Lessons Learned From Using the RIPE Atlas Platform for Measurement Research Lessons Learned From Using the RIPE Atlas Platform for Measurement Research RIPE 68, Warsaw Vaibhav Bajpai, Steffie Jacob Eravuchira, Jürgen Schönwälder (v.bajpai|s.eravuchira|j.schoenwaelder)@jacobs-university.de Computer Networks and Distributed Systems Group, Jacobs University Bremen, Bremen, Germany May 2014 Leone Project: leone-project.eu 1 / 15
15
Embed
Lessons Learned From Using the RIPE Atlas Platform for ... · [1, 1521) probev1 4570 Lantronix XPort Proprobev1.ripe.net (2000, 5000) probev2 4570 Lantronix XPort Proprobev2.ripe.net
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lessons Learned From Using the RIPE AtlasPlatform for Measurement Research
RIPE 68 Warsaw
Vaibhav Bajpai Steffie Jacob Eravuchira Juumlrgen Schoumlnwaumllder(vbajpai|seravuchira|jschoenwaelder)jacobs-universityde
Computer Networks and Distributed Systems GroupJacobs University Bremen
Bremen Germany
May 2014
Leone Project leone-projecteu
1 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Background
Background
Partners within the Leone Project The goal is to assess theQoE of end-users through active measurements1 Primarilydeveloping new metrics for the SamKnows2 platform
Started actively using the RIPE Atlas platform since the RIPEAtlas API was publicly released in 2013 Close collaborationwith Daniel Karrenberg Philip Homburg (RIPE NCC)
This talk is a subset of a larger measurement study using theRIPE Atlas platform (recently submitted to IMC 2014) Weshare our experience in using the platform during this journey
1ripe66ripenetarchivesvideo1259
2samknowscom
2 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 1
Rate Limits
We have ample credits Why canrsquot we provision measurements
Rate limits are setup on each account by default1 No more than 100 simultaneous measurements2 No more than 500 probes per measurement3 No more than 270K credits may be used each day
Although documented3 may not be well-known
Limits can be lifted off by proposing the research on atlas mailing list4
3httpsatlasripenetdocsudm
4Thanks to Vesna Manojlovic (RIPE NCC) for lifting off rate limits on our user accounts3 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationFirmware Variants
Lack of calibration can lead to uncertainity of results - [1]
Frequency of firmware releaseshas increased since 2013
Firmware release cycle since 2011 (as of May 2014)atlasripenetresultsgraphs
Each User-DefinedMeasurement (UDM) tags thefirmware version of the probewhen reporting results
prb_id 10678type traceroutefw 4560
A snippet of a traceroute measurement result froma probe (as of November 2013) 4 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationHardware Variants
Three hardware revisions ofdeployed probes v1 v2 v3v3 probes are more hardwarecapable5 than v1v2 probes
8162 probes are deployed (as of May 2014)atlasripenetresultsmapsnetwork-coverage
Anchors are dedicated serversAnchors are sourcessinks ofregional measurement traffic
56 anchors are deployed (as of May 2014)atlasripenetanchorsmap
5In terms of specifications5 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Background
Background
Partners within the Leone Project The goal is to assess theQoE of end-users through active measurements1 Primarilydeveloping new metrics for the SamKnows2 platform
Started actively using the RIPE Atlas platform since the RIPEAtlas API was publicly released in 2013 Close collaborationwith Daniel Karrenberg Philip Homburg (RIPE NCC)
This talk is a subset of a larger measurement study using theRIPE Atlas platform (recently submitted to IMC 2014) Weshare our experience in using the platform during this journey
1ripe66ripenetarchivesvideo1259
2samknowscom
2 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 1
Rate Limits
We have ample credits Why canrsquot we provision measurements
Rate limits are setup on each account by default1 No more than 100 simultaneous measurements2 No more than 500 probes per measurement3 No more than 270K credits may be used each day
Although documented3 may not be well-known
Limits can be lifted off by proposing the research on atlas mailing list4
3httpsatlasripenetdocsudm
4Thanks to Vesna Manojlovic (RIPE NCC) for lifting off rate limits on our user accounts3 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationFirmware Variants
Lack of calibration can lead to uncertainity of results - [1]
Frequency of firmware releaseshas increased since 2013
Firmware release cycle since 2011 (as of May 2014)atlasripenetresultsgraphs
Each User-DefinedMeasurement (UDM) tags thefirmware version of the probewhen reporting results
prb_id 10678type traceroutefw 4560
A snippet of a traceroute measurement result froma probe (as of November 2013) 4 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationHardware Variants
Three hardware revisions ofdeployed probes v1 v2 v3v3 probes are more hardwarecapable5 than v1v2 probes
8162 probes are deployed (as of May 2014)atlasripenetresultsmapsnetwork-coverage
Anchors are dedicated serversAnchors are sourcessinks ofregional measurement traffic
56 anchors are deployed (as of May 2014)atlasripenetanchorsmap
5In terms of specifications5 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 1
Rate Limits
We have ample credits Why canrsquot we provision measurements
Rate limits are setup on each account by default1 No more than 100 simultaneous measurements2 No more than 500 probes per measurement3 No more than 270K credits may be used each day
Although documented3 may not be well-known
Limits can be lifted off by proposing the research on atlas mailing list4
3httpsatlasripenetdocsudm
4Thanks to Vesna Manojlovic (RIPE NCC) for lifting off rate limits on our user accounts3 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationFirmware Variants
Lack of calibration can lead to uncertainity of results - [1]
Frequency of firmware releaseshas increased since 2013
Firmware release cycle since 2011 (as of May 2014)atlasripenetresultsgraphs
Each User-DefinedMeasurement (UDM) tags thefirmware version of the probewhen reporting results
prb_id 10678type traceroutefw 4560
A snippet of a traceroute measurement result froma probe (as of November 2013) 4 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationHardware Variants
Three hardware revisions ofdeployed probes v1 v2 v3v3 probes are more hardwarecapable5 than v1v2 probes
8162 probes are deployed (as of May 2014)atlasripenetresultsmapsnetwork-coverage
Anchors are dedicated serversAnchors are sourcessinks ofregional measurement traffic
56 anchors are deployed (as of May 2014)atlasripenetanchorsmap
5In terms of specifications5 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationFirmware Variants
Lack of calibration can lead to uncertainity of results - [1]
Frequency of firmware releaseshas increased since 2013
Firmware release cycle since 2011 (as of May 2014)atlasripenetresultsgraphs
Each User-DefinedMeasurement (UDM) tags thefirmware version of the probewhen reporting results
prb_id 10678type traceroutefw 4560
A snippet of a traceroute measurement result froma probe (as of November 2013) 4 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationHardware Variants
Three hardware revisions ofdeployed probes v1 v2 v3v3 probes are more hardwarecapable5 than v1v2 probes
8162 probes are deployed (as of May 2014)atlasripenetresultsmapsnetwork-coverage
Anchors are dedicated serversAnchors are sourcessinks ofregional measurement traffic
56 anchors are deployed (as of May 2014)atlasripenetanchorsmap
5In terms of specifications5 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe CalibrationHardware Variants
Three hardware revisions ofdeployed probes v1 v2 v3v3 probes are more hardwarecapable5 than v1v2 probes
8162 probes are deployed (as of May 2014)atlasripenetresultsmapsnetwork-coverage
Anchors are dedicated serversAnchors are sourcessinks ofregional measurement traffic
56 anchors are deployed (as of May 2014)atlasripenetanchorsmap
5In terms of specifications5 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 2
Probe Calibration
probev1 probev2 probev3 anchorv2
PROBE ID HARDWARE REVISION FIRMWARE VERSION HARDWARE WEBPAGE
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
Does probe hardware revision have effect on measurement results
Find probes whose
1 Hop 1 is private [2] and2 Hop 2 is public
This ensures
1 Probe does not cross a wireless link72 Probe wired directly behind a residential gateway
Provisioned traceroute measurements that lasted for a dayInvestigated latencies to hop 1 from v2v3 probes
7probes themselves cannot associate to a wireless access point7 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
VERIZON (US)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
VODAFONE (DE)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
BELGACOM (BE)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
XS4ALL (NL)
probev2probev3
Generated as of November 2013 (RIPE Atlas)
v3 probes show expected lt 1ms latencies to hop 1
v2 probes show surprisingly high latencies around 6ms to hop 1
8 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
We next investigated RIPE Atlas source code8
Measurement tools are adaptation of busybox utilities
Measurements modified to run in an evented manner using libevent
Measurements do not spawn new processes (instead invoked as function calls)
Circumvents absence of MMU in v1v2 probes
A single evented loop handles all measurement requests
8httpsatlasripenetget-involvedsource-code
9 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
VERIZON (US)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
VODAFONE (DE)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
BELGACOM (BE)
probev2probev3
0
01
02
03
04
05
06
07
08
09
1
0 1 2 3 4 5 6 7 8 9 10Percentile of number of probes
Latency to rst hop (ms)
XS4ALL (NL)
probev2probev3
Generated as of November 2013 (RIPE Atlas)
v3 probes show expected lt 1ms latencies to hop 1
v2 probes show surprisingly high latencies around 6ms to hop 1
8 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
We next investigated RIPE Atlas source code8
Measurement tools are adaptation of busybox utilities
Measurements modified to run in an evented manner using libevent
Measurements do not spawn new processes (instead invoked as function calls)
Circumvents absence of MMU in v1v2 probes
A single evented loop handles all measurement requests
8httpsatlasripenetget-involvedsource-code
9 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 3
Hardware Matters
We next investigated RIPE Atlas source code8
Measurement tools are adaptation of busybox utilities
Measurements modified to run in an evented manner using libevent
Measurements do not spawn new processes (instead invoked as function calls)
Circumvents absence of MMU in v1v2 probes
A single evented loop handles all measurement requests
8httpsatlasripenetget-involvedsource-code
9 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
) event_assign (ampbase -gtevent4 base -gtevent_base
base -gtv4icmp_rcv EV_READ | EV_PERSIST ready_callback4 base)
static void ready_callback4 (int __attribute (( unused )) unused const short __attribute (( unused )) event void s
) ms=(nowtv_sec -state -gtxmit_timetv_sec )1000 +
(nowtv_usec -state -gtxmit_timetv_usec )1e3
traceroute code snippet from v4570 running on v2 probes as of November 2013
RTT time-stamping is performed in user-space in the evented callbackA probe loaded with multiple measurements will witness time-stamping delaysDelays more pronounced on probes with hardware constraints v1v2
10 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 4
Proper Statistics Matter
Can per-hop averaging of RTT significantly vary the results
RIPE Atlas
1 evtraceroute sends 3 ICMP queries(default) to each hop
2 RTT from each ICMP response isseparately made available
SamKnows
1 mtr sends 3 ICMP queries (default)to each hop
2 RTT are averaged over each hop9
1 2 3 4 5 6 7 8 9 10 11 12 13
1700
1800
1900
2000
2100
2200
2300
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
Latency to
rst hop (ms)
Time (UTC)
BT (UK) ProbeID 2107 Probe Firmware 4570 Probev2
MedianAverage
Generated as of November 2013 (RIPE Atlas)
13131313
131313
$amp13($amp13)+13($amp13
)-amp13amp0-amp
Generated as of November 2013 (RIPE Atlas)9We replaced the traceroute test within SamKnows to expose results without aggregation
11 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 5
Heavy-Tailed AS-based Probe Distribution
56 (2307) of probes fallwithin AS ranks lt= 200ASes with Rank gt 200 haveless than 10 probes
13
13
Distribution of public connected and non-anchoredprobes (4133) as of October 2013
AS with Rank 1contributes 4 of all probesASes with Rank gt= 10contribute 19 of all probes
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Lesson Learned 6
Metadata is (Changing) Data
How is the probe connected to the Internet
The connection speedNetwork Type (CoreResearch IXP AccessHome) NetworkWAN Type DSL CableFibre to the X
Ability to track changes and APIaccess to metadata history
Probe registration page atlasripenetregister
13 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
Conclusion
1 Rate Limits2 Probe Calibration3 Hardware Matters4 Proper Statistics Matters5 Heavy-tailed AS-based Probe Distribution6 Metadata is Data
14 15
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt
15 15
Background
Lesson Learned 1
Lesson Learned 2
Lesson Learned 2
Lesson Learned 2
Lesson Learned 3
Lesson Learned 3
Lesson Learned 4
Lesson Learned 5
Lesson Learned 6
Conclusion
Lessons Learned From Using the RIPE Atlas Platform for Measurement Research
Conclusion
References
[1] P Eardley M Mellia J Ott J Schoumlnwaumllder and H Schulzrinne ldquoGlobalMeasurement Framework (Dagstuhl Seminar 13472)rdquo Dagstuhl Reports vol 3no 11 pp 144ndash153 2014 [Online] Availablehttpdropsdagstuhldeopusvolltexte20144440
[2] Y Rekhter B Moskowitz D Karrenberg G J de Groot and E Lear ldquoAddressAllocation for Private Internetsrdquo RFC 1918 (Best Current Practice) InternetEngineering Task Force Feb 1996 updated by RFC 6761 [Online] Availablehttpwwwietforgrfcrfc1918txt