Last night was Oracle Midlands Even #14 with Lothar Flatz.
The session started with “Raiders of the Data Dictionary I: Indexing for the Workload” where Lothar discussed a project he worked on based around completely revamping the indexing of a system. I guess the best way I can describe it is to say it’s a more scientific approach to indexing, using the contents of the dictionary to provide information about columns and column groups used in queries to determine the indexes to create. In some ways the approach was quite extreme (throw everything away and start again), but I also admire the bravery of that approach!
Next up was “Raiders of the Data Dictionary II: The Curse of the Buffer Cache” where Lothar discussed the problems associated with trying to keep large, frequently used tables in the buffer cache. Once again, some interesting points made and some things that will definitely influence my approach in future.
It was great to meet Lothar in person for the first time. After being in the game for 20+ years many subjects can start to feel a little repetitive, so it’s nice when someone comes along with a different spin on a subject. I certainly found myself asking a lot of questions of him and myself, which surely must be what this knowledge spreading thing is all about!
Thanks to Lothar for taking the time to come and speak to us and to everyone that came along to listen. Thanks to Mike for doing a great job in keeping Oracle Midlands going and to Red Stack Tech for their continued support.
See you at the next one! 🙂