RTT

• Average RTT

• Jitter

• Loss

• Derived throughput

• MOS

• Directness of Connection

• OthersUNIMAS



Unreachability


Unreachability: e.g. N. African uprisings Jan ‘11

NARSS (Cairo)

Helwan (Cairo)

EUN (Cairo)

23:59 Jan 28

23:59 Jan 27

12:00 Jan 27

• Impact varied: start time, recovery time, after effects• Egypt University Network (EUN) down least time

– NARSS via AlterNet->Italy->Egypt, Helwan &EUN via PCCW Global

• Libya first went dark 06:00 Feb 19 for 3 days, then again on Mar 4th more permanently

• Algeria, Morocco, Tripoli not noticeable

=No pings respond

19

Average Round Trip Time (RTT)• Mainly a distance related, but also congestion (i.e. at

the edge)

• For real-time multimedia (H.323) traffic RTT: 0-300ms =Good, 300-600ms=Acceptable, and > 600ms= poor.


20

Minimum RTT history by region• Minimize effects of congestion and queuing


21 UNIMASWorkshop, Sarawak, Dec 2012

•GEOS (Geostationary Earth Orbit Satellite)–Good coverage, but expensive in $/Mbps–& long delays min RTT >450ms easy to spot

N.b. RTTs > 250 ms bad for VoIP

Impact of GEOS vs. TerrestrialGEOSGEOS

Demo


Conversion history by country seen by min-RTT

Demo

23

Jitter• Mainly at edges, critical for real time: VoIP, gaming• Exponential improvement (factor 10 in 6 yrs.)• The optimum amount of one way latency is 11 ms for keeping time in

music. – Above that delay and they tend to slow down. – >50-70 ms performances tended to completely fall apart.


• For real time haptic control and feedback for medical operations <=80ms is needed.

N. America, Europe, E Asia & Oceania < 1msAfrica, S. Asia & S.E. Asia worst off

24 Summer Joint TechsStanford, July 2012

• Low (<1%) losses are good.• Huge impact on data transfer times (due to timeouts)• Real time impact due to recovery timeouts, e.g. echoing typing• Losses are mainly at the edge, so often distance independent• Losses improving roughly exponentially, ~factor 100 in 12 years

24

Loss has Similar

behavior to thruput

• Best <0.1%: N. America, E. Asia, Europe, Australasia

• Worst> 1%:• Africa & C. Asia

Loss



Derived ThroughputDerived throughput ~ 8*1460/(RTT*sqrt(loss))

Mathis et. al

Europe, E. Asia & Australasia merging

Behind Europe:

4 yrs: Russia, 7 yrs:L America,

M East, SE Asia

11 yrs.: India, C. Asia

13 yrs.: Africa

26

• ITU metric, based on quality of a conversation– Originally people listen and give quality 1-5– Can derive from RTT, jitter and loss


MOSMean Opinion Score MOS)

• >=4 is good,

• 3-4 is fair,

• 2-3 is poor.

Important for VoIP

Usa

ble

From the PingER project http://www-iepm.slac.stanford.edu/pinger


Directness of connection (Alpha)• Alpha to allow for delays in network equipment &

indirectness of actual route. D = 1 way distance• Distance=D(km)=Alpha*[RTT(msec)*100(km/msec)]– Alpha = D(km) / (min_RTT[msec] * 100 [km/msec])

• If know lat/long of monitor and remote host then know D, so with min-RTT can estimate Alpha– Max(Alpha) =1 = direct (great circle) route and no

network delays– Alpha > 1 probably identifies bad lat/long

coordinates for hosts.– Low values typically mean very indirect route, or

satellite or slow connection (e.g. wireless)– Alpha typically ~ 0.45


S.E. Asia Directivity


UNIMAS to …W MY A~0.45

•->PH via JP A~0.22

•->ID via HK A~0.17

•->TH via SG A~0.3

•->BR via ? A~0.07

•->KH via HK-VN A~0.16

•->SG A~0.23

Need measurements from W MY + routing info

TEIN3Trans Eurasia Info Net


Alpha worldwide

• Interest in Polar route with Global warming


Alpha=0.71A

lpha

=0.7

3

Alpha

=0.7

3SLAC

JP

AU

NZ

NSK.RU

TW0.180.16IN

DE

EG0.34

JP0.32

o.41

AU0.53

30

• Big improvements for C Asia, S Asia & Australasia


Directivity (Alpha) from SLAC to world


More stable year to year as add more hosts

31

Other metrics• Duplicate packets (try ping www.cern.ch, load

balancing?)

• Out of order packets (parallel paths)

• Conditional Loss Probability (non-random loss)

– one packet is lost the following packet is also lost– route change, loss of sync, spanning tree

reconfiguration

• Maximum packet loss (useful for buffer bloat?)

32

S.E. Asia

• Just started mining early data

• Where does Malaysia sit

• How much variation in SE Asia

• Variation in Malaysia

• Troubles at UNIMAS

• Top and bottom 3 sites monitored in Malaysia


33

Malaysia vs. Other Regions


34

Variation between SE Asian countries

• Factor of 10 between Singapore and Laos

• Singapore 4x better than next countries

• Exponentially improving with time

• On its ownSingaporeapproachesE Asia.


35

Avg-RTT, jitter & loss by Malay State• Need low values of all 3 metrics

• RTTs similar, big diffs in jitter & loss

• Allianze UniColl looks bad

• UTP next worst loss


Seen from SLAC, Nov 2012

36

Non lossy Malaysian hosts seen from SLAC Nov 2012


UNIMAS

MIMOS

MIMOS

UNISZA

Note monitoring host (SLAC) down

37

OCESB

Lo

ssy

ho

st s

een

fro

m S

LA

C N

ov

2012

UPSI

UTEM

MIU

AIU

Sabah

38

Diving deeper: packet loss Nov 27-28• Allianze University College unreachable• UTEM, MIU, UPSI, OCESB, AIU, SABAH experienced loss Nov

26• No PingER loss from the rest on Nov 26: Johor, Kelantan, KL,

Nigeri Sembilan, Sarawak, Terengganu


1930-midnight MST, backup?

UTEMLosses isolated, not correlated with large RTT

39

Another host, large RTTs correlate with time of day


40

UNIMAS Jitter


UNIMAS to

MalaysiaJitter

AllianzeUniversity

College

Universiti TechnologiPetronas

41

Why not show UNIMAS to Malaysia more

• Big changes in RTT affect throughput especially for Kuching

• UNIMAS was seeing congestion – This would be seen everywhere– Turn on shaping– Removes loss &

day-night variations


2

1RT

T m

s

Background loss colors

SLAC to UNIMAS Oct-Nov 2012


Increase of capacity to UNIMAS from 200Mbps to 500Mps


Huge spikes removed



Improvement

44

Unreachable Malaysian hosts

• Unreachable from SLAC Nov 1-26, 2012:1. 92% Allianze University College

2. 20% www.ocesb.com.my (Speedtest)

3. 17% Universiti Teknologi Petronas, Bandar Seri Iskandar

4. 4% University Malaysia Kelantan

• 100% Reachable– MIMOS, UNIMAS, Sultan Zainal Abidin University,

USIM, University Teknologi Malaysia, Sultan Idris University of Education


45

Case Study Africa


46

Submarine Cables

2012

1 cable, W Coast only, No competition (340Gb/s)

World Cup S Africa 2008 led to many submarine cables connecting Africa to rest of the worldCapacity, shorter RTT, competitionMost cables are now activeFuture Cables promise connectivity:

ACE Q3 2012: France-Spain-Morocco & many African countries

SAex 2014: Brazil-Angola-SA WASACE 2014: France-US-Brazil-Nigeria-

Angola-SABRICS 2014: Brazil-SA-Mauritius-India-

China-RussiaCable capacity increase from 0.34Tb/s in

2008 to 87.5 Tb/s by 2014 (factor ~3000)http://manypossibilities.net/african-undersea-cables

/

2008

47 Summer Joint TechsStanford, July 2012

Min-RTT satellite vs. fibre

2010

2009

2008

OK

to

US

48

Comparison in minimum from SLAC to African Countries in 2008 and 2012.


2008 2012800

600

400

200

0

800

600

400

200

0TerrestrialTerrestrial

GEOS GEOS

Minimum RTT from SLAC to African Countries 2008 vs.. 2012

Min

RT

T (

ms)

Note the countries that have gone from GEOS to terrestrial

49

RTT


• Angola step mid-May, more stable

• Zambia one direction reduce 720>550ms– Unstable, still

trying?• Tanzania, also

dramatic reduction in losses

• Uganda inland via Kenya, 2 step process

• Rwanda Sep 25• Many sites still

to connect

750ms750ms 450ms450ms

Aug 20Aug 20

SLAC to AngolaSLAC to Angola

SLAC to ZambiaSLAC to Zambia

SLAC to TanzaniaSLAC to Tanzania

SLAC to UgandaSLAC to Uganda

1 direction1 direction

Both directionsBoth directions

Sep 27Sep 27

1 direction1 direction Both directions?Both directions?

RTT, e.g. changes in 2009

50

Intra Africa Optical Fibre Networkhttp://www.ubuntunet.net/fibre-map


Just as important as the submarine cables serving the coasts, are the tens if not hundreds of millions of dollars being invested in new terrestrial fiber to move this capacity inland.

51

Other 3 billion (O3b)• Refers to population of world without

broadband

• Constellation of 8 Medium Earth Orbit satellites at altitude 8000km

• Min RTTs factor of 4 less than GEOS – ~125ms, similar to inter-continent land lines

• Backed by SES World Skies, HSBC, Google…

• Launch 2013

eGY Africa WorkshopNairobi Oct 2012

52

Differences within Africa• Southern Africa now leads, caught North • Central is worst off• Notice improvement since 2008

– satellite=> terrestrial, competition

eGY Africa 2012Workshop, Nairobi Oct 2012

Derived throughput from SLAC to African regions

• Africa improving, catching up?

53

Africa was 16 yrs. behind Europe in 2009


54

Africa is now 14 yrs. behind Europe in 2012

UNIMASWorkshop, Sarawak, Dec 201214 years

55

Africa might catch up with Europe in 20 years at current rate of improvement


Was falling further behind

70 times w

orse

26 years beh

ind

Eu

rop

e

56

• By 2011 prices had dropped only factor of 2– Alternative fibre often owned by electricity

companies, pipelines and not allowed to sell, lease or operate – needs deregulation and is happening

– Business model: ISPs sell to large corporations, gov, edu, NGOs=>small customer base to recover costs from =>high prices

• WiFI & Mobile to the rescue, overlay 3G with fibre net

• Alex Twinomugisha originally published |Why Are African Internet Access Prices Still High? on Africa Business Source

Prices


Then there is the cost

58

Africa Broadband costs vs. rest of the world


59

NRENs in Africa

eGY Africa WorkshopNairobi Oct. 2012

With connection to GÉANT going live end 2012, UbuntuNet will provide sub-Saharan Africa with infrastructure for global, and regional research collaboration and e-learning

N. Africa connected via EUMED to Europe. Also Arab States REN formed 1 year ago

From PingER Project

60

Traceroutes from S Africa & Burkina Faso in 2009


Despite having NRENs & terrestrial fibres along both East & West coasts of Africa connecting to most maritime countries, still most inter-African routes went via Europe and N America

Not only did this add large delays, but also resulted in costly inter-continental rates

61

Setting up IXPs for better connectivity


62

Opportunities: Square Kilometre Array (SKA)


• Build in in Sub-Saharan states with cores in South Africa and Australia,

• €1.5 billion, construction start 2016, initial observations 2019

“…equivalent to ten times the global internet traffic today”

“…equivalent to ten times the global internet traffic today”

63

Opportunities: Continued• Aug 30, 2012: 220 donated computer servers

from CERN, Switzerland, will start a journey to be delivered to the Kwame Nkrumah University of Science and Technology in Ghana.– This will provide a new computing center for

KNUST and boost African physics onto the international stage, helping African students, e.g. enable participation in simulations of LHC data.

• Strategic plan for a synchrotron light source in southern Africa – 2 day workshop Dec 2011, Pretoria– http://indico.saip.org.za/conferenceDisplay.py?ovw=True&confId=12

64

More Information

• Case Study– https://confluence.slac.stanford.edu/display/IEPM/N

ew+E.+Coast+of+Africa+Fibre

• Telegeography submarine cable map– http://www.submarinecablemap.com/

• Africa Undersea Cables– http://manypossibilities.net/african-undersea-cables/

• Ubuntunet– http://www.ubuntunet.net/fibre-map

eGY Africa WorkshopNairobi, Oct. 2012

65

Mediterranean Cable Cut

• See https://confluence.slac.stanford.edu/display/IEPM/Effects+of+Fibre+Outage+through+Mediterranean


66

Mediterranean Fibre Cuts Jan 31, 2008• 2 major cables: SEAMEW4, FLAG cut off

Marseilles and Alexandria

• Traffic falls over to SEAMEW3

• SEAMEW3 congested, resulted in doubling RTTs, increased jitter & Loss

• Start between 6:47am and 7:16am GMT• Reduced bandwidth by over 50% to over 20 countries


200=>400msmsLost connectionSLAC to tanta.edu.eg800

6004002000

RT

T m

s

67

Affected many countries


68

Impact on loss lasted several days


Afr

ica

M E

ast

S A

sia

SE

A

sia

69

World events Case studies

• Japanese Earthquake & Tsunami, 2011– https://confluence.slac.stanford.edu/display/IEPM/J

apanese+Earthquake+March+11th%2C+2011

• Chilean Earthquake 2010– https://confluence.slac.stanford.edu/display/IEPM/C

hilean+Earthquake+Feb+27th%2C+2010

• Syria Nov 29, 2012– https://confluence.slac.stanford.edu/display/IEPM/S

yria+shuts+down+its+Internet+connection UNIMAS


70

Effect of Japanese Earthquake & Tsunami March 11, 2011


2001000

RT

T (

ms)

SLAC to KEK

Traffic rerouted Eastwards

71

Chilean earthquake, Feb 27, 2010

• Telmex is an Internet service provider in Chile so it should have better connectivity than most organizations. The plot shows lack of connectivity (black), losses (colored background, see the legend below) and median RTT. The loss of connectivity following the earthquake at 6:34am UTC on February 27th is apparent. Connectivity was re-established at about 9:30pm (UTC) that evening followed by considerable instability.


72

Syria shuts down Internet, Nov-29-12• Renesys reported:

– Starting at 10:26 UTC on Thursday, 29 November (12:26pm in Damascus), Syria's international Internet connectivity shut down.

• Prior to this working well

SLAC to Thawra Online

100

200

0

mse

c

12:00:0000:00:0012:00:0000:00:0012:00:00

SLAC to iNET

100

200

0

mse

c

Source: PingER

29 Nov ‘12

Damascus

iNET

73

Other uses


74

Trouble shooting

• Identify when RTT etc. changed (or did not)

• Has congestion increased (jitter, diurnal changes)

• Is it worse than similar paths

• Is it related to a route change?– Also keep daily traceroutes

• Identify that it is no longer a problem

• Identify problem sites to focus on


75

Choosing ISP

• Choosing a service provider given a choice:– For SLAC ESnet vs. Internet2

• Got ISP to change routing to keep our business

– Choice of a DSL provider for residential coverage in Bay Area


76

Comparison to UN ITU ICT Development Index


100

1000

10000

ICT Development Index from the UN International Telecommunications Union

2 4 6 810

Pin

gER

Der

ived

thr

ough

put

(kbi

ts/s

ec)

Bubble size = Population

Demo

77

Malaysian and Brunei sites monitored from UNIMAS

eGY Africa 2012Workshop, Nairobi Oct 2012Demo

78

UNIMAS to SE. Asia


CambodiaThailand

Philippines

79

Viewing data


• Pingtable.pl– Select metric, tick, source, destination etc– Table downloadable for Excel– Details on hosts, and graphs of performance

• Table of country to country connections

• Map– See monitors, beacons, remotes– Select metric, src, dst, draw colored lines

connecting, colors = metric value– Graphs of metric and pinger hosts per country

• Maps of metric performance by Country

• Google Explorer and motion charts

80

Managing• Site map

– Introductions

• Meta database– Oracle database => perl require script and

pinger.xml files

• Data gathering status

• Measurements:– Run from cronjob each 30 mins

• Spotting anomalies, hosts with same IP addr, hosts with missing information (country, lat/lon

• Finding hostseGY Africa 2012

Workshop, Nairobi Oct 2012

81

Problems• Hosts change IP address

– Maybe v frequent, e.g. a cluster or slowly as host with same name is replaced with a new host with new IP address.

• Hosts not where you think they are

• No packet loss how to calculate Throughput ~ 1460(bytes)*8(bits)/(RTT*sqrt(loss))

• Throughput approx. only good for TCP Reno– OS’ such as Windows, Linux now allow other

congestion control algorithms


82

Joining • ~ 85 active monitors worldwide

– https://confluence.slac.stanford.edu/display/IEPM/Installation+Overview gives the state

• Invitation letter for monitoring sites– http://www-iepm.slac.stanford.edu/pinger/letters/invite-monit

or.doc

• Host requirements– http://www-iepm.slac.stanford.edu/pinger/tools/mon-req.html

• Download site for application etc.– Traceroute/ping server– Pinger2.pl measuring engine– https://confluence.slac.stanford.edu/display/IEPM/Installation

+Overview

83

Demo• Interactive demonstrations of the data mining

capabilities of public data sources provided by organizations such as the UN and ITU coupled with monitoring data from PingER

• http://www-iepm.slac.stanford.edu/pinger/explorer.html



More Information• PingER web home page

– http://www-iepm.slac.stanford.edu/pinger/

• Tutorial on network monitoring & PingER– http://www.slac.stanford.edu/comp/net/wan-mon/tutorial.html

• PingER data Explorer– www-iepm.slac.stanford.edu/pinger/explorer.html

• PingER Project site map– http://www-iepm.slac.stanford.edu/pinger/site.html

• Invitation to join– www-iepm.slac.stanford.edu/pinger/letters/invite-

monitor.doc

pinger: actively measuring the worldwide internet’s end-to-end performance

Documents

difference pinger

pinger monitoring site

simple pinger project

dozen countries

computer network

dozen research institutions

network node

internet help development