World Congress 2024 • Aug 20, 2024 • Session details

Why and when should we consider Stream Processing frameworks in our solutions

Soroosh Khodami

Are you over-engineering your data pipeline? Learn when to leverage complex stream processing frameworks like Apache Flink, and when simple SQL micro-batching is the smarter architectural choice.

Pause
Mute Enter Fullscreen
#1 about 3 min

Defining stream data processing versus standard event processing

Differentiating continuous data aggregation from discrete message decision making.

#2 about 3 min

Handling message order with event and processing time

Analyzing data by actual event occurrence to guarantee offline state consistency.

#3 about 3 min

Ensuring exactly once message delivery in distributed systems

Utilizing stream framework checkpoint mechanisms to completely avoid duplicate data processing.

#4 about 3 min

Constructing data pipelines with stream processing architecture operators

Connecting basic operator functions to correctly read, transform, and sync continuous inputs.

#5 about 6 min

Applying windowing functions to segment unbounded data streams

Segmenting unbounded continuous data streams using distinct tumbling, sliding, and session windows.

#6 about 4 min

Joining data streams and managing stateful memory processing

Combining separate message sources and retaining critical memory state for complex temporal analytics.

#7 about 2 min

Evaluating mature stream processing frameworks for production systems

Reviewing industry-standard data tools like Apache Flink and Beam for scalable application architecture.

#8 about 5 min

Comparing performance and complexity against standard microservice architectures

Benchmarking Apache Beam data pipelines against Spring Boot to assess throughput and setup overhead.

#9 about 3 min

Assessing the organizational cost of stream framework adoption

Weighing high execution speed against steep learning curves and heavy engineering maintenance bandwidth.

#10 about 1 min

Identifying real-world applications for stream processing technology usage

Exploring practical data implementations across gaming anti-cheat systems, telecommunications, and financial fraud detection.

#11 about 2 min

Establishing adoption prerequisites and business stakeholder strategic alignment

Ensuring transparent business requirements exist before committing complex engineering logic to pipeline structural changes.

Matching moments

1:33 min

Overcoming typical barriers to real-time stream processing adoption

Bobur Umurzokov · LIVE

1:56 min

Overview of current stream processing frameworks

Hartmut Armbruster Hartmut Armbruster · World Congress 2026 Europe

4:14 min

Recognizing architectural drivers pushing event streaming system adoption

Mary Grygleski Mary Grygleski · LIVE

4:30 min

Introducing data management and the shift to streaming

Mary Grygleski Mary Grygleski · LIVE

1:32 min

Comparing stream processing architecture to traditional batch processing

Bobur Umurzokov · LIVE

4:38 min

Operational complexities and performance optimization of stream applications

Denis Washington +1 · World Congress 2021