TPCTC 2011 followed the great success of previous TPCTC conferences. We were honored that Umeshwar Dayal, Director for the Intelligent Enterprise Technology Lab at HP Labs in Silicon Valley, agreed to give this year’s keynote about “Optimization of Analytic Data Flows for Next Generation Business Intelligence Applications”. Another highlight of the conference was the invited talk by Karl Huppler, Power Systems Performance Engineer at IBM Rochester entitled “Benchmarking With Your Head in the Cloud”.
Many deserving papers were not included in the conference due to time limitations. Out of 27 papers that were submitted to the conference 12 were accepted, 37% of which were submitted from academia, 26% from industry and 37% from collaborations between industry and academia. In addition to submissions from industry leading companies and accredited national and international universities we are especially proud to have received submissions from industry labs such as the Hasso-Plattner Institut and Hewlett-Packard Labs.
Every paper presented at the conference has the potential of making a significant impact on current and future benchmark development. However, the following four papers are already influencing the TPC. The first paper entitled “Extending TPC-E to Measure Availability in Database Systems” details extensions to the TPC-E benchmark for availability measurements, including High Availability scenario simulation, fault injection, and availability metric reporting, which are critical to database customers. The paper illustrates implementation details and exemplary test results on SQL Server 2008 Database Mirroring. The second and third papers entitled “Metrics for Measuring the Performance of the Mixed Workload CH-benCHmark” and “Normalization in a Mixed OLTP and OLAP Workload Scenario” describe ideas on how to implement a mixed workload benchmark that fuses Online Transaction Processing and Decision Support workloads on the same database. The fourth paper entitled “Introducing Skew into the TPC-H Benchmark” details plans on how to implement data skew into the well known TPC-H benchmark. Currently TPC-H uses uniform data distribution. Having data skew in TPC-H dramatically enhances the benchmark.
We are excited about the success of TPCTC 2011 and are looking forward to next year’s event. TPCTC 2012, to be held in conjunction with VLDB 2012 in Istanbul, Turkey from August 27 to 31. has been approved by the TPC and is now subject to approval of the VLDB endowment.