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.

The ubiquity of variable-length integers in data storage and communication necessitates efficient decoding techniques. In this paper, we present SFVInt, a simple and fast approach to decode the prevalent Little Endian Base-128 (LEB128) varints. Our approach effectively utilizes the Bit Manipulation Instruction Set 2 (BMI2) in modern Intel and AMD processors, achieving significant performance improvement while maintaining simplicity and avoiding overengineering. SFVInt, with its generic design, effectively processes both 32-bit and 64-bit unsigned integers using a unified code template, marking a significant leap forward in varint decoding efficiency. We thoroughly evaluate SFVInt's performance across various datasets and scenarios, demonstrating that it achieves up to a 2x increase in decoding speed when compared to varint decoding methods used in established frameworks like Facebook Folly and Google Protobuf.

Shared-nothing, distributed databases scale transactional and analytical processing over a large data volume by spreading data across servers. However, static sharding of data across nodes makes such systems fail to timely adapt to changing workloads and struggle to obey the cloud pay-as-you-go model. Migrating shards between nodes online is a key technique to react to dynamic changes of workloads for cloud elasticity. Existing approaches introduce severely degraded performance and service interruption, resulting in SLA violation on the cloud; or they are tailor-made to deterministic databases. In this paper, we propose Remus, a new live migration approach for shared-nothing, distributed databases with snapshot isolation. Remus migrates shards between nodes with zero service interruption and minimal performance impact. This is achieved by an efficient unidirectional dual execution during migration. We implement Remus on a shared-nothing, distributed version of PolarDB-PG and evaluate it against state-of-the-art approaches using standard OLTP workloads TPC-C and YCSB, and hybrid workloads consisting of long-lived and short transactions. The results demonstrate Remus is the only effective approach to achieve the goal of zero transaction interruption, zero downtime and marginal performance impact, paving the way for applying the shared-nothing architecture to a cloud database which needs to provide elasticity while guaranteeing strict SLAs.

Modern distributed database systems scale horizontally by partitioning their data across a large number of nodes. Most such systems build their transactional layers on a replication layer, employing a consensus protocol to ensure data consistency to achieve fault tolerance. Synchronization among replicated state machines thus becomes a significant overhead of transaction processing. Without careful design, synchronization could amplify transactions' lock duration and impair the system's scalability. Speculative techniques, such as Controlled Lock Violation (CLV) and Early Lock Release (ELR), prove useful in shortening lock's critical path and boosting transaction processing performance. To use these techniques to optimize geo-replicated distributed databases(GDDB) is an intuitive idea. This paper shows that a naive application of speculation is often unhelpful in a distributed environment. Instead, we introduce Distributed Lock Violation (DLV), a specialized speculative technique for geo-replicated distributed databases. DLV can achieve good performance without incurring severe side effects.

Huawei Fusion Insight Libr A (FI-MPPDB) is a petabyte scale enterprise analytics platform developed by the Huawei data-base group. It started as a prototype more than five years ago, and is now being used by many enterprise customers over the globe, including some of the world's largest financial institutions. Our product direction and enhancements have been mainly driven by customer requirements in the fast evolving Chinese market. This paper describes the architecture of FI-MPPDB and some of its major enhancements. In particular, we focus on top four requirements from our customers related to data analytics on the cloud: system availability, auto tuning, query over heterogeneous data models on the cloud, and the ability to utilize powerful modern hardware for good performance. We present our latest advancements in the above areas including online expansion, auto tuning in query optimizer, SQL on HDFS, and intelligent JIT compiled execution. Finally, we present some experimental results to demonstrate the effectiveness of these technologies.