For me, the day started bright and early thanks to the jet lag (I think). The day was action-packed with sessions, and I presented my Oracle rman:Don’t Forget the Basics to an enthusiastic crowd in Moscone West. The room was close to full with some hanging out at the back. It had been a while since I had presented in a room so close to being full. There was a handful of questions and comments during the session, and a group of attendees approached me afterwards to follow up. I received immediate feedback from a handful of attendees and some comments back from some Pythian colleagues.
I then sat in on a panel where DBAs and some non-technical personnel discussed “What I Wish Someone Had Told Me Before I Became a DBA”. Naturally, the free-form of the session allowed us to deviate from the topic, but that is expected. There were a handful of technical questions but the conversations centered more on the soft skills and communications between DBAs and other company staff. One question was asked about licensing, and the panel was not prepared to give any advice, which was a good thing as these questions are best answered by Oracle directly.
I ventured with two colleagues to the opening general session at 5pm. The crowd was enormous, and I know that conference organizers immensely enjoy these streams of attendees filing into the session hall. The Fujitsu address focused on the cloud and how its scope is all-encompassing. BigData will change the game by gathering huge amounts of societal data that will revolutionize the way many organizations do business. Then came the main event.
Ellison made four announcements at the event of which the release of Exadata 3 was the most dramatic. Key features of release 3 include flash write and the sheer quantity of memory – over 22Tb comprised of a layer of memory components. The entire active database will be able to sit in memory, thereby increasing speed of access and accelerating application interaction. Database 12c was first discussed in 2011, and Oracle’s beta program has been moving along briskly ever since. Larry did not say much if anything about 12c that we did not already know, but he did discuss the pluggable database feature, explaining its advantage to the traditional database/application approach.
I believe the course that Oracle is following with Exadata and 12c unfolds as follows. As far back as the mid 1990’s, many looked at Oracle as just a database company. Through the balance of the 20th century, Oracle invested huge dollars in the Applications tier and raised their image as the E-Business vendor of choice in the world’s eyes. The acquisition of Peoplesoft in 2004 put the icing on the cake and finalized Oracle’s hold on the Applications business they had already corralled. I believe that through this time period, Oracle was thirsting for a hardware company acquisition that could fill a niche in their complete solution from server to front-end Applications. Whilst many remember the plethora of hardware companies that could have been candidates, Oracle did not find one that suited their needs until the Sun acquisition on 2009.
Oracle was then poised to move in the direction that has culminated in their mantra to reduce I/O operations and move software into a non-CPU silicon layer. They have “conquered” the Applications tier and now, armed with a complete hardware solution, I believe they are now returning to their roots and concentrating again in the database tier. Their efforts will focus primarily on their multi-tenant database offering coupled with exponential acceleration of database processing, since the most commonly-used data will be sitting in memory.
Pythian is well-positioned in the Exadata space and has been for some time. We are involved in the 12c beta program, and a number of Pythian techies are rolling up their sleeves and getting into the technology as we speak. A heightening of the database tier in the eyes of Oracle is good for providers who dabble as heavily in the database tier as we at Pythain do.