Archive
A Brief History of Exadata Time by Juan Loaiza
The full article is available in the 106th issue of the NoCOUG Journal.
“we leveraged our 20–30 years of database experience to determine what would be the ideal platform for running the Oracle database. That’s the thinking that produced the Exadata platform as we know it today.”
“Exadata V1 used HP hardware. Exadata V2 used Sun hardware. Oracle has always worked very closely with Sun, but of course, with Sun becoming a part of Oracle, our relationship became much closer. And, we had a very clear directive from Larry Ellison that engineered systems were critical to Oracle’s overall strategy, and that was understood by both the database and hardware teams. We quickly got all the cooperation, all the features, all the fixes, and all the improvements that we wanted from Sun. When you are a single company and the direction is set very clearly by the leadership, then the hardware and the software integration can advance much faster. If you are two different companies, there are always different priorities in the different companies and this slows down progress.”
When you are a single company and the direction is set very clearly by the leadership, then the hardware and the software integration can advance much faster. If you are two different companies, there are always different priorities in the different companies and this slows down progress.
“When Exadata V1 first launched, it was a purely disk-based system. We added Flash in V2, but still the focus was primarily on disks. We had a disk focus, with Flash for acceleration. The big change with Exadata X3 is that we’ve increased the Flash memory capacity very significantly: we’ve quadrupled the amount of Flash memory.”
“Exadata X3 has been designed to work really well for all types of applications, including OLTP, warehousing, mixed workloads, and cloud”
“what is available today with Exadata is really just the beginning. We’re no longer focused on making the basic platform work , we’re now primarily focused on value add—things like improving consolidation, improving performance, improving compression—increasing the value of all the Exadata technologies even farther.”
The full article is available in the 106th issue of the NoCOUG Journal.
Oracle 12c Gives Fresh Life to the Relational Database Movement
FOR IMMEDIATE RELEASE
San Francisco (April 1, 2013) – In a dramatic move calculated to give fresh life to the moribund relational database movement, the latest version of Oracle Corporation’s flagship database has eliminated the famous “join penalty” by making it possible to store rows from multiple relational tables in the same database block. There are two flavors of the feature: …
What’s so sacred about relational anyway?
Dedicated to NoSQL and Big Data expert Gwen Shapira for forcing me to think.
Bring the past for judgment into the thousand-eyed present, and live ever in a new day—American philosopher Ralph Waldo Emerson in his essay on Self-Reliance.
First, a short but fun quiz. The answers are at the end of Page 4 or you can click on the information link provided after each question.
1. Before relational databases, there were network databases. Network databases were codified by the Conference on Data Systems Languages (CODASYL) in 1969. According to the website of a prominent software company: “CODASYL DBMS is a multiuser, CODASYL-compliant database management system for OpenVMS operating systems. CODASYL DBMS is designed for databases of all levels of complexity, ranging from simple hierarchies to sophisticated networks with multilevel relationships. CODASYL DBMS provides a reliable operating platform for application environments where stability, high availability, and throughput are essential.”
CODASYL DBMS was created by Digital Equipment Corporation (DEC) which went out of business in the 1990s. It continues to be supported by:
a) Oracle Corporation
b) IBM
c) Computer Associates
d) Software AG
Click here for the answer.
2. According to the IBM website:
- DBMS “X” manages a large percentage of the world’s corporate data;
- Over 95% of Fortune 1000 companies use DBMS “X;”
- DBMS “X” manages over 15 petabytes of production data;
- $2.5 trillion is transferred through DBMS “X” by one customer every day; and
- DBMS “X” can process 21,000 transactions per second
DBMS “X” is:
a) A pre-relational DBMS that helped put the first man on the moon
b) A relational DBMS
c) An object-oriented DBMS
d) A NoSQL DBMS
Click here for the answer.
3. Indian Railways is the world’s second-largest railway, with 6,853 stations; 63,028 kilometers of track; 37,840 passenger coaches; and 222,147 freight cars. Annually it carries some 4.83 billion passengers and 492 million tons of freight. Of the 11 million passengers who climb aboard one of 8,520 trains each day, about 550,000 have reserved accommodations. Their journeys can start in any part of India and end in any other part, with travel times as long as 48 hours and distances up to several thousand kilometers. The challenge was to provide a reservation system that can support such a huge scale of operations—regardless of whether it’s measured by kilometers, passenger numbers, routing complexity, or simply the sheer scale of India. (Source: HP website.)
What sort of DBMS was used by Indian Railways in 2000 to build a state-of-the-art passenger reservation system on HP OpenVMS AlphaServer systems in 2000?
a) No DBMS was used
b) A pre-relational DBMS
c) A relational DBMS
d) An object-oriented DBMS
e) A NoSQL DBMS
Click here for the answer. Search for the phrase “solution highlights.”
The surprising answers to the above questions should make us stop to think. As relational practitioners, we need to understand why the relational model is sacred. If we cannot explain why the relational model is sacred, we cannot hope to convert the unbelievers, can we?
So stick with me as I attempt to explain. First up is Simplicity and Naturalness.
Recent Comments