Overview
Concurrency is a crucial function for enterprises who wish to integrate data from multiple sources and ensure real-time data accessibility across the organization. Learn about the three key factors required to execute the level of concurrency outlined in this article.
What Is Concurrency?
Concurrency, or concurrent data processing, occurs when users access, interact with, or make changes to data at the same time. It’s a crucial function for enterprises using either traditional or cloud-based data warehousing to integrate data from multiple sources and glean business intelligence in real-time.
Businesses that need to scale the number of users and queries can find it challenging to manage these environments effectively. Teradata Senior Technologist Rob Armstrong calls this the “ten hundred thousand problem” — 10 users running 10 queries involving 10 rows or tables may not be hard to manage, but scaling to thousands or millions creates an environment that’s impossible for human beings to manage.
Moving from departmental point solutions to operational and production systems working at scale requires finely tuned concurrency. Three key factors are required to execute this level of concurrency effectively:
1. Optimizing individual queries
Users cannot necessarily understand all the rules or tools required to create optimized queries, so you need an optimizer that understands and leverages complex data relationships and cues.
2. Managing the diverse workload
Not every query is equal in importance. Effective workload management distinguishes between web access queries with stringent service levels that must be met versus long running queries where the run time is less of a concern.
3. Overall system monitoring
Keeping the entire environment running smoothly as a single system requires message traffic monitoring, error recovery, space management, ensuring continuity, and reducing failover.
Why Does Concurrency Matter?
In today’s digitized world, businesses across industries have to effectively manage and analyze the constant streams of data being produced by people, products, and machines, every second of every day. In order to capture intelligence from this flood of information, organizations must create a secure, open, accessible, and real-time repository of data that’s available to employees across the enterprise. By relying on people in various roles to bring their expertise and perspective to this data, businesses can surface answers more quickly as their people innovate together.
Concurrency is crucial to supporting this level of collaboration. If it’s possible that a user’s query on a data set could impact another user’s query, either by increasing lag times or hurting the integrity of the data, then people across the organization will no longer trust the accuracy and reliability of their analytics infrastructure. In some instances, individuals and teams within a corporation have created their own data marts in order to bypass slow processes. This can result in a “Wild, Wild West” environment, where different departments and teams have their own versions of the truth.
Concurrency ensures that everyone works from the same real-time data without having a negative impact on anyone else. This allows innovation to proceed without delay and for people to have full confidence that the data being used is accurate.
Teradata’s Perspective on Concurrency
At Teradata, we believe that to survive and thrive in today’s competitive business environment, organizations must be capable of managing all of the data, all of the time to deliver analytics that matter. Concurrency is an important part of a unified, integrated analytics environment that delivers the best analytics functions and engines at scale.
Teradata Vantage solves the “ten hundred thousand” concurrency problem, allowing your business to move from insight to production while driving your bottom line. The leading hybrid cloud data analytics software, Vantage leverages 100% of your data to analyze anything, anywhere, at any time. Vantage is built upon the Teradata database, which has the richest optimizer for analytic workloads, has world-class workload management, and provides a full suite of tools to simplify and automate multi-system ecosystems.