Real-Time Analytics: Building Blocks and Architecture
May 18, 2023
David Wang
There’s that saying “patience is a virtue”. But, in today’s day and age no one really wants to wait for anything. Is Netflix taking too long to load? Users will switch. Is the nearest Lyft too far? Users will switch.
That need for immediacy is also happening in data analytics, and it’s happening at scale on large data sets. The ability to deliver insights, make decisions and act on real-time data without users waiting (or requiring patience) is increasingly important. Companies like Netflix and Lyft but also Confluent and Target and 1000s of others are leaders in their industries in part because of real-time analytics and the data architectures that enable real-time, analytics-driven operations.
For data architects who are starting to think about real-time analytics, this blog unpacks what they are and the building blocks and data architecture that are preferred by many.
What are real-time analytics?
Real-time analytics are defined by two key attributes: fresh data and fast insights. They are used in latency-sensitive applications when it’s essential that new event-to-insight is measured in seconds.
Figure: Real-time analytics defined
In comparison, traditional analytics, which also goes by business intelligence, are static snapshots of business data used for reporting purposes and are powered by data warehouses like Snowflake and Amazon Redshift and visualized through BI tools such as Tableau or PowerBI.
While traditional analytics are built from historical data that can be hours, days, or weeks old, real-time analytics utilize recent data and are used in operational workflows that demand very fast answers to potentially complex questions.
Traditional Analytics
Real-Time Analytics
Long-running reports and exports
Minutes to hours to process
Historical, batch data
Catching queries is OK, as the data changes slowly
Rapid filters and aggregations
Sub-second queries
Real-time, streaming data
Data changes too fast to pre-compute queries
Figure: Decision criteria for real-time analytics
For example, a supply chain executive is looking for historical trends on monthly inventory changes: traditional analytics is perfect here. Why? Because the exec can probably wait a few minutes longer for the report to process. Alternatively, a security operations team is looking to identify and diagnose anomalies in network traffic. That’s a fit for real-time analytics as the SecOps team needs to rapidly mine thousands to millions of real-time log entries in sub-second to spot trends and investigate abnormal behavior.
Does the right architecture matter?
A lot of database vendors will say they’re good for real-time analytics and they are…to a degree. Take for example weather monitoring. Let’s say the use case calls for sampling temperature every second across 1000s of weather stations with queries that include threshold-based alerts and some trend analysis. This would be easy for SingleStore, InfluxDB, MongoDB, even PostgreSQL. Write a push API that sends the metrics directly to the database and then a simple query gets executed and voila…real-time analytics.
So when do real-time analytics actually get hard? In the example above, the data set is pretty small and the analytics are pretty simple. A single temperature event is only generated once every second and a SELECT with WHERE statement to capture the latest events doesn’t require much processing power. Easy for any time-series or OLTP database.
Things start getting challenging and pushing the limits of databases when the volume of events ingested gets higher, the queries involve a lot of dimensions, and data sets are in the terabytes (if not petabytes). Apache Cassandra might come to mind for high throughput ingestion. But analytics performance wouldn’t be great. Maybe the analytics use case calls for joining multiple real-time data sources at scale. What to do then?
Here are some considerations to think about that’ll help define the requirements for the right architecture:
Are you working with high events per second, from 1000s to millions?
Is it important to minimize latency between events created to when they can be queried?
Is your total dataset large, and not just a few GB?
How important is query performance – sub-second or minutes per query?
How complicated are the queries, exporting a few rows or large scale aggregations?
Is avoiding downtime of the data stream and analytics engine important?
Are you trying to join multiple event streams for analysis?
Do you need to place real-time data in context with historical data?
Do you anticipate many concurrent queries?
If any of these things matter, let’s talk about what that right architecture looks like.
Building blocks
Real-time analytics needs more than a capable database. It starts with needing to connect, deliver, and manage real-time data. That brings us to the first building block: event streaming.
1. Event streaming
When real-time matters, batch-based data pipelines take too long and that’s why messaging queues emerged. Traditionally, delivering messages involved tools like ActiveMQ, RabbitMQ, and TIBCO. But the new way is event streaming with Apache Kafka and Amazon Kinesis.
Apache Kafka and Amazon Kinesis overcome the scale limitations of traditional messaging queues, enabling high throughput pub/sub to collect and deliver large streams of event data from a variety of sources (Amazon lingo: producers) to a variety of sinks (Amazon lingo: consumers) in real-time.
Figure: Apache Kafka event streaming pipeline
Those systems capture data in real-time from sources like databases, sensors, and cloud services in the form of event streams and deliver them to other applications, databases, and services.
Because the systems can scale (Apache Kafka at LinkedIn supports over 7 trillion messages a day) and handle multiple, concurrent data sources, event streaming has become the de facto delivery vehicle when applications need real-time data.
So now that we can capture real-time data, how do we go about analyzing it in real-time?
2. Real-time analytics database
Real-time analytics need a purpose-built database, a database that can take full advantage of streaming data in Apache Kafka and Amazon Kinesis and deliver insights in real-time. That’s Apache Druid.
As a high-performance, real-time analytics database built for streaming data, Apache Druid has become the database-of-choice for building real-time analytics applications. It supports true stream ingestion and handles large aggregations on TBs to PBs of data at sub-second performance under load. And since it has a native integration with Apache Kafka and Amazon Kinesis it makes it the go-to choice whenever fast insights on fresh data is needed.
Scale, latency, and data quality are all important when selecting the analytics database for streaming data. Can it handle the full-scale of event streaming? Can it ingest and correlate multiple Kafka topics (or Kinesis shards)? Can it support event-based ingestion? Can it avoid data loss or duplicates in the event of a disruption? Apache Druid can do all of that and more.
Druid was designed from the outset for rapid ingestion and immediate querying of events on arrival. For streaming data, it ingests event-by-event, not a series of batch data files sent sequentially to mimic a stream. There’s no connectors to Kafka or Kinesis needed and Druid supports exactly-once semantics to ensure data quality.
Just like how Apache Kafka was built for internet-scale event data, Apache Druid was too. Its services-based architecture independently scales ingestion and query processing practically infinitely. Druid maps ingestion tasks with Kafka partitions, so as Kafka clusters scale Druid can scale right alongside it.
Figure: How Druid’s real-time ingestion is as scalable as Kafka
It’s not that uncommon to see companies ingesting millions of events per second into Druid. For example, Confluent – the originators behind Kafka – built their observability platform with Druid and ingests over 5 million events per second from Kafka.
But real-time analytics needs more than just real-time data. Making sense of real-time patterns and behavior requires correlating historical data. One of Druid’s strengths, as shown in the diagram above, is its ability to both support real-time and historical insights via a single SQL query with Druid managing up to PBs of data efficiently in the background.
So when you pull this all together you end up with a very scalable data architecture for real-time analytics. It’s the architecture 1000s of data architects choose when high scalability, low latency, and complex aggregations are needed from real-time data.
Figure: Data architecture for real-time analytics
Example: How Netflix Ensures a High-Quality Experience
Real-time analytics plays a key role in Netflix’s ability to deliver a consistently great experience for more than 200 million users enjoying 250 million hours of content every day. Netflix built an observability application for real-time monitoring of over 300 million devices.
Using real-time logs from playback devices streamed through Apache Kafka and then ingested event-by-event into Apache Druid, Netflix is able to derive measurements that understand and quantify how user devices are handling browsing and playback.
With over 2 million events per second and subsecond queries across 1.5 trillion rows, Netflix engineers are able to pinpoint anomalies within their infrastructure, endpoint activity, and content flow.
Parth Brahmbhatt, Senior Software Engineer, Netflix summarizes it best:
“Druid is our choice for anything where you need subsecond latency, any user interactive dashboarding, any reporting where you expect somebody on the other end to actually be waiting for a response. If you want super fast, low latency, less than a second, that’s when we recommend Druid.”
Conclusion
If you’re looking to build real-time analytics, I’d highly recommend checking out Apache Druid along with Apache Kafka and Amazon Kinesis. You can download Apache Druid from druid.apache.org or simply try out Imply Polaris, the cloud database service for Apache Druid, for free.
Other blogs you might find interesting
No records found...
Sep 21, 2023
Migrate Analytics Data from MongoDB to Apache Druid
This blog presents a concise guide on migrating data from MongoDB to Druid. It includes Python scripts to extract data from MongoDB, save it as CSV, and then ingest it into Druid. It also touches on maintaining...
How Druid Facilitates Real-Time Analytics for Mass Transit
Mass transit plays a key role in reimagining life in a warmer, more densely populated world. Learn how Apache Druid helps power data and analytics for mass transit.
Migrate Analytics Data from Snowflake to Apache Druid
This blog outlines the steps needed to migrate data from Snowflake to Apache Druid, a platform designed for high-performance analytical queries. The article covers the migration process, including Python scripts...
Apache Kafka, Flink, and Druid: Open Source Essentials for Real-Time Applications
Apache Kafka, Flink, and Druid, when used together, create a real-time data architecture that eliminates all these wait states. In this blog post, we’ll explore how the combination of these tools enables...
Visualizing Data in Apache Druid with the Plotly Python Library
In today's data-driven world, making sense of vast datasets can be a daunting task. Visualizing this data can transform complicated patterns into actionable insights. This blog delves into the utilization of...
Bringing Real-Time Data to Solar Power with Apache Druid
In a rapidly warming world, solar power is critical for decarbonization. Learn how Apache Druid empowers a solar equipment manufacturer to provide real-time data to users, from utility plant operators to homeowners
When to Build (Versus Buy) an Observability Application
Observability is the key to software reliability. Here’s how to decide whether to build or buy your own solution—and why Apache Druid is a popular database for real-time observability
How Innowatts Simplifies Utility Management with Apache Druid
Data is a key driver of progress and innovation in all aspects of our society and economy. By bringing digital data to physical hardware, the Internet of Things (IoT) bridges the gap between the online and...
Three Ways to Use Apache Druid for Machine Learning Workflows
An excellent addition to any machine learning environment, Apache Druid® can facilitate analytics, streamline monitoring, and add real-time data to operations and training
Apache Druid® is an open-source distributed database designed for real-time analytics at scale. Apache Druid 27.0 contains over 350 commits & 46 contributors. This release's focus is on stability and scaling...
Unleashing Real-Time Analytics in APJ: Introducing Imply Polaris on AWS AP-South-1
Imply, the company founded by the original creators of Apache Druid, has exciting news for developers in India seeking to build real-time analytics applications. Introducing Imply Polaris, a powerful database-as-a-Service...
In this guide, we will walk you through creating a very simple web app that shows a different embedded chart for each user selected from a drop-down. While this example is simple it highlights the possibilities...
Automate Streaming Data Ingestion with Kafka and Druid
In this blog post, we explore the integration of Kafka and Druid for data stream management and analysis, emphasizing automatic topic detection and ingestion. We delve into the creation of 'Ingestion Spec',...
This guide explores configuring Apache Druid to receive Kafka streaming messages. To demonstrate Druid's game-changing automatic schema discovery. Using a real-world scenario where data changes are handled...
Imply Polaris, our ever-evolving Database-as-a-Service, recently focused on global expansion, enhanced security, and improved data handling and visualization. This fully managed cloud service, based on Apache...
Introducing hands-on developer tutorials for Apache Druid
The objective of this blog is to introduce the new set of interactive tutorials focused on the Druid API fundamentals. These tutorials are available as Jupyter Notebooks and can be downloaded as a Docker container.
In this blog article I’ll unpack schema auto-discovery, a new feature now available in Druid 26.0, that enables Druid to automatically discover data fields and data types and update tables to match changing...
Druid now has a new function, Unnest. Unnest explodes an array into individual elements. This blog contains design methodology and examples for this new Unnest function both from native and SQL binding perspectives.
What’s new in Imply Polaris – Our Real-Time Analytics DBaaS
Every week we add new features and capabilities to Imply Polaris. This month, we’ve expanded security capabilities, added new query functionality, and made it easier to monitor your service with your preferred...
Apache Druid® 26.0, an open-source distributed database for real-time analytics, has seen significant improvements with 411 new commits, a 40% increase from version 25.0. The expanded contributor base of 60...
How to Build a Sentiment Analysis Application with ChatGPT and Druid
Leveraging ChatGPT for sentiment analysis, when combined with Apache Druid, offers results from large data volumes. This integration is easily achievable, revealing valuable insights and trends for businesses...
In this blog, we will compare Snowflake and Druid. It is important to note that reporting data warehouses and real-time analytics databases are different domains. Choosing the right tool for your specific requirements...
Learn how to achieve sub-second responses with Apache Druid
Learn how to achieve sub-second responses with Apache Druid. This article is an in-depth look at how Druid resolves queries and describes data modeling techniques that improve performance.
Apache Druid uses load rules to manage the ageing of segments from one historical tier to another and finally to purge old segments from the cluster. In this article, we’ll show what happens when you make...
What’s new in Imply Polaris – Our Real-Time Analytics DBaaS
This blog explains some of the new features, functionality and connectivity added to Imply Polaris over the last two months. We've expanded ingestion capabilities, simplified operations and increased reliability...
Wow, that was easy – Up and running with Apache Druid
The objective of this blog is to provide a step-by-step guide on setting up Druid locally, including the use of SQL ingestion for importing data and executing analytical queries.
Tales at Scale Podcast Kicks off with the Apache Druid Origin Story
Tales at Scale cracks open the world of analytics projects and shares stories from developers and engineers who are building analytics applications or working within the real-time data space. One of the key...
Real-time Analytics Database uses partitioning and pruning to achieve its legendary performance
Apache Druid uses partitioning (splitting data) and pruning (selecting subset of data) to achieve its legendary performance. Learn how to use the CLUSTERED BY clause during ingestion for performance and high...
Easily embed analytics into your own apps with Imply’s DBaaS
This blog explains how developers can leverage Imply Polaris to embed robust visualization options directly into their own applications without them having to build a UI. This is super important because consuming...
Building an Event Analytics Pipeline with Confluent Cloud and Imply’s real time DBaaS, Polaris
Learn how to set up a pipeline that generates a simulated clickstream event stream and sends it to Confluent Cloud, processes the raw clickstream data using managed ksqlDB in Confluent Cloud, delivers the processed...
We are excited to announce the availability of Imply Polaris in Europe, specifically in AWS eu-central-1 region based in Frankfurt. Since its launch in March 2022, Imply Polaris, the fully managed Database-as-a-Service...
Should You Build or Buy Security Analytics for SecOps?
When should you build—or buy—a security analytics platform for your environment? Here are some common considerations—and how Apache Druid is the ideal foundation for any in-house security solution.
Combating financial fraud and money laundering at scale with Apache Druid
Learn how Apache Druid enables financial services firms and FinTech companies to get immediate insights from petabytes-plus data volumes for anti-fraud and anti-money laundering compliance.
This is a what's new to Imply in Dec 2022. We’ve added two new features to Imply Polaris to make it easier for your end users to take advantage of real-time insights.
Imply Pivot delivers the final mile for modern analytics applications
This blog is focused on how Imply Pivot delivers the final mile for building an anlaytics app. It showcases two customer examples - Twitch and ironsource.
For decades, analytics has been defined by the standard reporting and BI workflow, supported by the data warehouse. Now, 1000s of companies are realizing an expansion of analytics beyond reporting, which requires...
Apache Druid is at the heart of Imply. We’re an open source business, and that’s why we’re committed to making Druid the best open source database for modern analytics applications
When it comes to modern data analytics applications, speed is of the utmost importance. In this blog we discuss two approximation algorithms which can be used to greatly enhance speed with only a slight reduction...
The next chapter for Imply Polaris: celebrating 250+ accounts, continued innovation
Today we announced the next iteration of Imply Polaris, the fully managed Database-as-a-Service that helps you build modern analytics applications faster, cheaper, and with less effort. Since its launch in...
We obviously talk a lot about #ApacheDruid on here. But what are folks actually building with Druid? What is a modern analytics application, exactly? Let's find out
Elasticity is important, but beware the database that can only save you money when your application is not in use. The best solution will have excellent price-performance under all conditions.
Druid 0.23 – Features And Capabilities For Advanced Scenarios
Many of Druid’s improvements focus on building a solid foundation, including making the system more stable, easier to use, faster to scale, and better integrated with the rest of the data ecosystem. But for...
Apache Druid 0.23.0 contains over 450 updates, including new features, major performance enhancements, bug fixes, and major documentation improvements.
Imply Polaris is a fully managed database-as-a-service for building realtime analytics applications. John is the tech lead for the Polaris UI, known internally as the Unified App. It began with a profound question:...
There is a new category within data analytics emerging which is not centered in the world of reports and dashboards (the purview of data analysts and data scientists), but instead centered in the world of applications...
We are in the early stages of a stream revolution, as developers build modern transactional and analytic applications that use real-time data continuously delivered.
Developers and architects must look beyond query performance to understand the operational realities of growing and managing a high performance database and if it will consume their valuable time.
Building high performance logging analytics with Polaris and Logstash
When you think of querying with Apache Druid, you probably imagine queries over massive data sets that run in less than a second. This blog is about some of the things we did as a team to discover the user...
Horizontal scaling is the key to performance at scale, which is why every database claims this. You should investigate, though, to see how much effort it takes, especially compared to Apache Druid.
When you think of querying with Apache Druid, you probably imagine queries over massive data sets that run in less than a second. This blog is about some of the things we did as a team to discover the user...
Building Analytics for External Users is a Whole Different Animal
Analytics aren’t just for internal stakeholders anymore. If you’re building an analytics application for customers, then you’re probably wondering…what’s the right database backend?
After over 30 years of working with data analytics, we’ve been witness (and sometimes participant) to three major shifts in how we find insights from data - and now we’re looking at the fourth.
Every year industry pundits predict data and analytics becoming more valuable the following year. But this doesn’t take a crystal ball to predict. There’s instead something much more interesting happening...
Today, I'm prepared to share our progress on this effort and some of our plans for the future. But before diving further into that, let's take a closer look at how Druid's core query engine executes queries,...
Product Update: SSO, Cluster level authorization, OAuth 2.0 and more security features
When you think of querying with Apache Druid, you probably imagine queries over massive data sets that run in less than a second. This blog is about some of the things we did as a team to discover the user...
When you think of querying with Apache Druid, you probably imagine queries over massive data sets that run in less than a second. This blog is about some of the things we did as a team to discover the user...
Druid Nails Cost Efficiency Challenge Against ClickHouse & Rockset
To make a long story short, we were pleased to confirm that Druid is 2 times faster than ClickHouse and 8 times faster than Rockset with fewer hardware resources!.
Unveiling Project Shapeshift Nov. 9th at Druid Summit 2021
There is a new category within data analytics emerging which is not centered in the world of reports and dashboards (the purview of data analysts and data scientists), but instead centered in the world of applications...
How we made long-running queries work in Apache Druid
When you think of querying with Apache Druid, you probably imagine queries over massive data sets that run in less than a second. This blog is about some of the things we did as a team to discover the user...
Uneven traffic flow in streaming pipelines is a common problem. Providing the right level of resources to keep up with spikes in demand is a requirement in order to deliver timely analytics.
Community Discoveries: multi-value dimensions in Apache Druid
Hellmar Becker is an Imply solutions engineer based in Germany, where he has been delving into the nooks-and-crannies of multi-valued dimension support in Druid. In this interview, Hellmar explains why...
Community Spotlight: Apache Pulsar and Apache Druid get close…
The community team at Imply spoke with an Apache Pulsar community member, Giannis Polyzos, about how collaboration between open source communities generates great things, and more specifically, about how...
Meet the team: Abhishek Agarwal, engineering lead in India
Abhishek is Imply’s first engineer in India. We spoke to him about setting up our operations in Bangalore and asked what kind of local talent the company is looking for.
Jihoon Son is a software engineer at Imply who works on Apache Druid®. He explains what drew him to Imply five years ago and why he’s even more inspired by the company today.