4

I recently downloaded Ubuntu 10.04 via HTTP from the Ubuntu website, the download reached 663.0 MB of 685.6 MB before it was accidentally canceled. Is it possible to continue the download using the already downloaded data?

A torrent is available for the exact same file.

Improve this question
3
  • Your question and its tags are confusing - is this a canceled torrent or a straight HTTP download?
    – Zayne S Halsall
    Dec 4, 2010 at 17:18
  • 1
    It is a canceled HTTP download. I am assuming that is is easier to continue via torrent.
    – Zaz
    Dec 4, 2010 at 17:40
  • It's better to use Torrent file than HTTP download because your sharing rate and resume support guarantee is more in torrent.
    – Mahesh
    Dec 4, 2010 at 18:46

3 Answers 3

Reset to default
6

You could use wget. Open a command line, navigate to the directory of your incomplete download, type 

wget --continue your_address_here
Improve this answer
4
  • 1
    Could you also demonstrate how to specify a target file name?
    – oKtosiTe
    Dec 4, 2010 at 18:01
  • 1
    wget looks in the current directory for a file with the same name as the file given in the download url. For example, wget --continue foo.com/bar.iso would look for a file named bar.iso in your current directory and continue to download it, if it was incomplete.
    – user54114
    Dec 4, 2010 at 18:09
  • 2
    +1. That will work, and will be simpler, if the partial download was done by something that only uses one TCP stream at a time (so the missing data is all at the end) and doesn't pre-allocate (so the end of file clearly marks where to restart). Most browsers' built-in download support works this way. If it was partially downloaded by something like DownThemAll and was properly cancelled (so the client forgot which parts it had no yet downloaded) you have to use something other than HTTP (i.e. bittorrent or rsync).
    – David Spillett
    Dec 4, 2010 at 18:11
  • I never knew this was possible! It would be pretty handy to have this feature built-in.
    – Zaz
    Dec 4, 2010 at 18:20
13

You can use the torrent to resume the HTTP download if they are exactly the same file (which in this case they are likely to be).

  1. Get the .torrent meta-file
  2. Start downloading it with your preferred bittorrent client
  3. Stop the download completely
  4. Replace the file the bittorrent client created with a copy of the one that resulted from the failed http download making sure the filename is exactly the same (no differences in letter case and so forth)
    (if you know for sure that the filenames are exactly the same, you can skip step 2 and 3)
  5. Reload your bittorrent client and start it downloading again. Before it starts the transfer it should do a full hash check to find out which parts of the file it needs to get and which parts are already complete, and so will know only to download the blocks it needs
Improve this answer
3
  • 1
    Also, some torrent clients offer the ability to manually force a check of the downloaded files. If that is the case in yours, you might want to do that before restarting the download as explained in step 5.
    – oKtosiTe
    Dec 4, 2010 at 17:40
  • 1
    Thanks, I had tried this already but I must have misnamed the file causing the torrent to start afresh. This has solved the problem.
    – Zaz
    Dec 4, 2010 at 17:55
  • 1
    @oKtosiTe: Thanks, I'm using Transmission which does this automatically.
    – Zaz
    Dec 4, 2010 at 17:56
2

When I have this happen I usually use rsync to continue the download. mirrors.kernel.org offers an rsync server, so I'll go to http://mirrors.kernel.org/ and navigate to the file I want to download. Then I copy the path to that file and replace the "http://" with "rsync://". For example, the Ubuntu 10.10 i386 desktop ISO is at:

http://mirrors.kernel.org/ubuntu-releases/10.10/ubuntu-10.10-desktop-i386.iso

So I can resume the download of it by running:

rsync -avPL rsync://mirrors.kernel.org/ubuntu-releases/10.10/ubuntu-10.10-desktop-i386.iso .

Note the "-L" argument, without that only the link to the file is copied because Ubuntu uses links to the files in different locations.

The above assumes that your file is the same name ("ubuntu-10.10-desktop-i386.iso") and is in the current directory. If that is not true, replace the trailing "." above with the actual file name and path to it.

Improve this answer
3
  • Good tip, but it only applies to certain files so no cigar.
    – Zaz
    Dec 12, 2010 at 13:48
  • Well, the original question was about grabbing Ubuntu ISOs, so ... cigar!
    – Sean Reifschneider
    Dec 12, 2010 at 20:54
  • Haha :P Man, I'm sorry, you deserve an upvote for that.
    – Zaz
    Jul 12, 2014 at 14:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .