continue interrupted downloading of packages without having to redownload [from apt-get download some-package]

Question

When installing packages with apt-get install, it first starts to download all needed packages and then installs them one by one.

But, If you use apt-get download, and interrupt the process during the first phase, and start over again, it downloads all packages again from scratch (erasing any partially-downloaded .deb files).

How can I continue the aborted download process without re-downloading the already received partial data?

Answer 1

Your claim that

If you interrupt the process during the first phase, and start over again, it downloads all packages again.

is completely wrong! Apt is a smart package management system and it never re-downloads packages required to satisfy an installation request twice, not even for a single package or .deb file.

Those .deb files which are half-downloaded reside in /var/cache/apt/archives/partial directory and with ls command (in 16.04, you need to use sudo ls) you can check those files there.

You can verify this by looking at the apt-get manpage, where it says about /var/cache/apt/archives/partial/

  /var/cache/apt/archives/partial/
  Storage area for package files in transit. Configuration Item:
  Dir::Cache::Archives (partial will be implicitly appended)

You can even pick of one of those files and manually resume download using wget -c.

In the background, Apt uses wget program with -c switch, which from man-page means

-c,  --continue                  resume getting a partially-downloaded file

Answer 2

wget -v -d -nc -c -S --random-wait --keep-badhash\
https://mirror.math.princeton.edu/pub/ubuntu/ubuntu/pool/universe/c/chromium-browser/chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb

REBUTTAL & REAL WORKING SOLUTION

Counter-point: If using apt-get download, and restarting, then yes it DOES RESTART OVER FROM SCRATCH.

Platform: I'm using Ubuntu Bionic - 18.04 LTS - x86 (32bit) (...as the engine under Bodhi_Linux_5.1.0's bootable ISO - LiveCD).

again... The answer to this is if using the --download-only option, or apt-get download some-package as opposed to apt-get install some-package ...

if using apt-get download somepackage then YES it WILL RESTART the download FROM SCRATCH.

SOLUTION

My solution to this:

1. backup the existing PARTIALLY-DOWNLOADED .deb file to .deb.BAK

2. use apt-get's --print-uris option to show me the URL of the .deb file to download

3. resume partial-download using wget -c instead of apt-get.... like so: wget -v -d -S -c [--random-wait] [--keep-badhash] https://url-to-file/path/some0package.deb

Code Template for the above steps:

1. $ cp some0package.deb some0package.deb.BAK

2. $ apt-get --print-uris download some0package

3. $ wget -v -d -c -S --random-wait --keep-badhash [copy & paste here the full https://URL you got from the previous command, but NOT the other rest of that output]

Code EXAMPLE using chromium-browser:

$ cp\
chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb\
chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb.BAK

$ apt-get download --print-uris chromium-browser

[RESULTING OUTPUT:]

https://mirror.math.princeton.edu/pub/ubuntu/ubuntu/pool/universe/c/chromium-browser/chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb' chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb 79183996 SHA512:bc34c6241b4200bf9c4c8a7dccba5415daec9623741328146bf4d8edb91d3db7489d5c467162075caf9266644369ac2af0c41cd6b4e320e9c110a6e71411b4ce

In the above output, the --print-uris option gives us several pieces of info; more than we need.

• the URL path to the package-file on the server-side ((THIS FULL URL IS WHAT WE NEED TO COPY!))
• the filename as it expects will be locally on your PC
• the total complete filesize on the SERVERside;
• an SHA512 hash (to do a comparison after it's been downloaded)

Of these, I only need to copy

• the first part, the URL
• optional: the SHA512 hash (if i wish to compare after-download that it matches the server's copy) :

$ wget -v -d -c -S --random-wait --keep-badhas https://mirror.math.princeton.edu/pub/ubuntu/ubuntu/pool/universe/c/chromium-browser/chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb

This worked for me! In summary:

0. I started with the command: $ apt-get download chromium-browser

1. It got interrupted / I canceled (ctrl-c).

2. I backed up the existing file: $ cp [local-filename.deb] [local-filename.deb.BAK]

3. I grabbed the web URL to use via: $ apt-get download --print-uris chromium-browser

4. I highlighted & COPIED TO CLIPBOARD the first part of the OUTPUT from that command, which (for ME) ended up being (don't use this; use the one you get in your terminal ) https://mirror.math.princeton.edu/pub/ubuntu/ubuntu/pool/universe/c/chromium-browser/chromium-browser_105.0.5195.102-0ubuntu0.18.04.1_i386.deb

5. I then used wget -c (the '-c' means 'continue an existing download')... (with added verbose, so i could see if anything went wrong what the problem is) .... and PASTED the URL i'd gotten from the previous command, like so:

$ wget -v -d -c -S --random-wait --keep-badhash\

https://[URL-you-got-and-copied-from-output-of-the-previous-command] somepackage.deb