NubiSave is a freely available user-space controller for Redundant Arrays of Inexpensive, Independent or even Optimal Clouds (RAIC/RAOC).
Through dispersion and other techniques, the data is secured beyond encryption and stored safely on the individual storage targets, devices, media and service providers. NubiSave and its associated tools have initially been developed in the context of the FlexCloud young investigator group and are now being continued under the auspices of OSSPRI.


  • Filesystem built on FUSE which integrates with a number of cloud-storage-related FUSE modules, including the advanced CloudFusion adapter
  • Integrates with the SPACE service platform, in particular the Cloud Storage Ontology, for optimal storage provider selection, allocation and scheduling
  • Offers extensible tree-like interfaces for nested splitters, encryption, steganography, compression, deduplication etc.
  • Includes a Java setup/configuration GUI, now with graph view, and a (not yet complete) alternative PyQT setup GUI
  • Associated projects: NubiVis - distributed data visualisation and browsing, peaCS - multi-cloud storage integration testsuite
  • Technical features: optimal redundancy, cache, parallelism, streaming, chunking, sessions, ...
  • Developer features: functional unit tests, UI unit tests


06. February 2014:
We now offer split (modular) Debian packages for installation through an APT repository. The latest version contains several usability improvements, including directory selectors and mouse hover display in the flow editor, as well as attribute preservation fixes in the splitter. There is now also a preliminary version of NubiGate available which runs a complete NubiSave and NubiVis installation inside a gateway VM.
30. November 2013:
A new snapshot has been packaged. It will be shown at CloudCom'13 and UCC'13.
23. October 2013:
We have successfully demoed NubiSave and the π-Box (personal secure cloud box) at the KIT/KSRI Service Summer school in Karlsruhe. About 20-30 participants, mostly PhD students, have received a USB stick with our tools, supported by SUCRE who ran the Young Researcher Forum as Cloud Computing track. This was the second big tutorial for us after an international practical course with KPI students.
25. August 2013:
The Cloud Storage Lab aka. NubiSave Lab contains preliminary results of our cloud storage integration measurement testsuite peaCS which has been worked on during a bachelor thesis. We are glad to be given the opportunity to present peaCS at the CLOUSO workshop on cloud storage optimisation!
11. June 2013:
We're working on a novel distributed data visualisation tool which lets users explore the NubiSave file/fragment database. Along with a graphical view, there will be a command-line tool and a RESTful service to query the database. Furthermore, we're currently refactoring the configuration GUI user interface and revamping the performance test suite. Stay tuned for more changes and follow the progress in the Git repository!
07. December 2012:
Two of our students have teamed up to create an awesome video which demonstrates the basics of dispersion and RAOC. You can watch it online at or download the whole explanation video. We celebrate this video by making a new Debian package of NubiSave available which includes the new graph GUI and a FUSE-J module fix.
26. November 2012:
A concept for a policy-driven cloud storage gateway, and in particular a policy language to express storage goals and constraints, has been published at the LatinCloud conference. Read about it on the Learn page.
07. June 2012:
Minor updates and corrections in today's snapshot.
20. April 2012:
A new snapshot has been packaged, with vastly extended functionality over the previous one. Get it from the installation page.
23. December 2011:
NubiTree is an experimental PyQt GUI for achieving a more intuitive representation of the cloud storage module tree. It works on both the actual NubiSave configuration and an easily demonstratable simulated tree. See screenshot!
16. December 2011:
The publication outlining the concepts behind NubiSave has been published and presented at the UCC 2011 in Melbourne, Australia. Apparently, it covers a hot topic since it has received a Best Paper Award! Meanwhile, development continues towards a more usable and powerful prototype.
29. October 2011:
Website is fairly informative now and goes public. We invite brave and curious people to try out the current state of NubiSave as long as only test data is stored.
08. September 2011:
Website goes online (in [p]review mode).
07. September 2011:
The NubiSave code repository is now publicly available. This snapshot represents the outcome of the practical lab exercise. Don't use it yet for storing data without additional backups!


NubiSave is developed in a git repository and released under open source licences as code and as pre-packaged application. Have a look at the installation instructions to learn about the installation options.

Latest release

Not yet. For the time being, please build from git.

Previous releases

Not yet.