We often work with clients to benchmark database performance during the vendor selection process and as a result have become experts at working with the various standardized testing data sets available to the public. Almost anyone who has been involved in a performance test is familiar with the Transaction Processing Performance Council (TPC) tests – in particular the TPC-H (see http://www.tpc.org/tpch/default.asp) which is geared towards decision support systems performance testing. This test provides a great starting point for database platform and configuration testing, but recently we have begun to utilize a newer testing suite which we feel provides a better representation of actual use case scenarios – the Star Schema Benchmark (SSB) (see https://github.com/electrum/ssb-dbgen).
SSB is a variation of the TPC-H benchmark that has been modified so that it represents a ‘Kimball’ style database design (i.e. fact tables that consist mainly of keys and very few, if any, text fields). If you’re familiar with utilizing the dbgen tool from the TPC-H test then you won’t have any issue converting to using SSB. As with all performance testing, it’s best to have a game plan as to what exactly your typical use case will look like so that you get the most out of your testing cycles. iOLAP has deep experience in helping develop test case scenarios and overall performance testing projects to help with DMBS platform selection. If you need some help, give us a call.
For detailed information on the SSB tool please reference this whitepaper from the test authors at http://www.cs.umb.edu/~poneil/StarSchemaB.PDF.