Skip to content

ErikWittern/web-api-benchmarking-data

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Web API Benchmarking Data

The here presented data (contained in data.zip) reflects measured qualities for 15 Web API endpoints. The collection and analysis of this data has been published in the research paper "Benchmarking Web API Quality", presented at the 2016 International Conference on Web Engineering (ICWE '16). We also published the tools to collect such data here.

Data description

The data has been collected over a period of 3 months, from August 20th, 2015 (16:00h CEST) until November 20th, 2015 (16:00 CEST). The data was collected for the following Web API endpoints:

API name Ping HTTP HTTPS URL
Flickr X X X api.flickr.com/services/feeds/photos_public.gne
OpenWeatherMap X X - api.openweathermap.org/data/2.5/weather?q=New%20York,us
Postcodes.io X X X api.postcodes.io/postcodes/GU50BD
ConsumerFinance X X X data.consumerfinance.gov/api/views/fphp-cr5a/rows.json
Police.uk.co - X X data.police.uk/api/crimes-street-dates
Wikipedia X X X en.wikipedia.org/w/api.php?action=query&titles=Albert_Einstein&prop=images&format=json
Apple iTunes X X X itunes.apple.com/search?term=u2&entity=musicVideo
Google Maps X X X maps.googleapis.com/maps/api/geocode/json?address=New%20York%20City%20US&sensor=false
MusicBrainz X X X musicbrainz.org/ws/2/recording?query=%22we%20will%20rock%20you%22%20AND%20arid:0383dadf-2a4e-4d10-a46a-e9e041da8eb3
Yahoo X X X query.yahooapis.com/v1/public/yql?q=select%20*%20from%20weather.forecast%20where%20woeid%20in%20(select%20woeid%20from%20geo.places(1)%20where%20text%3D%22nome%2C%20ak%22)&format=json&env=store%3A%2F%2Fdatatables.org%2Falltableswithkeys
Amazon S3 - X X s3.amazonaws.com/1000genomes
Twitter X X - urls.api.twitter.com/1/urls/count.json?url=http://www.tu-berlin.de
Spotify X X X ws.spotify.com/search/1/album?q=u2
BBC - X - www.bbc.co.uk/radio1/playlist.json
Google Books X - X www.googleapis.com/books/v1/volumes?q=isbn:0321751043

The data has been collected from virtual machines running in the following geographic regions: EU (Ireland), US West (Oregon), South America (Sao Paulo), Asia (Singapore), Australia (Sydney), Tokyo, US East. For every region, the data consists of 4 files:

  • httpget_output.csv contains data about performing an HTTP GET request to each endpoint every 5 minutes. Per request, the timestamp, the HTTP status code, the latency in milliseconds, and the response size in byte are stated.
  • httpgets_output.csv contains data about performing an HTTPS GET request to each endpoint every 5 minutes. Per request, the timestamp, the HTTP status code, the latency in milliseconds, and the response size in byte are stated.
  • ping_output.csv contains data about performing ping requests (using the ICMP protocol) to each endpoint every 5 minutes. Per request, the timestamp at the beginning and end of the request period are contained. The pingability denotes how many of 5 ping requests, which followed one another immediately, succeeded. Finally, we provide the average latency across the 5 ping requests.
  • errors.log contains logged error messages in case that HTTP or HTTPS GET requests failed. The error log entries can be correlated to the other HTTP and HTTPS request data by means of the invoked endpoint and the timestamp.

Web API anonymization

The measured qualities of Web APIs are subject to various factors, like server outages, network partitionings, or client errors. Not all of these factors lie under the control of the API provider. Thus, in order to not suggest poor performance of specific providers, we chose to anonymize the data. Correspondingly, APIs are randomly assigned with the names api_1 to api_15. Upon reasonable request, we will consider providing a mapping between the random names and the actual API endpoints. Please contact us under witternj@us.ibm.com or david.bermbach@tu-berlin.de.

Usage and license

If you use this data for research purposes with a resulting publication, we would appreciate if you cite our paper:

@inproceedings{bermbach_wittern_web_api_bench,
 author = {David Bermbach and Erik Wittern},
 title = {Benchmarking Web API Quality},
 booktitle = {Proceedings of the 16th International Conference on Web Engineering (ICWE2016)},
 series = {ICWE'16},
 year = {2016},
 publisher = {Springer}
}

About

Scripts to analyze data from benchmarking 15 public web APIs over 3 months in autumn 2015.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors