Your search did not match any results.
We suggest you try the following to help find what you're looking for:
By Chris Murphy
Juan Loaiza has been pounding away at Oracle Database for more than 27 years, helping refine the technology as it exists today. With that perspective, Loaiza thinks the careers of developers, architects, and IT implementers will depend on how well they prepare for the “big waves sweeping through the database industry.”
Juan Loaiza, Oracle senior vice president of database systems technology, says organizations need to embrace big change.
“It’s important for our database community to understand these changes and embrace them, because otherwise they‘re going to be swept away in the future,” says Loaiza, Oracle senior vice president of database systems technology. “The database technologies of the future are going to be very different than what we have today.”
Here’s one example: NoSQL databases are used by web giants to manage information for billions of users, and the web-scale approaches of those NoSQL databases are starting to appear as features of the enterprise mainstay Oracle Database. Take “sharding,” a type of database partitioning that separates very large databases into smaller parts called data shards. Web giants have used sharding for years to break an enormous NoSQL database into more manageable, smaller ones. But when that technique becomes part of Oracle Database (and it will), it inherits Oracle Database’s decades of sophisticated functionality, and thus becomes both more useful and more reliable.
So here are 10 transformational database technologies that Loaiza thinks developers, architects, and other IT pros need to track.
The idea’s pretty simple: Instead of users building a database platform from scratch, Oracle puts together the most efficient hardware and optimizes the software to run Oracle Database at its peak. For example, the system moves query processing directly into storage to make analytics run much faster.
Loaiza—and Oracle—are true believers in this approach, but the jury’s still out on whether this will sweep the industry the way some of the trends below will. After all, Oracle has beat this drum for nearly a decade and has thousands of customers in production, but no one else has followed its lead in a big way. “It sounds pretty straightforward, but no one else in the industry is doing that,” Loaiza says.
Maybe it’s because they can’t.
Learn how Oracle Database In-Memory powers the real-time enterprise.
There’s zero debate on this trend: Every major database vendor has added an in-memory product to its lineup. By using data structures and algorithms specialized for data in memory, a database can run analytics 10, 20, or even 100 times faster than a database tied to disk-based approaches. Such speed changes the questions people ask of their data, because analysts can iterate “what-if” queries knowing they’ll get answers back in seconds, rather than hours, as in the past.
Loaiza says Oracle is unique in offering “dual-format architecture”—a single database that uses the best approach depending on use: row analysis for OLTP, and in-memory columnar analysis for analytics. In-memory is part of the architecture of Oracle Database 12c, which means companies get it without changing existing applications running on Oracle Database, and they get features they expect, such as high availability and scalability. From on-disk formats to in-memory database, “the whole industry is going there,” Loaiza says.
“This is not just a product feature. It’s really a computer science advance,” Loaiza says.
The database technologies of the future are going to be very different than what we have today.
— Juan Loaiza, Oracle Senior Vice President of Database Systems Technology
Oracle calls the approach “software in silicon”—embedding algorithms directly onto microprocessors. The idea is that processors can’t just keep adding ever-more cores and threads, so speed and performance will come from putting algorithms to accelerate core tasks such as encryption and compression directly onto the chip.
Loaiza cites three aspects of this advance. First, SQL in silicon accelerates database in-memory performance. Second, capacity in silicon can get a lot more data into memory using on-processor tactics such as real-time decompression. Third, encryption support in silicon can help improve security as workloads move to in-memory, because without such protection in-memory could be less secure than data on disk.
“It’s an advance for the whole industry on how database processing works, and how chips are designed,” Loaiza says. “You’ll see more of these products appear as other vendors start copying this technology.”
Oracle Big Data Appliance delivers comprehensive big data capabilities at a low total cost of ownership.
Companies generate big data when pursuing strategies such as the Internet of Things, or tracking web clickstream data for customer trends. But the deep insights come from blending that new big data with data companies already have.
Database pros will increasingly face the question: “How do we integrate big data with existing operational data?” Loaiza says.
He points to two Oracle technologies as examples that can help. One, Oracle Big Data Appliance is a system that runs Apache Hadoop and Spark, letting IT teams set up a big data system much faster and at much lower cost than building their own. Two, Oracle Big Data SQL lets analysts run massively parallel, full Oracle SQL queries across relational, Hadoop, and NoSQL data. Oracle SQL is more sophisticated than what is generally run on these big data sets, Loaiza says. The goal is to “make it simple and efficient to integrate all this big data together.”
As we said at the start, big web companies with hundreds of millions and even billions of users depend on sharded databases; rather than have one massive database to manage a billion users, they shard the database so it’s broken into more manageable elements, but the company still can query all of the shards.
Sharding technology was developed many years ago and is easy to deploy using NoSQL. “It’s one of the reasons many of these web companies like NoSQL,” Loaiza says.
Oracle plans to offer “native sharding” inside Oracle Database, Loaiza says, since in the past users had to take manual steps to simulate sharding in Oracle Database. Native sharding will enable IT to manage the shards, while getting the usual benefits of Oracle Database, such as high availability and security features.
When disk-based database backups came on the scene about a decade ago, they brought two big advances over tape: more accessibility to data and deduplication that reduced the volume of data. “And that’s where it stopped in the last decade,” Loaiza says.
Oracle’s Zero Data Loss Recovery Appliance offers ultimate protection for always-on business.
Increasingly, the expectation will be that data backups happen constantly, in real time, so that data never gets lost. With batch-oriented, disk-based backups, data can be lost between backups. That doesn’t cut it in today’s digital business models. “Nobody wants to lose data.” Loaiza says. “It’s pretty much impossible to go to financial users and say ‘I’ve lost your data.’ No one wants to lose their seat on an airplane, or their day’s worth of shopping.”
Oracle meets this need with an approach similar to its engineered systems, using specialized algorithms on custom hardware called Zero Data Loss Recovery Appliance. In addition to preventing data loss, the appliance has minimum impact on production servers because it backs up only the changes to the database, rather than copying a whole database. It has database-aware recovery that validates data as it backs up, and it offers cloud-scale protection, so a single appliance can back up a whole data center.
JSON is a way to represent data that’s very popular with developers, Loaiza says. It largely replaces XML as the format for data with complex structures, such as web pages or user profiles. NoSQL database projects embraced JSON as more developers used it to build their apps.
Now developers can store JSON data natively as columns in Oracle Database. By using enhanced SQL to access JSON data more easily, database pros can do any relational database task with that data. “If you look at NoSQL databases, they kind of had two technologies that were really interesting—one was sharding, the other was JSON,” Loaiza says. “We’ve now introduced those into relational databases.”
Virtualization has been widely used for more than a decade, and has given way to lighter-weight, container-based virtualization. With Oracle Database multitenancy, container technology goes directly in the relational database, instead of virtualizing at the operating system level. With container databases, a single container manages many “pluggable” databases. Each pluggable database thinks it has its own private resources, but it’s really sharing them as part of one container, which makes managing and scaling easier.
Container databases are useful for consolidating lots of databases in an enterprise to reduce costs, Loaiza says. Also, they help reduce software administration costs, since there’s less software to manage compared with OS-based virtualization. OS-based virtualization reduces the hardware need by increasing the utilization rate of servers, but increases the software administration burden, because each virtual machine has a copy of all the software.
Oracle Database Exadata Cloud Service offers the performance of Oracle Database with all the advantages of the cloud.
Developers or testing teams want a database environment up and running as soon as they have work to do, without waiting weeks for IT to order and set it up. If an application is a huge hit, companies want to scale up their database quickly (and if it flops, they want to shut it down and move on to the next idea). They want to use a database with the latest performance and security features. And companies want to pay their technology teams to craft new applications and run new data analysis, rather than fine-tuning servers in a company data center. All those reasons and more will drive more work to cloud-based databases.
With systems such as Oracle Database Exadata Cloud Service, database pros can get cloud-based databases without compromises. The service provides the performance of Oracle Database running on Oracle Exadata, with all the advantages of the cloud—fast provisioning, Oracle experts running it, continual updates, elasticity, and subscription pricing. “It’s moving to the cloud but taking with you all the advantages that we’ve built during the past 40 years instead of going to some primitive database in the cloud,” Loaiza says.
Even with the rise of the cloud, database pros will be running on-premises systems for the next decade or more. Sometimes regulatory requirements may force a company to keep certain data inside its own data center, or within its country of origin. Sometimes it's a custom-built application that’s doing its job, so IT leaves it alone and surrounds it with modern apps.
Oracle is helping to address this reality by creating systems, called Oracle Cloud Machines, that can run a database exactly as if it were in Oracle Cloud, but that reside on premises in a company’s own data center. IT pays for Oracle Cloud Machine as a subscription, Oracle pros manage it remotely, and capacity can scale up and down depending on need; the only difference is that the device resides behind your firewall.
Safe Harbor Disclaimer
The preceding is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle's products remains at the sole discretion of Oracle Corporation.