ABSTRACT:
As one of the most popular cloud services, data storage has attracted great attention in recent research efforts. Key-value (k-v) stores have risen as a popular option for storing and querying billions of key-value pairs. So far, existing strategies have been deterministic. Giving such accuracy, however, comes at the cost of memory and CPU time. Interestingly, we display an inexact k-v storage for cloud-based systems that is more reduced than existing techniques. The tradeoff is that it might, hypothetically, return errors. Its design depends on the probabilistic data structure called “blossom filter”, where we stretch out the classical blossom channel to help key-esteem activities. We call the subsequent outline as the kBF (key-value bloom filter). We additionally build up a distributed version of the kBF (d-kBF) for the unique requirements of distributed computing platforms, where multiple servers participate to deal with a large volume of queries in a load balancing manner. Finally, we apply the kBF to a practical problem of executing a state machine to show how the kBF can be utilized as a building block for more complicated software infrastructures.
