Stream processing in the last decade has seen broad adoption in both commercial and research settings. One key element for this success is the ability of modern stream processors to handle failures while ensuring exactly-once processing guarantees. At the moment of writing, virtually all stream processors that guarantee exactly-once processing implement a variant of Apache Flink's coordinated checkpoints - an extension of the original Chandy-Lamport checkpoints from 1985. However, the reasons behind this prevalence of the coordinated approach remain anecdotal, as reported by practitioners of the stream processing community. At the same time, common checkpointing approaches, such as the uncoordinated and the communication-induced ones, remain largely unexplored. This paper is the first to address this gap by i) shedding light on why practitioners have favored the coordinated approach and ii) investigating whether there are viable alternatives. To this end, we implement three checkpointing approaches that we surveyed and adapted for the distinct needs of streaming dataflows. Our analysis shows that the coordinated approach outperforms the uncoordinated and communication-induced protocols under uniformly distributed workloads. To our surprise, however, the uncoordinated approach is not only competitive to the coordinated one in uniformly distributed workloads, but it also outperforms the coordinated approach in skewed workloads. We conclude that rather than blindly employing coordinated checkpointing, research should focus on optimizing the very promising uncoordinated approach, as it can address issues with skew and support prevalent cyclic queries. We believe that our findings can trigger further research into checkpointing mechanisms.@en

Stream processing has been an active research field for more than 20 years, but it is now witnessing its prime time due to recent successful efforts by the research community and numerous worldwide open-source communities. This survey provides a comprehensive overview of fundamental aspects of stream processing systems and their evolution in the functional areas of out-of-order data management, state management, fault tolerance, high availability, load management, elasticity, and reconfiguration. We review noteworthy past research findings, outline the similarities and differences between the first (’00–’10) and second (’11–’23) generation of stream processing systems, and discuss future trends and open problems.

@en

Stream processing has been an active research field for more than 20 years, but it is now witnessing its prime time due to recent successful efforts by the research community and numerous worldwide open-source communities. The goal of this tutorial is threefold. First, we aim to review and highlight noteworthy past research findings, which were largely ignored until very recently. Second, we intend to underline the differences between early ('00-'10) and modern ('11-'18) streaming systems, and how those systems have evolved through the years. Most importantly, we wish to turn the attention of the database community to recent trends: streaming systems are no longer used only for classic stream processing workloads, namely window aggregates and joins. Instead, modern streaming systems are being increasingly used to deploy general event-driven applications in a scalable fashion, challenging the design decisions, architecture and intended use of existing stream processing systems.

@en

Sliding windows are bounded sets which evolve together with an infinite data stream of records. Each new sliding window evicts records from the previous one while introducing newly arrived records as well. Aggregations on windows typically derive some metric such as an average or a sum of a value in each window. The main challenge of applying aggregations to sliding windows is that a naive execution can lead to a high degree of redundant computation due to a large number of common records across different windows. Special optimization techniques have been developed throughout the years to tackle redundancy and make sliding window aggregation feasible and more efficient in large data streams@en

In our data-centric society, online services, decision making, and other aspects are increasingly becoming heavily dependent on trends and patterns extracted from data. A broad class of societal-scale data management problems requires system support for processing unbounded data with low latency and high throughput. Large-scale data stream processing systems perceive data as infinite streams and are designed to satisfy such requirements. They have further evolved substantially both in terms of expressive programming model support and also efficient and durable runtime execution on commodity clusters. Expressive programming models offer convenient ways to declare continuous data properties and applied computations, while hiding details on how these data streams are physically processed and orchestrated in a distributed environment. Execution engines provide a runtime for such models further allowing for scalable yet durable execution of any declared computation. In this chapter we introduce the major design aspects of large scale data stream processing systems, covering programming model abstraction levels and runtime concerns. We then present a detailed case study on stateful stream processing with Apache Flink, an open-source stream processor that is used for a wide variety of processing tasks. Finally, we address the main challenges of disruptive applications that large-scale data streaming enables from a systemic point of view.

@en

Aggregation queries on data streams are evaluated over evolving and often overlapping logical views called windows. While the aggregation of periodic windows were extensively studied in the past through the use of aggregate sharing techniques such as Panes and Pairs, little to no work has been put in optimizing the aggregation of very common, non-periodic windows. Typical examples of non-periodic windows are punctuations and sessions which can implement complex business logic and are often expressed as user-defined operators on platforms such as Google Dataflow or Apache Storm. The aggregation of such non-periodic or user-defined windows either falls back to expensive, best-effort aggregate sharing methods, or is not optimized at all. In this paper we present a technique to perform efficient aggregate sharing for data stream windows, which are declared as user-defined functions (UDFs) and can contain arbitrary business logic. To this end, we first introduce the concept of User-Defined Windows (UDWs), a simple, UDF-based programming abstraction that allows users to programmatically define custom windows. We then define semantics for UDWs, based on which we design Cutty, a low-cost aggregate sharing technique. Cutty improves and outperforms the state of the art for aggregate sharing on single and multiple queries. Moreover, it enables aggregate sharing for a broad class of non-periodic UDWs. We implemented our techniques on Apache Flink, an open source stream processing system, and performed experiments demonstrating orders of magnitude of reduction in aggregation costs compared to the state of the art.

@en

Apache Flink is an open-source system for processing streaming and batch data. Flink is built on the philosophy that many classes of data processing applications, including real-time analytics, continuous data pipelines, historic data processing (batch), and iterative algorithms (machine learning, graph analysis) can be expressed and executed as pipelined fault-tolerant dataflows. In this paper, we present Flink’s architecture and expand on how a (seemingly diverse) set of use cases can be unified under asingle execution model.@en