What’s driving businesses’ thirst for real-time access to data?
Business stakeholders are demanding real-time data faster to keep up with customers, competitors, and partners. Enterprises often cite support for real-time as a top data management and analytics requirement, especially as users increasingly adopt modern business apps and streaming insights solutions. We find that traditional data platforms often slow down data access for critical business applications and insights, that’s impacting business growth and competitiveness. Key reasons for real-time access to data include:
- Competitive pressure that’s driving the need to improve data on-demand. Competitive pressure is forcing organizations to act more quickly to business needs more than ever before, especially where there are too many silos and inconsistent data being leveraged. Many business users are complaining about data being stale, and it’s slowing down their decision-making ability impacting growth and innovation.
- Real-time applications are requiring modern architecture. This is the most common usage scenario among commercial enterprises. For example, a telco handling an incoming sales call from an existing customer wants to put all of the relevant information at the salesperson’s disposal, including a detailed history of the most recent interactions with the customer and the sales offers that are most relevant to the specific customer calling in. Real time access to customer data helps deliver better customer experience, minimize churn and helps drive consistent and predictable business growth.
- Slow data that’s impacting operational decision making. When organizations need to determine why something is happening or the best course of action, they can’t wait for hours or days, they need information on-demand. Business stakeholders need to be able to aggregate and prepare data sets quickly without latency and with minimal effort.
- Lack of data sharing among employees and LOBs that’s impacting the organization. Insights and intelligence often sit with select analysts and managers in organizations, but employees need those insights to be more effective and efficient in their jobs. Data systems need to quickly get insights into services and systems that automate processes and guide employee actions.
- Mobile device usage that’s pushing for the need for real-time information on-demand. Imaging getting messages, emails and alerts on your mobile devices that come in minutes or hours instead of seconds. Mobile devices have changed the way we leverage information. No one wants yesterday’s data tomorrow; users want data in real-time, including business data and analytics. Mobile devices are transforming the way we now deal with business data, requiring organizations to invest in real-time data platforms.
What applications and insights are best suited for NoSQL?
Enterprises like NoSQL’s ability to scale out using low-cost servers and a flexible, schemaless model that can store, process, and access any type of business data. NoSQL platforms give organizations greater control over data storage and processing, along with a configuration that accelerates application deployments. While many organizations are complementing their relational databases with NoSQL, some have started to replace them to support improved performance, scale, and lower their database costs. The top workloads suitable for NoSQL include:
- Applications that need horizontal scale-out across many servers. NoSQL offers a scale-out platform to scale applications that need to support millions of users for transactional and operational workloads. Today, enterprises are building social-media apps like Twitter, Facebook LinkedIn, eCommerce apps like amazon, Airbnb and Uber, and other mobile and web applications by leveraging NoSQL to support web scale workloads.
- Insights that need to be in real-time. Accessing different pieces of information from multiple sources and performing business analysis and real-time insights, such as fraud detection, stock market analysis, security threats, or risk analysis, can be very complex, especially when dealing with large amounts of data in real-time. NoSQL offers new possibilities to leverage commodity servers and in-memory computing resources to support faster access of active business data.
- Embedded database applications for ISVs and VARs that need flexibility. NoSQL offers independent software vendors (ISVs) and value-added resellers (VARs) the ability to embed a low-cost NoSQL engine within their applications and solutions quickly. End-user companies can also take advantage of NoSQL as an embedded database to support various types of business applications.
- Mobile applications that need a variety of data quickly. NoSQL offers the ability to support real-time mobile platforms that need all kinds of information, such as for delivering a Customer 360 use case. Developers can store metadata and data in a schema-less data model allowing for the flexibility to introduce new features and functions to the applications without requiring database changes. This incremental and flexible approach allows the product to evolve based on customer feedback and market trends.
How can in-memory persistent memory help with a new generation of applications and insights?
Persistent memory is critical for all organizations, especially when supporting new and emerging business applications and insights that demand low-latency access to large amounts of data. Unlike DRAM, persistent memory offers a lower-cost alternative enabling organizations to store hundreds of Terabytes on a single server or petabytes in a scale-out architecture. With persistent memory, organizations can move their slower-moving data platforms to in-memory-based platforms quickly, with minimal effort. Large, scale-out, petabyte, in-memory data platforms are likely to emerge in the coming years to support all kinds of use cases, including translytical, customer 360, IoT and fraud detection.
How do you go about putting together a real-time data strategy?
A real-time data strategy requires people, processes and technology working together for various business use cases to succeed. The strategy should focus on key transactional and operational applications and insights initially that need low-latency access before expanding to others. A real-time data strategy should focus on:
- Creating a real-time data team to ensure success. A real-time data strategy is doomed to fail if you don’t involve the right personas from start, especially if there are too many applications and insights that need data in real-time. Your team should include enterprise architects, data architects, developers, data engineers, data security professionals, data stewards, business analysts, and data scientists.
- Start with a business use case in mind. Look for low hanging fruit; applications that are struggling with performance issues whether transactional or operational. Also, look for new applications that need low-latency access to critical business data to support use cases such as Customer 360, IoT analytics or fraud detection.
- Focus on operational and insights bottlenecks. Migrate from slow-moving batch and semi-batch data processing for operational and analytical workloads, moving towards real-time everything.
- Leverage in-memory extensively to support real-time initiatives. To succeed in a real-time data initiative, leverage DRAM, persistent memory, and Flash/SSD extensively. In-memory data platforms can dramatically improve performance of existing or new databases that support large mission-critical applications, such as ERP, CRM, supply chain management (SCM), and other OLTP applications.
- Use tiered storage for larger databases. Although data in-memory will give you extreme performance, look at tiered storage that leverages DRAM, Flash, SSD, and persistent memory to support larger databases that run into hundreds of terabytes or into petabytes.
To learn more about this topic, please see this webinar, where Noel serves as guest presenter: