Blog

The Power of Agile Data

Apple's singular focus on consumer usability has had a widespread impact on the future of enterprise applications and systems.

Apple’s singular focus on consumer usability has had a widespread impact on the future of enterprise applications and systems. The proliferation of easy to use apps on mobile phones, tablets, and other devices has reset consumer expectations worldwide. The lesson we’ve learned from great products is that our time matters. We want our apps and data where we want, when we want, in the form factor we want.

Enterprise apps—all the longstanding apps that drive businesses worldwide (ERP, CRM, HCM, SFA, etc.)—are notoriously difficult to use. Consumerization and even basic maintenance of enterprise apps can be painful and expensive. Every year IT organizations employ armies of application engineers and consultants on multi-million dollar projects that invariably take too long and cost too much.

Data is the lifeblood of applications, and compound data growth is progressively clogging the arteries of today’s application infrastructure. A senior executive at a major bank remarked: “There are only 3 things certain in life: death, taxes, and data growth.”

During an application’s lifecycle, developers and QA engineers need hardware and software environments loaded with the right data to properly test and re-test changes. Most enterprise apps store data in relational databases like Oracle and Microsoft SQL Server, which require careful management and handling. To load an environment with the right data, project teams rely on busy, overburdened IT operations teams. Delivering a single app environment can take hours of effort and days, weeks, or months of elapsed project time. Once a request gets approved, systems admins, storage admins, backup admins, and DBAs have to work in concert to prepare systems, move data, and verify correctness to present an environment to an app team.

Large businesses might have hundreds to thousands of application projects per year. Each project can require multiple development, test, QA, integration, system validation, user acceptance, performance, training, and pilot environments—all of which require the right datasets loaded at the right times. Moving datasets from production systems into project environments requires forcing balloons of data through multiple choke points. Data often needs to be read from production systems, into backup systems, and then across networks, before being written into development systems—taxing all systems along the way. The larger the dataset, the larger the tax, and the more time it takes to navigate the choke points.

Shepherding data through these choke points burdens IT teams, reducing their value and effectiveness to businesses. At StubHub, a division of eBay, they compare data copy work to breaking rocks: dispiriting and highly repetitive. When IT teams and systems get busy or run out of capacity, development requests get delayed, and application projects fail to meet deadlines.

This domino effect is why businesses fail to innovate at the rate they expect.

Once you virtualize data, everything moves faster. Delphix virtualizes the data in relational databases, providing an Agile Data Platform that breaks through choke points and enables businesses to roll out application projects in less time, while simplifying IT operations and reducing hardware impact and costs. Delphix introduces ten “super powers” for IT organizations to address the data management problem:

  • Plastic Man: Delphix turns database hardware into big, elastic workspaces. An average installation of our software allows customers to run 20 databases in the space of 1—on the same storage and database servers. This allows IT organizations to provide app teams with the environments they need, instead of what they can afford.
  • Virtual Speed: by virtualizing databases, Delphix can deliver, migrate, synchronize, and move large data environments, even 50TB databases, in minutes. At physical speeds, loading a 50TB database could take weeks or months—or could be completely impractical.
  • Web Spinners: with the ability to pull data from 20+ databases (including Oracle, Oracle Exadata, Oracle RAC, Microsoft SQL Server, and Microsoft SQL clusters), a single Delphix installation can efficiently manage data transfer, synchronization, and delivery within and across data centers—in near real time and with full transactional consistency. Delphix can also pull data from Delphix to Delphix for high availability or remote site access.
  • Fresh Lantern: project environments often make do with stale data, due to the impact and time to move fresh data from production systems. Unfortunately, stale test environments can lead to downtime and high severity support cases. Delphix can swap out stale data with fresh data in seconds—with no impact to production systems—so tests can always use realistic, current datasets.
  • Time Travel: apps evolve and data is dynamic and ever changing. Mistakes happen. Errors percolate through the stack. Delphix can quickly dial data back in time to recover lost, corrupt, or deleted data.
  • Wonder Twins: determining root cause can be hard when you can’t afford to try things on production. Delphix stamps exact replicas of production sources in minutes, providing immediate firefighting or break/fix environments, where performance probes can be inserted, tables replaced, or indexes can be rebuilt to quickly implement fast fixes. Wonder twins can also be used to spin up parallel development or integration environments.
  • Deflector Shields: today’s production systems face a barrage of workloads, which can swamp available resources: business intelligence extracts, backup runs, operational reports, etc. By providing a parallel virtual environment, Delphix can deflect a multitude of workloads to help protect the enterprise starship.
  • Tide Control: by managing, processing, and storing only changes to data (instead of moving full datasets over and over again), Delphix reduces the rising tide of data load on production systems, networks, and development environments, ensuring more resources can be used for real, productive work. Delphix shrinks systemic data loads and storage by as much as 90% on an ongoing basis.
  • Invisibility Cloak: virtual databases look and behave like physical databases, connecting to existing database and application servers—so developers and QA engineers won’t see a difference—until they use a “super power.” In addition, Delphix can slipstream into programs in a day or two, saving time and costs for projects already in flight.
  • Rocket Boosters: by providing self-service data control for developers and QA engineers, Delphix adds a 20-80% productivity boost to application projects in customer deployments—a significant competitive advantage in evolving and modernizing apps.

These “super powers” help at various stages of the app lifecycle, at different levels within an IT organization, and across different systems in a datacenter. Together, they enable project agility and return time and budget to IT teams.

Prospects often say our software sounds too good to be true. After testing the product, they tell us it works as advertised. But when they see the impact on app projects, they say it’s like magic. KLA-Tencor has seen their teams accomplish five times the number of SAP projects on our Agile Data Platform, with the same number of people and the same hardware systems. StubHub estimates we’ve helped their release schedules by over 20% by quickly delivering virtual data environments to developers. We’ve helped Facebook reduce their monthly financial close from 21 to 2 days. By consumerizing elements of data and database management, we deliver an Apple-like experience for some of the most complex moving parts in an IT datacenter, which helps our customers consumerize their applications at business speed.

Virtual infrastructure is clearly superior to physical systems for app projects. With over 50 Fortune 500 customers in production across all industries, it’s not too good to be true. And it’s not magic.

It’s just great software.