Oracle Open World 2013 started today and San Francisco is ready to rock. This year approx. 60.000 visitors are attending to this great event – this is a 20% increase compared with last year. Very impressive! From my point of view there is no other event where you can get so many useful information in just five days. Keynotes with the latest anouncements, many customer cases, sessions, networking, demos, hand-on labs and of course personal meetings with certain members of the product management team – a perfect combination to be prepared for our daily business and to be successful in our projects.
This year our schedule is full with the following topics: Big Data / Fast Data, Real Time Analytics, Event Processing, Governance with Oracles Enterprise Repository, large File Transfers, Business Intelligence, Web Center, Oracle ADF, Internet of Things (IoT), Mobile and of course SOA & BPM. This is going to be a very busy week 🙂
In his Welcome keynote Oracle CEO Larry Ellison was in a good mood because Oracle Team USA won two America’s Cup races today. Overall he came with the following announcements:
Oracle Database In-Memory Option
With Oracle’s new In-Memory option, he explained, queries are going to run at least 100-times faster, INSERT of rows run 3- to 4-times faster, UPDATES are 2-times faster and the JOIN of tables are at least 10-times faster. Furthermore reports run 20-times faster without predefined cubes. How is that possible? As Larry Ellison described transactions run faster in row format and analytics run faster in column format. Oracle 12c stores the same data in row AND column format simultaneously – it is a Dual Format In-Memory database. But how can you do more things in less time? Replace the indexes with the In-Memory columnar technology. Indexes work well for predictable access patterns. They are mostly used for analytic queries but the administrator needs to decide what to index and what not to index. Larry Ellison described that the column store replaces the analytic indexes which means less tuning and less administration.
Oracle In-Memory requires zero application changes. It requires no restrictions on SQL and will be turned on with the following three steps:
- Configure Memory Capacity (inmemory_size = XXX GB)
- Configure tables & partitions to be in memory (ALTER TABLE …)
- Drop analytic indexes
Afterwards Larry Ellison showed a live demonstration on a two socket server and a database table with 3 billion rows about searches that people have done on wikipedia. A traditional database with index achieved 2005 million row scans / second. Without index the result was much less – only 5 million row scans / second. Finally after enabling the In-Memory option the same test achieved an outstanding value – 7151 million rows / second.
Big Memory Machine M6-32
The second announcement was a new member of Oracle’s Engineered System family, the Big Memory Machine M6-32. It contains 32 Terabytes of DRAM memory and 32 SPARC M6 Chips. Larry Ellison said: „It’s the fastest machine in the world for databases stored in memory“. The live demonstration with this machine was again impressive – 341072 million row scans / second.
The third announcement was that the M6-32 is also available in SuperCluster form.
Oracle Database Backup, Logging, Recovery Appliance
The last announcement was the one with a name that is definitely distinctive, the „Oracle Database Backup, Logging, Recovery Appliance“. It is architected for the protection of critical business data. While the database is running the delta / changes are shipped to the appliance. The backup appliance is designed to backup databases (not files) and it doesn’t have to be in the same data center. The key characteristics of the product are:
- Real-time log shipping
- Fast restore to any time point
- Real time log & change deltas
Oracle is also offering the product as a public cloud service. It allows the backup of databases directly to the cloud while the data is encrypted at source.
The first day was already great and we are looking forward to take a lot of useful content with us.