1. Archiving and backup

1. We can use gzip and bzip2 to compress one or more files:

   ls -l /etc > foo.txt

   ls -lh foo.*

   gzip foo.txt

   ls -lh foo.*

   gunzip foo.txt

   ls -lh foo.*

   ls -l /etc | gzip > foo.txt.gz

   gunzip -c foo.txt.gz

   zcat foo.txt.gz | less

   zless foo.txt.gz

   bzip2 foo.txt

   ls -lh foo.*

   bunzip2 foo.txt.bz2

   ls -lh foo.*

2. We can use tar to archive files.

   Let's create a test directory:

   mkdir -p testdir/dir-{001..100}

   touch testdir/dir-{001..100}/file-{A..Z}

   ls testdir/

   ls testdir/dir-001/

   Create a tar archive of the entire directory:

   tar -c -f testdir.tar testdir

   tar -cf testdir.tar testdir

   ls -lh

   The option -c means create, and the option -f is for the filename of the archive.

   The option -t is used to list the contents of the archive, and -v is for verbose:

   tar -tf testdir.tar | less

   tar -tvf testdir.tar | less

   Now let's extract the archive in a new location:

   mkdir foo

   cd foo

   tar -xf ../testdir.tar

   ls

   tree -C | less -r

   cd .. ; rm -rf foo/

3. By default, tar removes the leading / from absolute filenames:

   echo $(pwd)/testdir

   tar cf testdir2.tar $(pwd)/testdir

   tar tf testdir2.tar | less

   mkdir foo

   tar xf testdir2.tar -C foo/

   tree foo -C | less -r

   rm -rf foo

4. We can extract only some files from the archive (not all the files):

   mkdir foo

   cd foo

   tar tf ../testdir.tar testdir/dir-001/file-A

   tar xf ../testdir.tar testdir/dir-001/file-A

   tree

   tar xf ../testdir.tar testdir/dir-002/file-{A,B,C}

   tree

   We can also use --wildcards, like this:

   tar xf ../testdir.tar --wildcards 'testdir/dir-*/file-A'

   tree -C | less -r

   cd .. ; rm -rf foo

5. Sometimes it is useful to combine tar with find and gzip:

   find testdir -name 'file-A'

   find testdir -name 'file-A' \
        -exec tar rf testdir3.tar '{}' '+'

   tar tf testdir3.tar | less

   find testdir -name 'file-B' \
        -exec tar rf testdir3.tar '{}' '+'

   tar tf testdir3.tar | less

   The option 'r' is for appending files to an archive.

   find testdir -name 'file-A' \
       | tar cf - -T - \
       | gzip > testdir.tgz

   The first - makes tar to send the output to stdout instead of a file. The option -T or --files-from includes in the archive only the files listed in the given file. In this case we are reading the list of files from -, which means the stdin and is the list of files coming from the command find. Then we are passing the output of tar to gzip in order to compress it.

   We can also use the options z or j to compress the archive:

   find testdir -name 'file-A' \
       | tar czf testdir.tgz -T -

   find testdir -name 'file-A' \
       | tar cjf testdir.tbz -T -

   ls -lh

   The option j uses bzip2 compression, instead of bzip.

6. The zip program is both a compression tool and an archiver:

   zip -r testdir.zip testdir

   ls -lh

   The option -r is for recursion.

   mkdir -p foo

   cd foo

   unzip ../testdir.zip

   tree | less

   unzip -l ../testdir.zip testdir/dir-007/file-*

   rm -rf testdir
   unzip ../testdir.zip testdir/dir-007/file-*
   tree | less

   cd .. ; rm -rf foo

7. We can use rsync to synchronize files and directories:

   rsync -av testdir foo

   ls foo

   rsync -av testdir foo

   Notice that in the second case no files are copied because rsync detects that there are no differences between the source and the destination.

   touch testdir/dir-099/file-Z

   rsync -av testdir foo

   With the option --delete we can also delete the files on the destination directory that are not present on the source directory.

   rm testdir/dir-099/file-Z

   rsync -av testdir foo

   ls foo/testdir/dir-099/file-Z

   rsync -av --delete testdir foo

   ls foo/testdir/dir-099/file-Z

   rsync can be used over the network as well, usually combined with ssh.

Download lesson08/part1.cast

Loading asciinema cast...