AMFETCHDUMP(8)

✨ NAME ✨

amfetchdump β€” pull backup images off one or more Amanda tapes, no big deal πŸ’–


πŸŽ€ SYNOPSIS πŸŽ€

amfetchdump [-c | -C | -l] [-p | -n] [-a] [-O directory] [-d device] [-h
            | --header-file filename | --header-fd fd] [--release-tapes
            | --reserve-tapes] [--decompress | --no-decompress |
            --server-decompress | --client-decompress]
            [{--extract | --extract-client=HOSTNAME} --target target [--data-path amanda|directtcp] [--run-client-scripts] [--prev-level level] [--next-level level] [--application-property NAME=VALUE]* [--include-file file]* [--include-list filename]* [--include-list-glob filename]* [--exclude-file file]* [--exclude-list filename]* [--exclude-list-glob filename]*]
            [--init] [--restore] [--decrypt | --no-decrypt |
            --server-decrypt | --client-decrypt] [--exact_match]
            [-o configoption...] config hostname
            [disk [ date [ level [ hostname [...] ] ] ]]

πŸ¦‹ DESCRIPTION πŸ¦‹

Okay so amfetchdump yanks one or more matching dumps off tape, or off the holding disk, and it's literally doing the most: it handles reassembling multi-tape split dump files and any tape autochanger operations for you. By default the dump gets decompressed and decrypted. Periodt. πŸ’…

It's gonna automatically hit up the Amanda catalog to find which dumps are actually available on tape β€” same energy as the find feature of amadmin(8) when it lists available dumps. Iconic behavior, honestly.

The hostname, diskname, datestamp, and level dump specs are described in way more detail over in amanda-match(7). But know this: at minimum you gotta specify a hostname. Non-negotiable, bestie. 🫢

Unless -p is used, the backup images get extracted to files in the current directory, named like this:

hostname.diskname.datestamp.dumplevel

And if a changer error happens, or if the -d option is passed, then amfetchdump will prompt you for each required volume. Yes, it talks to you. Answer it. πŸ™ˆ


πŸ’– OPTIONS πŸ’–

  • -p β€” Pipe exactly one complete dump file to stdout instead of writing the file to disk. Heads up: this restores only the first matching dumpfile (and "first" is decided by the dump log search facility, so don't argue with it).
  • -h β€” Output the amanda header as a 32K block to the same output as the image.
  • --header-fd fd β€” Output the amanda header to the numbered file descriptor.
  • --header-file filename β€” Output the amanda header to the filename.
  • -d device_or_changer β€” Restore from this device or changer instead of the default, prompting for each volume.
  • -O directory β€” Output the restored files to this directory instead of the current working directory.
  • -c β€” Compress the output, using the fastest method available.
  • -C β€” Compress the output, using the smallest file size method available.
  • --decompress β€” Always do the decompression. This is the default.
  • --no-decompress β€” Never do the decompression.
  • --server-decompress β€” Do the decompression only if the compression was done on the server.
  • --client-decompress β€” Do the decompression only if the compression was done on the client.
  • --decrypt β€” Always do the decryption. This is the default.
  • --no-decrypt β€” Never do the decryption.
  • --server-decrypt β€” Do the decryption only if the encryption was done on the server.
  • --client-decrypt β€” Do the decryption only if the encryption was done on the client.
  • --init β€” Prepare for a restore. Some devices need this for a faster restore.
  • --no-restore β€” Do not start the restore. Useful when you're pairing it with --init.
  • --exact-match β€” The host and disk are parsed as exact values.
  • --extract β€” Extract the backup on the server, in the target path.
  • --directory directory β€” Deprecated, bestie. Use --target instead. πŸ‘‘
  • --target target β€” Where to extract the backup to when you're using the --extract option.
⚠️ Warning: all files in that path can be removed. Like, all of them. Be so serious right now.
  • --data-path amanda|directtcp β€” The data path to use with --extract. The default is to use the fatest data path. πŸ’€
  • --run-client-scripts β€” Run the client script, if it is set.
  • --prev-level level β€” The previous level that was restored for this dle. If it's not set, run the pre-recover script. If it is set, run the inter-level-recover script.
  • --next-level level β€” The next level we will restore for this dle. If it's not set, run the post-recover script.
  • --application-property NAME=VALUE β€” Application property to send to the application with --extract.
  • --include-file file β€” The name of a file to restore.
  • --include-list filename β€” The name of a file containing files to restore.
  • --include-list-glob filename β€” The name of a file containing files to restore, using --include-list-glob of the application.
  • --exclude-file file β€” The name of a file to exclude from the restore list.
  • --exclude-list filename β€” The name of a file containing files to exclude from the restore list.
  • --exclude-list-glob filename β€” The name of a file containing files to exclude from the restore list, using --exclude-list-glob of the application.
  • --release-tapes β€” Release the tapes that were already reserved for the restore.
  • --reserve-tapes β€” Reserve the tapes needed for the restore.
  • -l β€” Leave dumps in whatever compressed/uncompressed and encrypted/unencrypted state they were found in on tape. It's a synonym for --no-decompression and --no-decryption.
  • -a β€” Assume that all tapes are already available (via tape changer or otherwise) instead of prompting the operator to make sure all tapes are loaded. Vibes: trusting. 😌
  • -n β€” Do not reassemble split dump files at all β€” just restore each piece as its own individual file.
  • -o configoption β€” See the "CONFIGURATION OVERRIDE" section in amanda(8).

🌈 EXAMPLES 🌈

All the examples down here assume your configuration is called SetA. πŸ“

Here's a simple case β€” restoring all known dumps of the host vanya to the current working directory:

$ amfetchdump SetA vanya

A more likely scenario involves restoring a particular dump from a particular date. We'll pipe this one to GNU-tar as well, so it automatically extracts the dump:

$ amfetchdump -p SetA vanya /home 20051020 | gtar -xvpf -

😭 CAVEATS 😭

Real talk: amfetchdump depends on being able to access your server's config, tape changer, and (normally) dump logs. So like… it's not necessarily the most useful tool when all of those have been completely wiped out and you're desperately trying to drag things off your tape. Pains have been taken to make it as capable as possible, fr fr, but for seriously minimialist restores, go look at amrestore(8) or dd(8) instead. πŸ’œ


⭐ SEE ALSO ⭐

The Amanda Wiki: : http://wiki.zmanda.com/


πŸ‘‘ AUTHORS πŸ‘‘

  • John Stange <building@nap.edu> β€” National Academies Press
  • Ian Turner <ian@zmanda.com> β€” Zmanda, Inc. (http://www.zmanda.com)

Okay that's the whole thing, you ate it up πŸŽ€ This was Amanda 3.5.4, dated 08/26/2026 β€” go forth and restore responsibly, bestie. Slay. πŸ’…βœ¨πŸ¦‹

πŸ“¦ where this page comes from