Download archives such as Info-Mac, Tucows and CPAN maintained worldwide networks mirroring their content accessible over HTTP or anonymous FTP.
[10] On occasion, some mirrors may choose not to replicate the entire contents of the upstream server because of technical constraints, or selecting only a subset relevant to their purpose, such as software written in a particular programming language, runnable on a single computer platform, or written by one author.
[25][26][27] Some notable partial mirrors include free and open-source software projects such as GNU,[28] in particular Linux distributions CentOS,[29] Debian,[30] Fedora,[31] and Ubuntu;[32][33] such projects provide mirrors of the download sites (since those are expected to have high server load).
[35] It was once common for tech companies such as Microsoft, Hewlett-Packard or Apple to maintain a network of mirrors accessible over HTTP or anonymous FTP, hosting software updates, sample code and various freely-downloadable utilities.
[38] Occasionally, some people will use web scraping software to produce static dumps of existing sites, such as the BBC's Top Gear and RedFlagDeals.