This paper focuses on the micro-service migration problem with affinity, stemming from the cloud computing industry. Because of periodically creating and deleting micro-services to satisfy users’ demands, the deployment of micro-services in the cloud needs to be regularly adjusted, which is referred to as a micro-service migration. An optimal migration schedule should minimize the number of activated physical machines as well as maximize total internal invoking traffic (affinity). A cooperative multi-agent reinforcement learning (MARL) is proposed, which is enhanced by integrating Hindsight Reward Shaping and by fine-tuning the state encoder using a pre-trained ResNet model. The proposed MARL is validated on both synthetic datasets and real cloud traces of ByteDance and Alibaba, compared with four baseline algorithms: Migration Ant Colony Optimization , Migration Neighborhood Search, Single-Agent Reinforcement Learning, and the optimization solver CPLEX. Finally, an evaluation mechanism called Matching Score is proposed to explain the superior performance of MARL.

Text-to-SQL, the task of translating natural language questions into SQL queries, plays a crucial role in enabling non-experts to interact with databases. While recent advancements in large language models (LLMs) have significantly enhanced text-to-SQL performance, existing approaches face notable limitations in real-world text-to-SQL applications. Prompting-based methods often depend on closed-source LLMs, which are expensive, raise privacy concerns, and lack customization. Fine-tuning-based methods, on the other hand, suffer from poor generalizability due to the limited coverage of publicly available training data. To overcome these challenges, we propose a novel and scalable text-to-SQL data synthesis framework for automatically synthesizing large-scale, high-quality, and diverse datasets without extensive human intervention. Using this framework, we introduce SynSQL-2.5M, the first million-scale text-to-SQL dataset, containing 2.5 million samples spanning over 16,000 synthetic databases. Each sample includes a database, SQL query, natural language question, and chain-of-thought (CoT) solution. Leveraging SynSQL-2.5M, we develop OmniSQL, a powerful open-source text-to-SQL model available in three sizes: 7B, 14B, and 32B. Extensive evaluations across nine datasets demonstrate that OmniSQL achieves state-of-the-art performance, matching or surpassing leading closed-source and open-source LLMs, including GPT-4o and DeepSeek-V3, despite its smaller size. We release all code, datasets, and models to support further research.

Workload forecasting is pivotal in cloud service applications, such as auto-scaling and scheduling, with profound implications for operational efficiency. Although Transformer-based forecasting models have demonstrated remarkable success in general tasks, their computational efficiency often falls short of the stringent requirements in large-scale cloud environments. Given that most workload series exhibit complicated periodic patterns, addressing these challenges in the frequency domain offers substantial advantages. To this end, we propose Fremer, an efficient and effective deep forecasting model. Fremer fulfills three critical requirements: it demonstrates superior efficiency, outperforming most Transformer-based forecasting models; it achieves exceptional accuracy, surpassing all state-of-the-art (SOTA) models in workload forecasting; and it exhibits robust performance for multi-period series. Furthermore, we collect and open-source four high-quality, open-source workload datasets derived from ByteDance's cloud services, encompassing workload data from thousands of computing instances. Extensive experiments on both our proprietary datasets and public benchmarks demonstrate that Fremer consistently outperforms baseline models, achieving average improvements of 5.5% in MSE, 4.7% in MAE, and 8.6% in SMAPE over SOTA models, while simultaneously reducing parameter scale and computational costs. Additionally, in a proactive auto-scaling test based on Kubernetes, Fremer improves average latency by 18.78% and reduces resource consumption by 2.35%, underscoring its practical efficacy in real-world applications.

Containerization has garnered substantial favor among cloud service providers. Nevertheless, the notable network overhead incurred between containers has prompted concerns within the community. In cloud resource scheduling, collocating service containers that frequently communicate to the same machine - termed “service affinity” - is instrumental in enhancing application performance. In response to this concern, we present a solution that harnesses service affinity and collocates containers to enhance the overall system performance and stability. To maximize the benefits of collocating containers, it is necessary to calculate a new schedule that optimally and efficiently maximizes service affinity, especially within the expansive domain of industry-scale cloud environments. In pursuit of this, we leverage the skewness property of affinity and machine learning to fuse solver-based algorithms, thereby assuring both quality and efficiency for problems at scale. Our methodology encompasses the partitioning of a given task into discrete subproblems, with a keen focus on resolving the most critical ones. Via a graph neural network classifier, we assign each subproblem to be solved independently using methods based on off-the-shelf solvers in our algorithm pool - namely, MIP-based, or column generation. This strategic approach enables the efficient computation of a schedule for a cloud cluster that fully optimizes the overall service affinity. We further propose a heuristic algorithm to compute executable container migration plans for practical use, facilitating the transition to the new placement where service affinity is well optimized. Our solution has been deployed in our large-scale production environment, covering over a million cores within ByteDance. Through the successful real-world production deployment, our approach exhibits an average improvement in end-to-end latency by 23.75% and a reduction in request error rates by 24.09% compared to the original system.

Storage space usage forecasting is critical for the scalability and stability of storage systems. Cloud providers estimate storage usages based on the forecast and allocate resources accordingly. Overestimated space usages require a redundant storage buffer that brings unnecessary cost, and underestimated space usages will cause capacity shortages that may lead to data loss and Service-Level Agreement (SLA) failures. While accurate storage forecasting is important, it is highly challenging due to various storage usage patterns on different workloads and storage systems. Moreover, some operations from users or administrators may cause transient workload burst in historical data, which makes forecasting even harder.In this paper, we propose the Storage Usage Forecasting Service (SUFS) that combines deep neural networks and statistical models adaptively to make predictions for multiple major storage systems in ByteDance. SUFS carries comprehensive analyses of storage usage time series from various storage systems in real business scenarios. To handle workload bursts in historical data, we enhance regular LSTMs using a control signal that is installed on the input gate. When the burst is detected, the control signal reduces the input influences to the cell state. To further improve the prediction accuracy, SUFS integrates the Enhanced-LSTM (ELSTM) with a novel adaptive ensemble method. Different from previous works, our approach learns dynamic ensemble weights for each prediction step on-the-fly, making our model more accurate for multiple-step predictions. SUFS has been deployed to serve more than 150,000 storage instances. We conducted extensive experiments on the storage systems that are widely-used in ByteDance, and the results show that SUFS outperforms the state-of-the-art methods and significantly reduces storage cost.

In large-scale systems, due to system complexity and demand volatility, diverse and dynamic workloads make accurate predictions difficult. In this work, we address an online interval prediction problem (OnPred-Int) and adopt ensemble learning to solve it. We depict that the ensemble learning for OnPred-Int is a dynamic deterministic Markov Decision Process (Dd-MDP) and convert it into a stateful online learning task. Then we propose IPOC, a lightweight and flexible model able to produce effective confidence intervals, adapting the dynamics of real-time workload streams. At each time, IPOC selects a target model and executes chasing for it by a designed chasing oracle, during which process IPOC produces accurate confidence intervals. The effectiveness of IPOCis theoretically validated through sublinear regret analysis and satisfaction of confidence interval requirements. Besides, we conduct extensive experiments on 4 real-world datasets comparing with 19 baselines. To the best of our knowledge, we are the first to apply the frontier theory of online learning to time series prediction tasks.

Volatility of international crude oil prices is influenced by various external factors on different time scales. User search data (USD) which reflects investor attentions has been widely researched and proved to be associated with crude oil price change at different frequency bands. In this paper, a novel hybrid approach that utilizes bivariate empirical mode decomposition (BEMD) with user search data and machine learning is developed for crude oil price forecasting. First, BEMD is adopted to simultaneously decomposed the crude oil price data and USD into a finite set of components. Second, each component is modelled and predicted by random vector functional link (RVFL) network and the corresponding final results are obtained via an ensemble model. Third, Brent crude oil spot price is used to test the proposed approach empirically. Forecasting results are analyzed with various evaluation criteria and verified robustness. Results show that the proposed approach statistically outperforms traditional forecasting machine learning techniques and similar counterparts (with USD or EMD-based method) in terms of prediction accuracy.

Stock index forecast is a complicated problem since the financial market is influenced by various underlying factors and some of them are unobservable. Hidden Markov Model (HMM) is an effective probabilistic graphic model which can models the hidden states of the observed sequence. So the underlying patterns of the stock movements can be learnt by HMM models. In this paper, HMM is employed to forecast the Shanghai Stock Exchange Composite Index and Shenzhen Stock Exchange Composite Index, which are the two representative stock indices in China. The proposed method is empirically tested on the target data sets and compared with other models based on HMM using Mean Absolute Percentage Error (MAPE). Results show that the proposed HMM model achieves good forecasting performance, and at the same time reveals meaningful hidden states of stock market.

This paper aims at predicting the volume of online movie play on an Video On Demand(VOD) platform using movie characteristics together with related social media data collected from the Internet. An ordinal support vector machine classification approach is employed to distinguish movies playback volume levels. This study collected 1,266 online movie data from year 2013 to 2015, and divided it into three types: high, medium, and low, according to view counts. For each movie, the first two months of video broadcasts are collected from the VOD website. Country, movie type, director, actor, box office are collected from professional movie database. Total number of news and tweets of the movies from Baidu and Sina Weibo are crawled to reflect the publicity level of the movie and the attention of the audience. After feature selection and feature creation, the ordinal support vector machine model is used to predict movie view counts with focus on identifying movies with low play counts and high play counts. The empirical results show that our ordinal support vector machine approach has better out-of-sample prediction accuracy, comparing with standard support vector machine models and other classification models.