Task 1-2

  • Scalability. Currently, web preemption results are established with a KG stored in an HDT [36] file. However, HDT does not scale, since HDT requires to load all the index of the KG into the memory. We need to study how web preemption can be deployed on scalable datastores such as Cassandra, HBase or BigTable while preserving linear scalability of the underlying datastore. If many works explore how to store RDF data in a distributed datastore, the original workload generated by the web preemption approach suggests that an optimal exists.

Comments are closed.