Software & Links

Software Results 

• Pravega: Pravega provides a new storage abstraction - a data stream - for continuous and unbounded data. A Pravega stream is a durable, elastic, append-only, unbounded sequence of bytes that has good performance and strong consistency.

• Crystal: Crystal is the first transparent, dynamic and open Software-Defined Storage system for OpenStack Swift. With Crystal, administrators can easily develop and install new "storage filters" in the object store. Then, administrators can define simple "storage policies" to execute such storage filters on object requests of users. The enforcement of storage filters can be done statically or dynamically based on real time workload metrics. Workload metrics and filters can be dynamically plugged into the system to expand the capabilities of the object store.  

• SDGen: SDGen is a synthetic data generator for storage benchmarks. The objective of this framework is to enable users creating methods to generate realistic data to feed storage benchmarking tools.

• BenchBox: BenchBox is a tool that adopts a user-centric performance evaluation methodology based on user stereotypes. An stereotype describes the behavior of a group of users with similar storage activity, which enables flexible and realistic performance evaluation of Personal Clouds. BenchBox enables researchers and practitioners to deploy benchmarks with varying rates of user stereotypes, as well as to monitor the desktop clients activity.

• MChannel: M-Channel is a Topology-Aware Group Communication Middleware for MANETs based on the JGroups Tookit. We have modified the JGroups protocol stack to incorporate two protocols: jOLSR and OMOLSR. jOLSR is a Java implementation of the MANET OLSR unicast protocol using UDP. On top of it, we have developed an overlay Multicast protocol (OMOLSR) that directly benefits from the jOLSR protocol to improve communication efficiency. In our middleware, group membership is obtained from the overlay multicast protocol (OMOLSR), failure detection is received from the jOLSR protocol, and our modified flow control protocol benefits from jOLSR topology information.

Datasets

• Sharing traces of NEC Personal Cloud: The NEC dataset integrates two sources of information: storage layer and sharing interactions. The information to build our dataset was collected directly by the provider from the back-end (SQL Server & OpenStack Swift).

• Personal Clouds Measurement: This dataset comes from a measurement study of three major Personal Clouds: DropBox, Box and SugarSync. Actively accessing to free accounts through their REST APIs, we analyzed important aspects to characterize their QoS, such as transfer speed, variability and failure rate. Our measurement, conducted during two months, is the first to deeply analyze many facets of these popular services and reveals new insights, such as important performance differences among providers, the existence of transfer speed daily patterns or sudden service breakdowns.

• UbuntuOne trace: This trace contains the back-end activity of UbuntuOne desktop clients (1.29 million clients) for one month.