Characterizing the file hosting ecosystem: A view from the edge

Abstract

We present a comprehensive, longitudinal characterization study of the file hosting ecosystem using HTTP traces collected from a large campus network over a one-year period. We performed detailed multi-level analysis of the usage behaviour, infrastructure properties, content characteristics, and user-perceived performance of the top five services in terms of traffic volume, namely RapidShare, Megaupload, zSHARE, MediaFire, and Hotfile. We carefully devised methods to identify user clickstreams in the HTTP traces, including the identification of free and premium user instances, as well as the identification of content that is split into multiple pieces and downloaded using multiple transactions. Throughout this characterization, we compare and contrast these services with each other as well as with peer-to-peer file sharing and other media sharing services.

Design and Evaluation of HTTP Protocol Parsers for IPFIX Measurement

“…In both cases a ground truth sample is needed, which is a topic addressed by [22]. In [7] and [12] the authors use DPI to obtain information from the HTTP headers. Our approach is orthogonal to these works, since we are interested in extending IP flow records with HTTP data.…”

Holiday Pictures or Blockbuster Movies? Insights into Copyright Infringement in User Uploads to One-Click File Hosters

“…They are typically implemented in a centralised fashion with thousands of servers located in computing centres [2,16,20]. According to previous studies, there are more than 300 OCHs [12].…”

“…A file can be downloaded only when the corresponding download link is known. Therefore, uploaders who wish to disseminate their files post the download links on blogs, social networking sites, discussion boards, or they even submit their links to specialised search engines such as Filestube [2,12,15,16]. Mahanti et al [15] observed OCHs were receiving incoming traffic from up to 8,000 indexing sites.…”

“…There is a wide body of peer-reviewed research in the area of OCHs [2,8,12,13,[15][16][17]20]. However, only Antoniades et al [2] specifically investigated whether the shared files were infringing copyright.…”

“…It tends to underestimate legitimate use cases that do not involve publishing the download links, such as exchanging holiday pictures and other private files, or storing backups. Later works analysed the content types of files downloaded from OCHs as seen in network traces gathered at university networks [16,20] or in crawls of public indexing sites [15], but potential copyright infringement was not investigated.…”