Scalable in-memory data store systems based on mixed-consistency data types and replication algorithms for efficient and anomaly-free data management in the emerging NVM-based computer architectures
The amount of data stored in data centers and processed by cloud services is growing constantly and rapidly. Therefore a new generation of data store systems is required, which will offer high efficiency, availability, and robustness for a large number of concurrent users. The main goal of the project is to propose cutting-edge solutions that can be used to build robust data store systems which scale, i.e., the system throughput grows with an increasing number of nodes or processors. The solutions include novel replication methods and algorithms with support for transactions (taking advantage of multi-core CPUs), replicated data structures that support both strong and eventual consistency, and novel recovery algorithms, which envisage the use of persistent non-volatile memory (NVM) in the future computer systems. The results of our work will be evaluated formally and experimentally. The project will be done in collaboration with a Foreign Partner.
Twój opis projektu
MiAby dodać zdjęcie, opis lub załączyć inne materiały dotyczące projektu, powinieneś wypełnić poniższy formularz. Po kliknięciu przycisku „wyślij” zostanie przesłana do Ciebie wiadomość e-mail z prośbą o potwierdzenie. Do przesłania formularza niezbędne jest zaakceptowanie Regulaminu Mapy Dotacji UE oraz zgoda na przetwarzanie danych osobowych. Materiały zostaną przesłane do redaktora serwisu, który po sprawdzeniu opublikuje je na stronie.