Transactional cloud applications such as payment, booking, reservation systems, and complex business workflows are currently being rewritten for deployment in the cloud. This migration to the cloud is happening mainly for reasons of cost and scalability. Over the years, application developers have used different migration approaches, such as microservice frameworks, actors, and stateful dataflow systems. The migration to the cloud has brought back data management challenges traditionally handled by database management systems. Those challenges include ensuring state consistency, maintaining durability, and managing the application lifecycle. At the same time, the shift to a distributed computing infrastructure introduced new issues, such as message delivery, task scheduling, containerization, and (auto)scaling. Although the data management community has made progress in developing analytical and transactional database systems, transactional cloud applications have received little attention in database research. This tutorial aims to highlight recent trends in the area and discusses open research challenges for the data management community.

Executing applications in the cloud is becoming increasingly popular, primarily developed as microservices containing imperative code. In our previous work, we have made the case that such applications can benefit from using dataflow-based runtimes in a cloud environment. In particular, dataflow-based runtimes offer significant advantages over imperative code, namely, exactly-once processing, transparent message handling, and coarse-grained fault tolerance offered by dataflow systems. However, dataflow programming is not preferred by developers. In this work we bridge this gap, namely, we present our progress towards creating a suitable intermediate representation (IR) that can be used to compile stateful imperative code into dataflows, enabling seamless migration to the cloud. We then present a compiler pipeline prototype that offers two key benefits: i) it enables program optimizations and data parallelism, and ii) it decouples the input program from the target execution environment, while allowing interesting optimizations. Preliminary experiments demonstrate that our IR optimizations speed up the p50 request latency by 267x on average.

The rapid growth of large-scale machine learning (ML) models has led numerous commercial companies to utilize ML models for generating predictive results to help business decision-making. As two primary components in traditional predictive pipelines, data processing, and model predictions often operate in separate execution environments, leading to redundant engineering and computations. Additionally, the diverging mathematical foundations of data processing and machine learning hinder cross-optimizations by combining these two components, thereby overlooking potential opportunities to expedite predictive pipelines. In this paper, we propose an operator fusion method based on GPU-accelerated linear algebraic evaluation of relational queries. Our method leverages linear algebra computation properties to merge operators in machine learning predictions and data processing, significantly accelerating predictive pipelines by up to 317x. We perform a complexity analysis to deliver quantitative insights into the advantages of operator fusion, considering various data and model dimensions. Furthermore, we extensively evaluate linear algebra query processing and operator fusion utilizing the widely-used Star Schema and TPC-DI benchmarks. Through comprehensive evaluations, we demonstrate the effectiveness and potential of our approach in improving the efficiency of data processing and machine learning workloads on modern hardware.

Traditional monolithic applications are migrated to the cloud, typically using a microservice-like architecture. Although this migration leads to significant benefits such as scalability and development agility, it also leaves behind the transactional guarantees that database systems have provided to monolithic applications for decades. In the cloud era, developers build transactional and fault-tolerant distributed applications by explicitly programming transaction protocols at the application level.
In this paper, we argue that the principles behind the streaming dataflow execution model and deterministic transactional protocols provide a powerful and suitable substrate for executing transactional cloud applications. To this end, we introduce Styx, a transactional application runtime based on streaming dataflows that enables an object-oriented programming model for scalable, faulttolerant cloud applications with serializable guarantees.

Developing stateful cloud applications, such as low-latency workflows and microservices with strict consistency requirements, remains arduous for programmers. The Stateful Functions-as-a-Service (SFaaS) paradigm aims to serve these use cases. However, existing approaches provide weak transactional guarantees or perform expensive external state accesses requiring inefficient transactional protocols that increase execution latency. In this paper, we present Styx, a novel dataflow-based SFaaS runtime that executes serializable transactions consisting of stateful functions that form arbitrary call-graphs with exactly-once guarantees. Styx extends a deterministic transactional protocol by contributing: i) a function acknowledgment scheme to determine transaction boundaries required in SFaaS workloads, ii) a function-execution caching mechanism, and iii) an early commit-reply mechanism that substantially reduces transaction execution latency. Experiments with the YCSB, TPC-C, and Deathstar benchmarks show that Styx outperforms state-of-the-art approaches by achieving at least one order of magnitude higher throughput while exhibiting near-linear scalability and low latency.

Multiple works in data management research focus on automating the processes of data augmentation and feature discovery to save users from having to perform these tasks manually. Yet, this automation often leads to a disconnect with the users, as it fails to consider the specific needs and preferences of the actual end-users of data management systems for machine learning. To explore this issue further, we conducted 19 semi-structured, think-aloud use-case studies based on a scenario in which data specialists were tasked with augmenting a base table with additional features to train a machine learning model. In this paper, we share key insights into the practices of feature discovery on tabular data performed by real-world data specialists derived from our user study. Our research uncovered differences between the user assumptions reported in the literature and the actual practices, as well as some areas where literature and real-world practices align.

The advent of Large Language Models (LLMs) provides an opportunity to change the way queries are processed, moving beyond the constraints of conventional SQL-based database systems. However, using an LLM to answer a prediction query is still challenging, since an external ML model has to be employed and inference has to be performed in order to provide an answer. This paper introduces LLM-PQA, a novel tool that addresses prediction queries formulated in natural language. LLM-PQA is the first to combine the capabilities of LLMs and retrieval-augmented mechanism for the needs of prediction queries by integrating data lakes and model zoos. This integration provides users with access to a vast spectrum of heterogeneous data and diverse ML models, facilitating dynamic prediction query answering. In addition, LLM-PQA can dynamically train models on demand, based on specific query requirements, ensuring reliable and relevant results even when no pre-trained model in a model zoo, available for the task.

While the concept of large-scale stream processing is very popular nowadays, efficient dynamic allocation of resources is still an open issue in the area. The database research community has yet to evaluate different autoscaling techniques for stream processing engines under a robust benchmarking setting and evaluation framework. As a result, no conclusions can be made about the current solutions and problems that remain unsolved. Therefore, we address this issue with a principled evaluation approach.

This paper evaluates the state-of-the-art control-based solutions in the autoscaling area with diverse, dynamic workloads, applying specific metrics. We investigate different aspects of the autoscaling problem as performance and convergence. Our experiments reveal that current control-based autoscaling techniques fail to account for generated lag cost by rescaling or underprovisioning and cannot efficiently handle practical scenarios of intensely dynamic workloads. Unexpectedly, we discovered that an autoscaling method not tailored for streaming can outperform others in certain scenarios.

In recent years, researchers have developed several methods to automate discovering datasets and augmenting features for training Machine Learning (ML) models. Together with feature selection, these efforts have paved the way towards what is termed the feature discovery process. Data scientists and engineers use automated feature discovery over tabular datasets to add new features from different sources and enrich training data. By surveying data practitioners, we have observed that automated feature discovery approaches do not allow data scientists to use their domain knowledge during the feature discovery process. In addition, automated feature discovery methods can leak private features or introduce biased ones.

In this paper, we introduce the first user-driven human-in-the-loop feature discovery method called HILAutoFeat. We demonstrate the capabilities of HILAutoFeat, which effectively combines automated feature discovery with user-driven insights. Our demonstration is centred around two scenarios: (i) an automated feature discovery scenario -- HILAutoFeat acts as a steward in a large data lake where the user is unaware of the quality and relevance of the data, and (ii) a scenario where HILAutoFeat and the user work together -- the user drives the feature discovery process by adding his domain and business knowledge, while HILAutoFeat performs the intensive computations.

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.

Although the cloud has reached a state of robustness, the burden of using its resources falls on the shoulders of programmers who struggle to keep up with ever-growing cloud infrastructure services and abstractions. As a result, state management, scaling, operation, and failure management of scalable cloud applications require disproportionately more effort than developing the applications' actual business logic. Our vision aims to raise the abstraction level for programming scalable cloud applications by compiling stateful entities – a programming model enabling imperative transactional programs authored in Python – into stateful streaming dataflows. We propose a compiler pipeline that analyzes the abstract syntax tree of stateful entities and transforms them into an intermediate representation based on stateful dataflow graphs. It then compiles that intermediate representation into different dataflow engines, leveraging their exactly-once message processing guarantees to prevent state or failure management primitives from "leaking" into the level of the programming model. Preliminary experiments with a proof of concept implementation show that despite program transformation and translation to dataflows, stateful entities can perform at sub-100ms latency even for transactional workloads.

The proliferation of pre-trained ML models in public Web-based model zoos facilitates the engineering of ML pipelines to address complex inference queries over datasets and streams of unstructured content. Constructing optimal plan for a query is hard, especially when constraints (e.g. accuracy or execution time) must be taken into consideration, and the complexity of the inference query increases. To address this issue, we propose a method for optimizing ML inference queries that selects the most suitable ML models to use, as well as the order in which those models are executed. We formally define the constraint-based ML inference query optimization problem, formulate it as a Mixed Integer Programming (MIP) problem, and develop an optimizer that maximizes accuracy given constraints. This optimizer is capable of navigating a large search space to identify optimal query plans on various model zoos.

How can we perform similarity joins of multi-dimensional streams in a distributed fashion, achieving low latency? Can we adaptively repartition those streams in order to retain high performance under concept drifts? Current approaches to similarity joins are either restricted to single-node deployments or focus on set-similarity joins, failing to cover the ubiquitous case of metric-space similarity joins. In this paper, we propose the first adaptive distributed streaming similarity join approach that gracefully scales with variable velocity and distribution of multi-dimensional data streams. Our approach can adaptively rebalance the load of nodes in the case of concept drifts, allowing for similarity computations in the general metric space. We implement our approach on top of Apache Flink and evaluate its data partitioning and load balancing schemes on a set of synthetic datasets in terms of latency, comparisons ratio, and data duplication ratio

The increasing need for data trading across businesses nowadays has created a demand for data marketplaces. However, despite the intentions of both data providers and consumers, today’s data marketplaces remain mere data catalogs. We believe that marketplaces of the future require a set of value-added services, such as advanced search and discovery, that have been proposed in the database research community for years, but are not yet put to practice. With this paper, we report on the effort to engineer and develop an open-source modular data market platform to enable both entrepreneurs and researchers to setup and experiment with data marketplaces. To this end, we implemented and extended existing methods for data profiling, dataset search & discovery, and data recommendation. These methods are available as open-source libraries. In this paper we report on how those tools were assembled together to build topio.market, a real-world web platform for trading geospatial data, that is currently in a beta phase.

In this work, we evaluate autoscaling solutions for stream processing engines. Although autoscaling has become a mainstream subject of research in the last decade, the database research community has yet to evaluate different autoscaling techniques under a proper benchmarking setting and evaluation framework. As a result, every newly proposed autoscaling solution only performs a shallow performance evaluation and comparison against existing solutions. In this paper, we evaluate autoscaling solutions by employing two streaming queries and a dynamic workload that follows a cosinus pattern. Our experiments reveal that current autoscaling techniques fail to account for generated lag due to rescaling or underprovisioning and cannot efficiently handle practical scenarios of intensely dynamic workloads.

Machine learning (ML) researchers and practitioners are building repositories of pre-trained models, called model zoos. These model zoos contain metadata that detail various properties of the ML models and datasets, which are useful for reporting, auditing, reproducibility, and interpretability. Unfortunately, the existing metadata representations come with limited expressivity and lack of standardization. Meanwhile, an interoperable method to store and query model zoo metadata is missing. These two gaps hinder model search, reuse, comparison, and composition. In this demo paper, we advocate for standardized ML model metadata representation, proposing Macaroni, a metadata search engine with toolkits that support practitioners to obtain and enrich that metadata.

Machine learning (ML) practitioners and organizations are building model repositories of pre-trained models, referred to as model zoos. These model zoos contain metadata describing the properties of the ML models and datasets. The metadata serves crucial roles for reporting, auditing, ensuring reproducibility, and enhancing interpretability. Despite the growing adoption of descriptive formats like datasheets and model cards, the metadata available in existing model zoos remains notably limited. Moreover, existing formats have limited expressiveness, thus constraining the potential use of model repositories, extending their purpose beyond mere storage for pre-trained models. This paper proposes a unified metadata representation format for model zoos. We illustrate that comprehensive metadata enables a diverse range of applications, encompassing model search, reuse, comparison, and composition of ML models. We also detail the design and highlight the implementation of an advanced model zoo system built on top of our proposed metadata representation.

Machine learning (ML) training data is often scattered across disparate collections of datasets, called data silos. This fragmentation poses a major challenge for data-intensive ML applications: integrating and transforming data residing in different sources demand a lot of manual work and computational resources. With data privacy and security constraints, data often cannot leave the premises of data silos, hence model training should proceed in a decentralized manner. In this work, we present a vision of how to bridge the traditional data integration (DI) techniques with the requirements of modern machine learning. We explore the possibilities of utilizing metadata obtained from data integration processes for improving the effectiveness and efficiency of ML models. Towards this direction, we analyze two common use cases over data silos, feature augmentation and federated learning. Bringing data integration and machine learning together, we highlight new research opportunities from the aspects of systems, representations, factorized learning and federated learning.

Given a set of pre-trained Machine Learning (ML) models, can we solve complex analytic tasks that make use of those models by formulating ML inference queries? Can we mitigate different tradeoffs, e.g., high accuracy, low execution costs and memory footprint, when optimizing the queries? In this work we present different multi-objective ML inference query optimization strategies, and compare them on their usability, applicability, and complexity. We formulate Mixed-Integer-Programming-based (MIP) optimizers for ML inference queries that makes use of different objectives to find Pareto-optimal inference query plans.

Recent advances in Graphic Processing Units (GPUs) have facilitated a significant performance boost for database operators, in particular, joins. It has been intensively studied how conventional join implementations, such as hash joins, benefit from the massive parallelism of GPUs. With the proliferation of machine learning, more databases have started to provide native support for the basic building blocks of ML algorithms, i.e., linear algebra operators such as matrix multiplication (MM). Despite the recent increasing interest in processing relational joins using matrix multiplication (MM-join), two crucial questions still remain open: i) how efficient are current MM-join implementations compared to the GPU-based join algorithms; ii) how should practitioners choose among MM-join and conventional GPU-based joins given different data characteristics.In this paper, we compare the execution time, and memory I/O of MM-join against multiple GPU hash joins. An empirical analysis of our experimental results reveals that the state-of-the-art hash join implementation shows substantial scalability for various data characteristics. In contrast, MM-join outperforms the SOTA hash join in low join selectivity and low table cardinality but shows unsatisfactory scalability due to synchronous data movement and computation.