A simple way to securely mirror your data to a remote server. This tutorial requires Passwordless SSH in order for it to work without user intervention (ie, cronjob). Please read my Passwordless SSH tutorial before proceeding.
One of the questions I am frequently asked, is What is the best way to backup my server data? As any good sysadmin will tell you, the answer is backup to as many different medium that you can! A
lot of people new to systems administration make the fatal mistake of
believing that any one backup method alone will be good enough. In
time, we all learn that sooner or later, a failure is going to occur
with that single backup solution. Then what do you do? You will
probably start looking for a new job!
Rsync is a great utility for performing full or incremental backups.
One simple command can copy a single file, or and entire file system to
wherever you want. But that’s not all! Not only does it copy the data
to a new location, but it also compares the files to see if anything
has changed. This results in saved space, and bandwidth (if mirroring
to a remote system).
The first thing we need to do is make sure that rsync is installed on
the client machine. The client is the system we are backing up. We will
refer to these systems in this tutorial as client & server, for
easier understanding. This does not mean that they can’t both be
servers. On an RPM based system such as RedHat, Fedora, CentOS, or SuSE, simply query the RPM database: rpm -qa | grep rsync
If it returns a result, you’re ready to go. if not, you need to
download and install the rsync package for your distribution of linux.
Next, we need to locate the rsync binary. In most cases it will be in
your path already, but when scheduling jobs, or writing scripts, I
usually put the full path to binaries just to be safe. Type which rsync and it should return something like /usr/bin/rsync
if it doesn’t find it, and you made sure that the package was
installed, then it is not in your path. You will need to search for
this using either locate or find.
It is important to remember that trailing slashes do matter!
Unlike the cp command, rsync uses trailing slashes to decide how data
is mirrored. A trailing slash on the source directory means that you
will copy only the files within the source directory. For example let’s
say you have a directory called /stuff which contains 2 files file1
rsync -avz /stuff/ -e ssh username@server:/backup/dir/
Will produce /backup/dir/file1 & /backup/dir/file2
Now take the same command, and remove that trailing slash on /stuff:
rsync -avz /stuff -e ssh username@server:/backup/dir/
Will produce /backup/dir/stuff/file1& /backup/dir/stuff/file2
Now for the actual commands! Rsync has many options, and I strongly suggest you read the man page (man rsync) to learn more. For this tutorial, we will use -a, -v, -z,–delete, -exclude, & -e.
-a = Archive mode. It is equal to using -rlptgoD (man rsync)
-v = Verbose
-z = Compression. This compresses all data before sending it using the same compression method as gzip.
–delete = Delete files on the server that don’t exist on the client
-exclude = Exactly what is says. Excludes a file or pattern
-e = Choose remote shell. We use this to specify SSH
Backup the entire server (deleting missing files, and excluding the proc filesystem):
rsync -avz –delete –exclude=/proc / -e ssh username@server:/backup/dir/
Backup a single folder/filesystem (deleting missing files):
rsync -avz –delete /etc -e ssh username@server:/backup/dir/
Making sense? Good! Now lets automate this sucker.
07 03 * * * /usr/bin/rsync -avz –delete /etc -e ssh username@server:/backup/dir/ > /dev/null 2>&1
Setting up cron is beyond the scope of this tutorial, but a quick rundown of this is:
Mirror /etc to our remote server in /backup/dir/ every night at 3:07am.
Once you get the basics down, It’s easy to make a solid backup
solution. I like to rsync to a 2nd hard drive on my system, then rsync
that to a remote server. This gives me an extra level of protection, in
case my server blows up, or one of the bad guys get in and wipe out my
data. Another cool thing to do, is use hard links (man ln) with
rsync, then setup a rotation schedule say daily for a week. This gives
you 7 full backups, but only uses the space of 1 full backup plus
whatever incremental changes happen over the week. The advantage versus
normal incremental backups, is that it appears as a full backup
everyday. So if you need to restore from say Friday, you don’t have to
restore Sunday first. Just copy the Friday backup.