Welcome to a Supplemental Executive Interview in Module 4. This interview provides a context for the assignment in Module 4. You will learn about summary data management practices from a former Data Warehouse administrator and now Oracle technical evangelist. Let me welcome Kellyn Gorman, consulting member of the Oracle Strategic Customer Program Team, focused on the enterprise manager. Kellyn has an extensive background about database administration, data warehouse performance and Oracle expertise. I'm excited and honored to have this opportunity, to share her insights with learners involved in this capstone course. Kellyn, briefly tell the learners about your background, especially your expertise in data warehouse administration, data warehouse performance. >> I've been a database administrator for approximately 17 years, another 3 years as a consultant with Oracle. I worked in everything from multi terabyte environments to databases of service, AW warehouse. We incorporate a number of environments for high level customers that are looking for specialists that are able to implement systems that they may not be able to do on their own. >> Kellyn, thanks for sharing your background with the learners. Now can you tell the learners about the importance of materialized views in data warehouse environments as you've seen, are they under utilized? How's the SQL axis advisers? Is that important to helping customers that you've seen utilize materialized views? >> Materialized views are a object that many people kind of confuse with regular views, but they are a physical object that you're creating. Many times that people like to have standard views and they'll actually migrate those to materialized views and that's not the best way of utilizing them. You should be seeing what is the highest hit in the SQL tuning advisory, any of those will tell you what those are. And take that piece of maybe a standard view or a query, and turn that into a materialized view and then rewrite the query that you may have used originally. So you can take down what is heavy hash joins, heavy sorting on the system. That is going to eliminate a lot of IO times, especially when we talk about an OLAP system or DSS. Anything that's large. This is going to take down that IO actual execution time. >> Well, Kellyn, thanks for sharing about the importance of materialized views in a data warehouse environment. Now, can you talk about the process of analyzing query performance. And why the material that is used may be important, such as the use of the AWR, or the Automatic Workload Repository tool that's provided by Oracle. >> The AWR, as you stated, the Automatic Workload Repository is the de-facto tuning environment for any DBA. The ability to run AWR reports as well as ASH reports, which ASH is Active Session History. Allows you the ability to get reports that give you information specially about a lapsed time. They'll break it down by CPU, IO, any kind of resource, standard resource usage. But a lapse time is important. Most people who tune never think about time, and if you're not tuning time, you're really wasting time. So if you use an AW report and see what your top elapse time SQL, you can utilize that and see what might be a good example query to turn into a materialized view. But you're going to look at the execution plans and see what is taking up the most time in the processing of that statement. >> Kellyn, thanks for sharing your insights about performance analysis of data warehouse queries. Now can you provide some lessons? Lessons that may involve trends, or may involve best practices customers as have difficulty with. >> Many times you see where DBAs and developers aren't working together. And both of them provide very different skills, that can create an environment that's going to run most effectively and productively for the company. If they work together, and where the DBA may understand the database environment overall, where the developer is actually fulfilling requirements. If they work together, they can create code that is going to be released production, and work as flawlessly as possible. Many of the companies that I go to, my whole job is to optimize the app, the code, and the hardware to work together, in the best fashion. If you're going to do that, that's one of the biggest deals for me. I do a lot of just facilitating between those groups and getting them to cooperate. So that's important and I see that a lot of times where the DBA against the developer, the developer against the DBA. And IT as a whole is extremely important to building the best infrastructure, to provide the business what they need. >> Kellyn, given your extensive experience in data warehouse administration now consulting with Oracle, what advice can you offer to learners? Maybe learners who taking a specialization about starting a career in data warehouse administration? >> There are tons, it sounds hilarious, but there's tons of meetups out there. There's a lot of opportunities to learn about different technologies. Never limit yourself to one. As an OLAP or a DSSDBA. I also learned about OLTP or transactional systems. I learned about different platforms of databases, because you found that one platform may know a little bit more than another platform, or something this feature here was interesting. I learned about a different languages. I've taken the time to learn Ruby on Rails and Python and other computer languages. Even though it may not be part of my day to day job, it made me better in my day-to-day job. >> I want to thank Kellyn Gorman, for sharing her extensive experience about Data Warehouse Performance in Oracle solutions. You can find some brief supplemental slides, about this interview in the class website, and I hope this interview provides some motivation for embarking on a career as a data warehouse administrator or data warehouse analyst.
