Blog

Oracle’s database technology is getting cloud-friendly

Oracle’s database product line is getting increasingly cloud-friendly, even if it doesn’t get credit for it. While Oracle’s Public Cloud has yet to make its presence felt, the Oracle 12c and NoSQL databases are becoming better citizens on Amazon Web Services and Microsoft Azure. Perhaps more important, as more customers and ISV partners build SaaS apps, whether delivered from their own premises or public clouds, Oracle’s technology is making life increasingly easier for them.

On the admin side, features that have come out over the past year have dramatically reduced server sprawl and, therefore, cost of operations. On the developer side, better support for JSON as a data type makes it easier to build web and mobile experiences.

Oracle’s cloud technology cred
Three technologies distinguish Oracle’s new cloud-friendly cred. First, its flagship database 12c, released in the summer of 2013, has an option that makes creating multi-tenant applications much simpler. The feature makes it possible to consolidate many independent database applications or instances of the same application on a single database server.

Second, the flagship database 12c’s in-memory option, introduced this past summer, offers a major performance acceleration for analytic queries, making it possible to consolidate what used to be separate OLTP and OLAP databases.

Third, Oracle is embracing deeper support across both 12c and its NoSQL databases for JSON, the new data lingua franca for web and mobile user experiences.

Putting the three new technologies together makes it easier to build modern SaaS applications with Oracle database technology.

For those evaluating database technology to support cloud applications, let’s look at the first two of these technologies: Oracle 12c’s containers and its in-memory facilities.

Oracle 12c Containers
Containers and pluggable databases, introduced last year with 12c, simplify delivery of SaaS apps. A DBA can now run one database server that shares many customer databases, each isolated from the other, and each having its own security or availability options.

Before containers, SaaS pioneers like Salesforce.com had to code their applications to isolate one customer from another. Failing the use of that technique, VMWare-style server virtualization could be used, but then each customer database needed its own operating system, database software, and data, leaving little efficiency for a DBA.

But with containers and pluggable databases, SaaS apps now require no changes to application code, making for a pretty concrete demonstration of bringing customers’ existing investments forward.

Oracle 12c Database In-Memory
In-memory technology helps consolidation through performance. Traditionally, operational data had to be moved from OLTP systems to data marts via ETL. That made it somewhat stale and inflexible.

Oracle’s in-memory option makes queries orders of magnitude faster, not just by using memory, but by storing the relevant data there in columnar format, as many analytic databases do. This enables scanning of only the small subset of columns typically required by an analytic query.

With this performance acceleration, customers can simplify administration by consolidating data marts with OLTP systems.

What it all means
At the NoSQL Now conference in August, Oracle’s database chief, Andy Mendelsohn, put Oracle’s database cloud support into a big-picture perspective. Mendelsohn opined that any technology vendor’s long-term success depends on its ability to carry partners’ and customers’ investments forward from one technology era to the next.

With containers and pluggable databases, Oracle has enabled mere mortal enterprises and ISVs to build multi-tenant SaaS apps that previously took the heroics of a Salesforce.com R&D department. The in-memory technology makes it possible to combine analytics formerly done in a separate data mart with live OLTP data, simplifying the problem of database sprawl that is the bane of all well-designed SaaS applications. Thus Oracle is, at the very least, meeting its own database chief’s standard for long-term success.

Next up
In the next post, we’ll look at how Oracle is delivering another key piece of cloud-era technology, JSON support. Meanwhile, in terms of cloud-readiness, open source software and metered pricing count as much as technology issues do, if not more so. A third post will look at those challenges as well.