This paper tackles issues encountered in storing and querying services dealing with information described with Semantic Web languages, e.g. OWL and RDF(S). Our work considers RDF triples stored in relational databases. We assume that depending on the applications and queries asked to RDF triple stores, different partitioning approaches can be considered : either storing all triples in a single relation or using a vertical partitioning where each property is associated to a given relation. We believe that several solutions lie in between these two approaches and we have already proposed roStore has one of them. It consists of an ontology-guided, column-oriented approach particularly efficient for ontologies containing property hierarchies. In a previous paper, we have emphasized that this approach is efficient for queries retrieving information. Due to the adoption of a column-oriented relational approach, an obvious question is : How does it perform on update operations ? The main contribution of this paper is to reply to this question through an evaluation on knowledge bases generated from LUBM. Moreover, our previous work is also extended with (i) a solution to adapt the roStore schema when the underlying ontology is modified and (ii) a formalization of an inference-based query translation from SPARQL to SQL for roStore.