How Modern Active/Active Data Infrastructures Unlock Digital Transformation

Print Friendly, PDF & Email

In this special guest feature, Srini Srinivasan, Founder and Chief Product Officer at Aerospike, begins by discussing the limitations of an ACTIVE-PASSIVE data infrastructure approach. He’ll then explain how an ACTIVE-ACTIVE approach works, along with an example, and the technical and business transformation benefits it delivers. Srini has two decades of experience designing, developing, and operating high-scale infrastructures. He also has more than 30 patents in database, web, mobile, and distributed systems technologies. He co-founded Aerospike to solve the scaling problems he experienced with internet and mobile systems while he was senior director of engineering at Yahoo.

In the current environment, businesses continue to drive digital programs that reduce the need for physical presence or human intervention and are now adding a whole new layer of vital business continuity that bolsters systems to deal with unprecedented human and technical challenges.

One of the places ripe for digital transformation in the “new normal” are many traditional, straightforward business processes that are stuck in older systems with little automation and too much human intervention to ensure the integrity of a transaction. These processes create unnecessary business delays and overhead—and fail to meet the “digital expectations” of customers.

Many of these business transactions aren’t going to be digitally “disrupted” out of existence. They are fundamental processes woven into the business fabric. But with the modernization of the underlying data infrastructure, these processes will leapfrog into the expectations of the 21st century—delivering both efficiencies and savings while increasing the continuity and integrity of the transaction.

Let’s look at one industry in particular: banking. According to Gartner, the top business priority for one-third of all banking CIOs is digital transformation. With the rise of PayPal, Venmo and other modern instant, digital payments, consumers expect instant transactions. Yet many core back-end bank business processes—account changes, inter-bank funds transfers, etc.—remain slow, manual and held back by legacy data infrastructure.

There are significant outdated technology components and business processes that slow these transactions down. First, time delays help prevent fraud and errors and allow for human intervention to physically confirm and approve transactions. But today, this could be easily replaced with modern approaches used by fintech companies like PayPal.

Second, the underlying data infrastructure, often using an active-passive (or even active/active eventual consistency approach) wasn’t designed to meet the real-time expectations of digital-savvy consumers and applications. In active/passive (and active/active with eventual consistency) systems, the database is typically replicated across two or more geographically distributed sites but only one of them—the primary—takes application requests. All changes happening in the primary will be sent to the other site—the secondary—in an asynchronous manner. (The secondary site is sometimes described as a “hot standby.”) This asynchronous replication can cause conflicting writes that require all kinds of manual conflict detection and remediation.

How Active/Active Data Infrastructure Works

An active/active database implementation is an underlying data architecture that makes transformation possible, affordable and resilient. There are many intricate configuration possibilities for an active/active database, but let’s look at the basics of how it works.

An active/active system is defined as a network of independent processing nodes, each having access to a common replicated database such that all nodes can participate in a common application. Today, there exist active/active systems that can ensure synchronous replication across geographically distributed sites that are thousands of miles apart while at the same time preserving strong consistency (with no data loss) and 100% availability during site failures without operator intervention.

Even with a potential increase in write latency (between 2 and 100 milliseconds), an active/active system can still support data-intensive applications with sub-millisecond read SLAs while allowing every change to be instantly visible executing the transaction across multiple sites. Note that such a system can also provide lower cost and higher scale of operation without sacrificing any of the correctness and safety that was available in previous semi-automated processes and systems. Configured in a rack-aware multi-site cluster, an active/active configuration that preserves strong consistency can guarantee that all writes will be replicated across sites without data loss. Such a system can survive the loss of an entire site (rack) with no loss of data and continue to operate.

For a practical application, let’s revisit that money transfer example I mentioned. An active/active strongly consistent data infrastructure eliminates a lot of process friction and redundant systems. Customers can simply use their own bank accounts for transferring funds to anyone else who has a bank account without needing special additional electronic wallets, etc. This kind of transformation is essential for financial services to survive and thrive in the new era, and it is emerging in systems like the Zelle network in Europe (vs. older money transfers, which can take hours or days).

New Life to Old, But Essential, Processes

Modernizing existing active/active systems with eventual consistency to use an active/active data infrastructure with strong consistency has many technical benefits. The most obvious is a return on investment by dramatically reducing the time to market for new features and services for existing clients, attracting new clients as well eliminating the technical debt of needing to continuously enhance and maintain legacy data systems. According to Forrester Research, more than 70% of technology budgets are consumed by maintaining older applications that don’t make sense for today’s digital economy. Many active/active data infrastructures with strong consistency can survive site failures without operator intervention and also capitalize on commodity hardware or the cloud—further reducing the up-front costs of switching.

Beyond the technology, the real gain lies in business transformation. First, adding automation and removing delays brings critical, foundational business processes into the speed and delivery expectations of modern life. As digital-native companies emerge, this allows older, foundational companies to compete—delivering the stability and experience of a time-tested institution with the modern business processes that allow it to better compete—and fend off—the disruptors who have painted targets on their backs.

Sign up for the free insideBIGDATA newsletter.

Speak Your Mind