In this paper, we describe veDB-HTAP, a highly integrated, efficient, and adaptive HTAP system recently built in ByteDance. veDB-HTAP adopts a highly integrated system architecture by leveraging the Secondary Engine mechanism provided by MySQL and provides a seamless query processing experience across OLTP and OLAP engines. In addition, we introduce a cost-based and machine-learning-based smart query router that significantly outperforms the rule-based query router used in ByteHTAP, a precursor of veDB-HTAP. A key design principle of veDB-HTAP is the collaboration and adaptability of major system components, including query planning, query execution, and unified storage. Our adaptive query execution can be classified into two categories: 1) adaptive execution that dynamically collects and utilizes runtime statistics for better query performance; 2) utilizing runtime resource information to achieve a high quality of service even under heavy workloads. The experiments show that veDB-HTAP can achieve more than 3× speedup for TPC-H while consuming only one-third of the resources compared to ByteHTAP.

In recent years, at ByteDance, we see more and more business scenarios that require performing complex analysis over freshly imported data, together with transaction support and strong data consistency. In this paper, we describe our journey of building ByteHTAP, an HTAP system with high data freshness and strong data consistency. It adopts a separate-engine and shared-storage architecture. Its modular system design fully utilizes an existing ByteDance's OLTP system and an open source OLAP system. This choice saves us a lot of resources and development time and allows easy future extensions such as replacing the query processing engine with other alternatives. ByteHTAP can provide high data freshness with less than one second delay, which enables many new business opportunities for our customers. Customers can also configure different data freshness thresholds based on their business needs. ByteHTAP also provides strong data consistency through global timestamps across its OLTP and OLAP system, which greatly relieves application developers from handling complex data consistency issues by themselves. In addition, we introduce some important performance optimizations to ByteHTAP, such as pushing computations to the storage layer and using delete bitmaps to efficiently handle deletes. Lastly, we will share our lessons and best practices in developing and running ByteHTAP in production.

With the increasing demand for express delivery, a courier needs to deliver many tasks in one day and it's necessary to deliver punctually as the customers expect. At the same time, they want to schedule the delivery tasks to minimize the total time of a courier's one-day delivery, considering the total travel time. However, most of scheduling researches on express delivery focus on inter-city transportation, and they are not suitable for the express delivery to customers in the “last mile”. To solve the issue above, this paper proposes a personalized service for scheduling express delivery, which not only satisfies all the customers' appointment time but also makes the total time minimized. In this service, personalized and accurate travel time estimation is important to guarantee delivery punctuality when delivering shipments. Therefore, the personalized scheduling service is designed to consist of two basic services: (1) personalized travel time estimation service for any path in express delivery using courier trajectories, (2) an express delivery scheduling service considering multiple factors, including customers' appointments, one-day delivery costs, etc., which is based on the accurate travel time estimation provided by the first service. We evaluate our proposed service based on extensive experiments, using GPS trajectories generated by more than 1000 couriers over a period of two months in Beijing. The results demonstrate the effectiveness and efficiency of our method.

This paper presents SMoCoR, a smart mobile contact recommender based on smart phone data. It recommends the most appropriate way to contact friends according to friends' current condition. In no emergency condition, SMoCoR achieves two goals. First, the recommended contacts disturb friends least, that means, it will tell whether it's suitable to call friends. Second, the recommended contact makes the communication information accessible to the friends and gets replies from the friends as soon as possible. For the goals above, SMoCoR takes friends' calendar data and smart phone data as inputs, and after several steps of calculation it will recommend a list of contacts ranked by intelligent algorithm according to the appropriateness. The experimental results based on real-user data show that SMoCoR provides an efficient and accurate means for contact recommendation.

This paper presents SMoCoR, a smart mobile contact recommender based on smart phone data. It recommends the most appropriate way to contact friends according to friends’ current condition. In no emergency condition, SMoCoR achieves two goals. First, the recommended contacts disturb friends least, that means, it will tell whether it’s suitable to call friends. Second, the recommended contact makes the communication information accessible to the friends and gets replies from the friends as soon as possible. For the goals above, SMoCoR takes friends’ calendar data and smart phone data as inputs, and after several steps of calculation it will recommend a list of contacts ranked by intelligent algorithm according to the appropriateness. The recommendation includes two aspects. Firstly, SMoCoR recommends whether it is suitable to make calls, which is named call recommendation. Secondly, SMoCoR recommends a list of the ranked text contacts, which is named text contact recommendation. The experimental results based on real-user data show that SMoCoR provides an effective method for contact recommendation.