a practical internet route oracle
TRANSCRIPT
![Page 1: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/1.jpg)
Sibyl A Practical Internet Route Oracle
Ítalo Cunha
P. Marchetta, M. Calder, Y-C. Chiu
B. Schlinker, B. Machado, A. Pescapè
V. Giotsas, H. Madhyastha, E. Katz-Bassett
![Page 2: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/2.jpg)
Lots of use cases
Topology mapping
AS relationship inference
Route performance and inflation
Locating congestion
Identifying outages
Detecting prefix hijacks
2
Traceroute is Widely Used
“The number one go-to tool is traceroute.” NANOG Network operators troubleshooting tutorial, 2009.
![Page 3: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/3.jpg)
Lots of use cases
Topology mapping
AS relationship inference
Route performance and inflation
Locating congestion
Identifying outages
Detecting prefix hijacks
3
Traceroute is Widely Used
“The number one go-to tool is traceroute.” NANOG Network operators troubleshooting tutorial, 2009.
Lots of vantage points
PlanetLab
Ark
RIPE Atlas
Traceroute servers
MobiPerf, Dasu, BISmark
![Page 4: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/4.jpg)
Lots of use cases
Topology mapping
AS relationship inference
Route performance and inflation
Locating congestion
Identifying outages
Detecting prefix hijacks
4
Lots of vantage points
PlanetLab
Ark
RIPE Atlas
Traceroute servers
MobiPerf, Dasu, BISmark
Traceroute is Widely Used
“The number one go-to tool is traceroute.” NANOG Network operators troubleshooting tutorial, 2009.
But traceroute only supports one query:
“What is the path from vantage point s to destination d?”
![Page 5: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/5.jpg)
What we do
17
What we want to do
Next-gen measurements
![Page 6: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/6.jpg)
Provide support for rich queries on Internet paths
18
Goal
![Page 7: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/7.jpg)
Paths that go through Sprint’s Chicago PoP to USC:
^.*[Sprint&Chicago].*[USC]$
From NANOG: “Problem between Level3 in LA and
GTT in Seattle?”
^.*[Level3&LA].*[GTT&Seattle].*$
19
Querying Internet Paths with Regular Expressions
![Page 8: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/8.jpg)
21
Limited VPs Limited Path Coverage
![Page 9: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/9.jpg)
More VPs Richer Path Coverage
![Page 10: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/10.jpg)
23
Combining Platforms Improves Coverage
![Page 11: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/11.jpg)
24
Combining Platforms Improves Coverage
Support for multiple measurement platforms
![Page 12: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/12.jpg)
25
Rate Limits Cannot Issue All Measurements
![Page 13: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/13.jpg)
26
Rate Limits Cannot Issue All Measurements
Need to target probes intelligently
![Page 14: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/14.jpg)
In each round, allocate probing budget to best serve
queries
28
Optimize Use of Probing Budget
Pick traceroutes
Tr that maximize
the number of
answered queries
Subject to the
rate limits of each
platform V
![Page 15: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/15.jpg)
“I suspect problems on peering between GBLX-AT&T on way to
Akamai. Give me a matching path.”
35
Optimize Across Candidates
RIPE1
RIPE2
Unlikely to match
![Page 16: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/16.jpg)
“I suspect problems on peering between GBLX-AT&T on way to
Akamai. Give me a matching path.”
36
Optimize Across Candidates
RIPE1
RIPE2
Likely match
![Page 17: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/17.jpg)
Train system to recognize unlikely predictions
Features include:
Peering relationship at splice point
Path length inflation vs shortest prediction
Evaluation shows system can identify measurements more likely to match queries
37
How Likely Is a Spliced Path Correct?
Predicted Jaccard Index
Real Jaccard
Index
![Page 18: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/18.jpg)
Prediction is effective: Sibyl satisfies 81% as many queries
as an Oracle that knows which candidates match each query
Important to assess likelihood: Sibyl satisfies 264% more
than Randomly selecting among spliced candidates
38
Evaluation: How Effective Is Probing Allocation?
Random Candidates
Sibyl
Oracle
![Page 19: A Practical Internet Route Oracle](https://reader033.vdocuments.net/reader033/viewer/2022041802/62542dc1b2a87166d624c326/html5/thumbnails/19.jpg)
Improve path prediction and ranking
Better formalism, richer training sets
Balance between serving current queries and expected benefit in serving future queries
Fill in gaps in routing knowledge
Refresh stale knowledge
Unify queries over historical and live data
“Give me a path that used to look like X but now looks like Y.”
Queries over path performance
Latency, bandwidth, loss, length
39
Future Work