Building a modern app with Oracle's Converged Database

Synergized, Secure, Scalable: Oracle’s Converged Database helps Development Teams Get Application-Smart

Business today runs on high-performance applications. But in order to innovate and differentiate in an always-on world, application development needs to be swift, efficient, and responsive to complex data. Oracle’s converged database delivers a unified solution loaded with the power to move business forward at speed.

When it comes to doing business in the 21st century, a company’s success depends on its applications. Not only do applications differentiate a company, their strength, speed, responsiveness, reliability, and performance can make or break the customer experience.

More Info: 1Z0-068: Oracle Database 12c - RAC and Grid Infrastructure Administration

As a result, modern applications are more prevalent than ever. To customers, they represent brand experience. To employees, they’re the key to efficient operations and empowerment. To executives, apps drive business growth and competitiveness in the marketplace — factors measured externally by visibility as well as customer acquisition, retention, and loyalty. Internally, apps drive analytics and insight into all operations, which in turn facilitate better corporate decision-making.

In light of all this, drop a layer deeper (across any enterprise) and an important truth becomes crystal clear. Databases are the backbone of every application.

Databases not only contain rafts of crucial data, they keep it safe, secure, and available 24/7. But this data is diverse. Gone are the days of capturing simple transactions and generating monthly reports. Now, customers are their own transaction agents. They travel and transact from mobile phones, laptops, and other devices, any time of day or night. They nurture relationships to other people, places, and things. Their needs and behaviors drive the always-on world.

When it comes to capturing all this data, this trend of transactions pushing out to customers creates exponential compounding of data. But let that data remain latent or untapped, and not only will a company’s applications suffer, they’ll miss the real-time cues that are most valuable to any organization.

Many companies are making their way through this complex landscape with piecemeal solutions — technologies and architectures reached for as needed over time; some of them proprietary, some home-grown, others open-source. Regardless, the resulting patchwork can’t handle the data load. Data fragments. Security becomes inconsistent. 

Administration is ad hoc. The systems become brittle as data compounds, which quickly hampers a company’s agility, as well as its ability to innovate. 

This exacts a high price. A literal and metaphoric cost that can be radically mitigated with Oracle’s converged database. 

Let’s unpack some of the key ways Oracle’s synergized solution delivers the power to keep you smart. Application Smart. On our way there, we’ll begin with a fairly common view of the challenges many companies face, and how IT organizations typically respond — and can course-correct.

The modern application experience is continuously evolving. This change in business needs demands fluid adaptation when it comes to the back-end. Tech stack, application architecture, dev/test, release processes, support: IT’s responsiveness, and its ability to efficiently handle various workflows, becomes the fulcrum of the enterprise — the leverage-point that requires deliberate strategies and savvy choices to deliver the necessary simplification, unification, and automation of architectures, technologies, and processes.

Without these deliberate choices, applications that combine analytics, geo-location, network relationships, machine learning, and transaction activity are often fueled by separate, single-purpose databases and their resulting technology stacks. That may seem viable on the surface, at first-blush. But within this solution construct, the more data moves, the more it degrades an application’s performance. Dev-time increases. Release-schedules lag. Apps become less competitive. And each one requires IT specialists trained in specific stacks.

The overall hit to efficiency can be devastating.

To demonstrate this with specifics, let’s take a quick look at the data, data types, and demands of a truly modern application.

For a long time, the nature of data, and data types, has been row/columns representing relational data. As IT has advanced, data such as key-value, spatial, graph, blockchain, document, time series, and IoT have subsequently emerged. Each of these datasets must co-exist with relational data. When seen in its totality, every modern, data-driven application has inherent data-type complexity. As a result, workloads like transactional, analytics, machine learning, and IoT require different database algorithms to solve for unique demands. This in turn has given rise to single-purpose database solutions.

But while single-purpose databases might appear to be best-of-breed in isolation, they quickly become worst-of-weakness when cobbled and patched together.

As an example, let’s take a look at the retail application of “eShop.biz” — a fictitious company with a common challenge who would come to us after having chosen independent, single-purpose databases as their architecture for online retail.

As you can see in the graphic below, all “eShop.biz” retail services, which included product and customer catalogs; payment, geo-location, cache, and recommend engine services; plus payment, finance, inventory, and delivery services — benefit from a single-purpose database.

A visual example of eShop.biz’s various single-purpose databases and some of the tech stacks each database holds. 

Committed to meeting the ever-changing demands of B2C online retail, eShop.biz opted to run their prime retail application on the compute layer, hosting the processing of all complex application logic for multiple services within auto-scaling compute environments. These services needed to communicate with the data layer, which, in this case, was built from linked, single-purpose databases — each of which solves a unique and specific problem (such as provisioning a spatial database, or provisioning a document database for a specific catalog or spatial requirement). When viewed one by one, these discreet databases appeared to offer the easiest, most natural, specific solutions for the dev team tasked with adopting the APIs.

But quickly, the reality proved otherwise.

For one, the product-catalog service needed to propagate data from its own unique database to three other databases — in this case, the cache, spatial, and warehouse databases. This meant eShop.biz’s dev team needed to write additional code for the job, which in turn meant any developer assigned to the job needed to deeply understand cache and spatial databases, in order to both make sure transaction communications were frictionless across proprietary APIs, and that response times were fast enough. 

Meanwhile, as the above change was being implemented within the production system, an additional, multi-pronged change needed to be considered. This involved the addition of a geo-spatial service that would empower the recommendation engine to suggest relevant items to the multi-faceted production system. All of this needed to work seamlessly with the warehouse databases, as well as the customer catalog for regional performance, while adding machine-learning capabilities to the warehouse databases in order to optimize eShop.biz’s product-delivery engines.

Complexity grew logarithmically.

In this case, each single-purpose database had its own DDL and DML. As a result, adding the above new features required merging data and application calls, which in turn involved additional code, written in the proprietary APIs of each discreet database. Simultaneously, IT needed to make absolutely sure they weren’t impacting any earlier change with their required additional code.

Very soon, IT realized every change had a base cost that was a function of compute, network, storage, integration, coding, management, and more. Most important, in addition to base-cost per change, the dev team also needed to address the security profiles related to each change, making sure security was in sync with the prime application logic — all while eShop.biz continued to scale.

Needless to say, eShop.biz’s IT organization had a tough time meeting the demands of its retail application. Increasing complexity; proprietary APIs, languages, and transaction models; fragmented data; inconsistent management; fractured security; limited views into various datasets (as opposed to a holistic view); apps that became locked to a discreet database; slow development cycles … the problems multiplied. eShop.biz’s prime application was neither fault-tolerant, nor able to shard data. Far worse, scalability became markedly difficult if not impossible.

eShop.biz knew they needed a multi-pronged solution with the power to address the myriad challenges they suddenly faced.

At the core of the problem, eShop.biz needed to make data-driven application development and deployment simple, streamlined, and swift. They needed faster development cycles (agile development with programming languages), along with easier management, data consistency, data security (the kind that would give them granular security controls, and security profiles at the database level), plus native high-availability, scalability, and cross-site consistency.

Equally important, eShop.biz needed to host multi-model data types, such as JSON, spatial, graph, timeseries, and ledger data. They also needed to run different workloads together, while hosting multiple tenants and in-memory area/cache streams without any code changes. Above all, they needed to streamline to one language, versus many disparate, specialized languages.

Oracle was uniquely positioned and ready to help, as the Oracle’s converged database addressed eShop.biz’s many challenges.

To use a perfect analogy, at the highest level, a converged database is akin to a smart phone — a single device that houses what used to entail multiple separate tools and functions. What once required an analog calendar, desk phone, audio player, alarm clock, pager, camera, etc. became highly simplified and collapsed into a single device — a platform that delivered everything within a unified operating system.

The smart-phone — a single device that seamlessly runs multiple apps, tools, and functions — is the perfect analogy for the workload aggregation offered by Oracle converged database.

As the following graphic illustrates, a converged database operates in a similar way. It supports the seamless mixing of workloads, data types, and algorithms. It enables SQL, as well as transactions across any data type; it provides unified security and management across all data; and it prevents data fragmentation and copy contagion. 

Oracle’s converged databases delivers a single, unified system that supports IT organizations in achieving peak efficiency and cost savings.

Beyond that, by enabling powerful synergies across features, Oracle’s converged database removes initial and recurring integrations costs, while being simpler, more reliable, and far more economical overall.

eShop.biz soon realized that the multi-model flexibility of Oracle’s converged database encompassed relational, JSON, XML, graph, spatial, and OLAP. It also facilitated multiple workloads (including transactions, analytics, in-memory, IoT, steaming, and blockchain), while offering multitenant functionality with container, isolation, orchestration, and scale-out with transactional consistency.

Oracle’s converged database delivers synergy, efficiency, security, and performance within a single, unified system.

For eShop.biz’s developers, Oracle’s converged database provided one database for all development, while allowing APIs and queries to be written across data types. Meanwhile, dev ops was able to provision dev and test environments for support and UAT much more quickly. Plus, fewer tech-stacks meant dev teams could work efficiently without specialized knowledge. Simultaneously, system admins and DBAs enjoyed consistent, high availability and disaster recovery across all data types and data processing. Even better, data was easily and consistently governed, secured, redacted, and masked for testing and support. 

When it came to infrastructure management, eShop.biz gained a platform that could run as easily in an on-premises data center as it could in the cloud, while providing the ease of migration as budgets and business dictated.

Diving deeper into detail, once deployed, Oracle’s converged database empowered eShop.biz with predictable performance, even as it scaled, with real application clusters (RAC), sharding, and real-time cache analytics in-memory requiring no code changes. When it came to data availability, eShop.biz was also able to build solutions with Oracle’s Active Data Guard, while implementing maximum availability architecture (MAA) with RAC, GG, and DG for the converged database.

Security was also strengthened, with controls like Encryption Everywhere, VPD, redaction, masking, Oracle Label Security, Oracle Audit Vault, and Oracle Database Firewall that secured databases with granular profiles while providing an easy way to implement them. Furthermore, with solutions like Oracle GoldenGate and Oracle Database Advanced Queuing (AQ), the database could handle feeds from distributed systems. All while eShop.biz developers were able to make use of REST services, JSON, SML, Python, Node, SQL, PL/SQL, and machine learning — all of which are native to Oracle’s converged database.

But perhaps the most powerful benefits delivered by the synergy, efficiency, security, and overall performance of Oracle’s converged database resulted in the ability of eShop.biz to accelerate innovation. No longer hampered by compounding complexity, eShop.biz’s IT organization was able to measurably simplify IT operations. This allowed them to quickly respond to eShop’s business needs with smart solutions that reduced business risk in the process.

As a final example of the synergy, efficiency, and multiple benefits of the Oracle converged database, let’s look at how eShop.biz made an architectural change at the database layer.

eShop.biz streamlined and optimized its architecture at the database layer by adopting the Oracle converged database.

Setting the stage, Oracle’s converged database natively hosts unique data types, including JSON, XML, spatial, graph, timeseries, blockchain, tables, and more. These are run from a single transaction engine, thereby eliminating logarithmic complexity at scale. 

Oracle’s converged database also seamlessly handles different transaction modes, such as analytics, transaction, in-memory, IoT, and blockchain. And its multi-tenant deployment provides a container platform, along with the necessary isolation among the various services. 

Taken together, the above benefits (plus more itemized below) introduced efficiencies and cost savings within eShop.biz’s IT organization, while enabling innovation that improved eShop.biz’s offering and performance.

Because Oracle’s converged database is a single, unified system, that treats all data types with a single DML (Oracle SQL), eShop.biz’s dev team was able to achieve transaction consolidation, as actions that had formerly been performed across single-purpose databases — each of which required additional code — now took place in a single architecture. This consolidation eliminated additional cost, effort, maintenance, and time-consuming integration.

Additionally, by leveraging Pluggable Databases in achieving microservice data isolation with multitenant, eShop.biz simplified its microservice architecture by improving simultaneous isolation among containers, even at scale. This enabled eShop’s prime retail app to quickly respond to business demand. 

Furthermore, because Oracle’s converged database automatically generates REST APIs on top of SQL, eShop.biz developers were able to code less while accessing data like any other service. Not only did this standardize their API workflows, it allowed them to generate new services across warehouses, which helped the business optimize its logistics.

As any dev knows, moving data takes time. eShop.biz developers were able to minimize that time — especially when it came to analytics — with in-place machine learning, since Oracle’s converged database allows machine-learning training to occur with the data in place (by bringing the algorithms to the data ). With 30 algorithms (and counting), this led to additional cost savings for eShop IT.

In-place graph analytics offered a similar advantage to eShop.biz developers, as they found it simpler to discover influencers and dependencies, while better understanding various insights revealed through relationships.

Summing it up, eShop.biz was able to radically simplify their architecture, and thereby, their IT operations, with the many benefits delivered through Oracle’s converged database.

Best of all, the ongoing advantages delivered by Oracle’s unified system have made it possible for eShop.biz developers to efficiently address application requirements while achieving greater operational, security, availability, scalability, and deployment efficiencies. This meant they were able to activate innovation that has optimized and transformed eShop’s business.

Source: oracle.com