Using Your Data for Testing
Scalability Testing for DB2
Tuning the Oracle RAC Interconnect with BMF and Spotlight
Optimize Oracle 10g on Linux, Non-Rac ASM vs. LVM
Load Testing Oracle RAC for Performance and Scalability
The Top Five Database Benchmarking Questions
We have added several new documents to our library. The first discusses how to scale test your DB2 programs.
The second shows how to incorporate application specific data into your programs and queries by reading values from a text file or by simply using random data.
Here’s an updated list of our documents -
This shows how to take values in a text file and send to queries in a BMF project.
 Last Modified: Jun 4, 2007 1:14 PM
Presentation showing how to stress test your programs for a DB2 instance.
 Last Modified: Jun 4, 2007 1:15 PM
Tuning the RAC Interconnect
I have been asked many times how to check the tuning of the interconnect used in an Oracle RAC installation. Oracle themselves provide numerous views that the industrious DBA can build reports from to perform this monitoring and then of course the data can be placed into a spreadsheet and graphed to show the results. Of course all of this reporting, moving to spreadsheet and graphing can be time consuming and be prone to error (for example, if the DBA doesn’t use delta values.) In this series of tests I used the Quest Benchmark Factory to run a TPC-D Power test and used the Quest Spotlight on RAC product to monitor the interconnect performance as I modified the TCP buffer settings. Figure 1 shows the start screen for the Benchmark Factory TPC-D Job. The TPC-D job was configured to perform 10 runs with 10 users spread out as 5 users per node in a 2-node RAC environment.
 Last Modified: Jun 4, 2007 1:16 PM
Bert returns after his series of Whitepapers regarding Oracle on Linux with ASM with this followup attempts to answer the question, “Does ASM perform as well as Linux file systems using a logical volume manager?”
In my first ASM paper, I simply assumed that people either would or would not be utilizing ASM—without considering Real Application Cluster (RAC) usage ramifications. What numerous people have told me this past year at shows, conferences and on-site visits is that while ASM makes obvious sense for RAC environments, they also desired to know whether ASM was in fact a viable alternative for non-RAC environments. Specifically, does ASM perform as well as Linux file systems using a logical volume manager?
Of course that’s a challenge far too enticing to pass up—especially when there are tools like Quest Software’s Benchmark Factory® that make such tests trivial. So, on to the races.
 Last Modified: Jun 4, 2007 1:17 PM
An Oracle Real Application Cluster (RAC) combines several independent servers/nodes into one database system. RAC provides IT database infrastructures with a variety of
features that include:
When a node in an Oracle RAC cluster fails, an Oracle database and the remaining nodes continue to run, providing uninterrupted service to users. Oracle RAC distributes
database demands over several nodes, performing the function of load balancing. When more capacity is required, additional nodes can be added with minimal effort. Organizations spending time and money setting up a RAC cluster must ensure their systems have the stability to withstand user environment demands when adding nodes.
Without knowing how their system reacts to these demands, DBAs operate blindly and jeopardize their entire IT enterprise.
This white paper presents a load testing methodology to implement and manage Oracle RAC clusters, to obtain optimum user service regardless of the demands placed on the
 Last Modified: Jun 4, 2007 1:18 PM
This whitepaper will address the five most common benchmarking questions. Read the whitepaper