Good options to use for httrack to mirror a large-ish site.
Quick copy and paste
httrack --connection-per-second=50 --sockets=80 --keep-alive --display --verbose --advanced-progressinfo --disable-security-limits -n -i -s0 -m -F 'Mozilla/5.0 (X11;U; Linux i686; en-GB; rv:1.9.1) Gecko/20090624 Ubuntu/9.04 (jaunty) Firefox/3.5' -A100000000 -#L500000000 'YOURURL'
- ignores robots.txt
- allows for a queue of 500M unfetched URLS
- custom useragent
- pretty fast (uses several connections at once)
- will re-write links so they work offline
A rundown of the previous options
--connection-per-second=50: This allows for up to 50 connections per second.
--sockets=80: Opens up to 80 sockets. If this gives you errors, lower this to 48.
-A100000000: By default, HTTrack attempts to play nicely with webservers, and tries not to overload them by limiting the download speed to 25kbps. On text-based sites this is normally good, but it becomes a hassle when the site is image-heavy. The first option disables the forced limit and the second one raises the limit to a large amount.
-s0: Tells HTTrack to disobey robots.txt.
-F: Sets the user agent.
-#L500000000: Raises the maximum amount of links HTTrack fetches to 500M. Raise if needed.
-n: This gets all nearby files (all files shown on a page), rather than only those on the domain name, which is HTTracks default behavior.
- A full rundown of all possible options (including those on site structure) can be found here: https://www.httrack.com/html/fcguide.html .
NOTE: httrack runs java internally (I believe) and is limited to 2GB of ram. Not sure if a 64-bit version of it will allow for a larger crawl queue.