Two way filesystem synchronization utility based on rclone https://rclone.org.
UpBack assumes a star backup topology, that means it can be used to keep one or more "local" filesystem branches (on your laptop, workstation, workplace pc, etc...) synchronized with a single "remote" backup. rclone is used to access the remote storage so you should be able to use UpBack with all the cloud backup services supported by rclone (this includes DropBox, Google Drive, Amazon Cloud Drive, Microsoft OneDrive and others). Before using UpBack the remote storage must be configured with rclone.
In its current form UpBack is far from being production-ready code: USE IT AT YOUR OWN RISK.
The current version of UpBack requires Python 3. If you need a version supporting Python 2.7 you should check out the branch python27.
UpBack now supports doublestar matches in local exclude files, check the docs below.
First of all you must install rclone. See https://rclone.org/install/ for details. Here we assume rclone is installed and configured with a storage that is accessible with name remote: so, for example, you should be able to list it by running:
rclone ls remote:
UpBack is written in Python and the easiest way to install it is to use Python's own packaging/distribution infrastructure. That means that, if you have Python (currently version 3 is supported, a version running with Python 2.7 is available but should be condidered EOL), you should be able to install UpBack for your user with:
pip install --user git+https://github.com/DavideRossi/upback.git
Alternatively download the git repository as a zipfile ("Clone or download" green button on the top left on https://github.com/DavideRossi/upback then select "Download ZIP"), expand it, and from the upper upback directory run:
pip install --user .
On first run you must tell UpBack the directory on the remote to use for synchronization, the local directory you run this from is assumed to be local root of this backup branch:
upback init-push remote:backup
or:
upback init-pull remote:backup
init-push copies the content of the local branch to remote:backup (removing everything else from remote), init-pull copies from local to remote. To avoid losing data it is recommended to just run init from an empty directory and make sure that remote:backup is empty too.
UpBack always assumes you want to synchronize from the current directory so the typical invocation simply is:
upback
this will look for the configuration files previously created with the init option in the current dir or in one of its ancestors.
If the current directory is not the root of a local branch only a partial synchronization from the current level and below will be performed.
This allows you to perform a partial synchronization without the need to operate on the whole branch.
UpBack will then perform all the operations needed so that the local and the remote branches are synchronized without the risk of losing data (that means that a file is overwritten only by newer versions, or that a file is removed only if it was removed from the other side and the other side contains a fresher version of the files).
If this is not possible and a conflict is detected UpBack aborts creating a conflict file that you can edit.
One by one you can tell UpBack how to deal with a conflict (such as: the local version of a file is to be considered the good one, or both files in local and remote should be deleted, etc...).
Once the conflict file contains all the resolutions you can resume UpBack:
upback resume
These are some command line options that can be used to configure the behavior of UpBack. The most common ones are:
- -v
verbose. Prints various information on what UpBack is doing.
- -vv
more verbose.
- -i
interactive. Asks before performing synchronization operations.
There are two ways to exclude single files or whole branches from the fileset that is synchronized.
The first way is to use the global_excludes field in .upback.config. This is simply a list of the (relative path of the) elements that should not be considered.
The second way is to use .upback.exclude files.
Each line in a .upback.exclude file is a pattern against which elements in THE SAME directory containing the .upback.exclude are matched.
It is also possible to apply the exclusion to files in subdirectories by prepending a **/ in front of the pattern; so **/*.csv excludes all *.cvs files in the current directory and in all its subdirectories as well.
If the match succeeds the element (can be a file or a sub directory) is ignored.
Of course, in the case the matching element is subdirectory, everything inside that subdirectory is ignored as well.
Notice that this applies only to elements of a local filesystem; .upback.exclude in remote filesystems are synched but are NOT processed.
Be warned that this could result in some counter intuitive behavior when items are excluded locally by using .upback.exclude but files with the same path are available at the remote: UpBack will see no local file and will try to copy from remote to local to achieve synchronization, but this way it could end up overwriting local files (it is unaware of, since they are excluded) with the versions in remote (that could be older revisions or have a completely different content).
Please understand that this can cause DATA LOSS so be very careful. I’m thinking about a better way to deal with these cases, when I find a convincing approach I will implement it (and I’m open to suggestions, you can create an enhancement issue in GitHub to get in touch).
- Are symlinks supported?
No, symlinks are deliberately skipped.
If you need them consider to store the "real" files and directories in the local backup branch and link to it from somewhere else.
For example, if you want to backup a src/myproject directory that is outside an UpBack backup branch, put the real myproject directory inside the local UpBack branch, then link it from src/.
This could change in the future, maybe I could use the git approach: when copying from local to remote transform the link to a special text file contaning the path pointed by the symlink; when copyng from remote create a symlink in local pointing to the path stored in the text file in remote. But then I need a way to understand that a file in remote is a special link file...
- What happens if I decide to ignore a conflict
One of the options that can be specified in the conflict file is to ignore a path. Be aware that, the next time you run UpBack it will keep complaining about that conflict until you finally decide how to handle it.
- How do I report a bug?
Please use github's issue tracker https://github.com/DavideRossi/upback/issues
- What is that
.upback.configfile I see on the local root of an UpBack branch?
This is a configuration file storing the details about the backup branch. It is a JSON file and can be edited.
- What is that
.upback.remotefile I see on the local root of an UpBack branch?
That is the last contents of the remote branch seen by UpBack, it is used to decide which operations on the remote branch are to be considered new with respect to the last time UpBack has been executed.