Design, implementation, and evaluation of the linear road bnchmark on the stream processing core
Proceedings of the 2006 ACM SIGMOD international conference on Management of data
The CQL continuous query language: semantic foundations and query execution
The VLDB Journal — The International Journal on Very Large Data Bases
Experiences with MapReduce, an abstraction for large-scale computation
Proceedings of the 15th international conference on Parallel architectures and compilation techniques
SPADE: the system s declarative stream processing engine
Proceedings of the 2008 ACM SIGMOD international conference on Management of data
Clustera: an integrated computation and data management system
Proceedings of the VLDB Endowment
Exploiting the power of relational databases for efficient stream processing
Proceedings of the 12th International Conference on Extending Database Technology: Advances in Database Technology
Extend UDF Technology for Integrated Analytics
DaWaK '09 Proceedings of the 11th International Conference on Data Warehousing and Knowledge Discovery
Efficiently support MapReduce-like computation models inside parallel DBMS
IDEAS '09 Proceedings of the 2009 International Database Engineering & Applications Symposium
Cooperating SQL Dataflow Processes for In-DB Analytics
OTM '09 Proceedings of the Confederated International Conferences, CoopIS, DOA, IS, and ODBASE 2009 on On the Move to Meaningful Internet Systems: Part I
Dynamic migration of processing elements for optimized query execution in event-based systems
OTM'11 Proceedings of the 2011th Confederated international conference on On the move to meaningful internet systems - Volume Part II
| Hi-index | 0.00 |
Many mobile applications are based on cloud services such as location service, messaging service, etc. Currently most cloud services are based on statically prepared information rather than the real-time analytics results of dynamically captured events. A paradigm shift is to take Continuous Stream Analytics (CSA) as a cloud service, which, however, poses several specific challenges in scalability, latency, time-window semantics and transaction control. In this work we extend the SQL query engine to unify the processing of static relations and dynamic streams for providing the platform support of CSA service. This platform is significantly differentiated from the current generation of stream processing systems which are in general built separately from the database engine thus unable to take advantage of the functionalities already offered by the existing data management technology, and suffer from the overhead of inter-platform data access and movement. To capture the window semantics in CSA, we introduce the cycle-based query model and support it in terms of the cut-and-rewind query execution mechanism. This mechanism allows a SQL query to run cycle by cycle for processing the unbounded stream data chunk by chunk, but without shutting the query instance down between chunks for continuously maintaining the application state across the execution cycles, as required by sliding-window oriented operations. We also propose the cycle-based transaction model with cycle-based isolation and visibility. To scale-up analytics computation, we introduce the parallel infrastructure with multi-engines cooperated and synchronized based the common data chunking criteria without centralized coordination. To scale-up service provisioning, we investigate how to stage the continuously generated analytics results efficiently through metadata manipulation without physical data moving and copying. We have prototyped our approach by extending the PostgreSQL, resulting in a new kind of tightly integrated, highly efficient platform for providing CSA service. We tested the throughput and latency of this service using a well-known stream processing benchmark and with WebOS based Palm phones. The test results show that the proposed approach is highly competitive. Providing CSA cloud service using HP Neoview parallel database engine is currently explored.