The current information technology trends toward the distribution of data among numbers of autonomous and heterogeneous repositories. Object Query Service is specified by OMG to provide uniform interfaces for users to retrieve information based on predicate-based query despite of data heterogeneity. In this paper, we propose a sophisticated architecture for querying federated information for CORBA paradigm based on Object Query Service framework. The dissimilarity between our system and other implementation of Query Service, which focus on the query processing and system performance is that we concentrate on the requirements, behaviors, scalability as well as the extensibility of the federated query system from different participants point of view. We design a prototype called Octopus to demonstrate the feasibility of our architecture. We also share the experiences learned while designing the federated query service.