Greenplum 部署最佳实践: 网络篇
1 使用高速硬件，降低节点间的延迟 2 尽量降低网络中断的次数。 3 尽量提高每一个数据包的有效载荷 本文为摘录，原文为： attachments/pdf...
Posts
Greenplum 资源管理——Resource Group使用和实现分析
1 Resource Group 简介 2 Resource Group 使用 2.1 创建资源组 2.2 将资源组分配给角色 2.3 修改资源组配置 2.4 删除资源组 2.5 取消资源组中的正在运行或已排队的事务 3 Resource Group 监控 3.1 查看资源组...
Hash Index of PG
1 Hash Indexing 2 Page Addressing 本文为摘录，原文为： ../../../Work/pg_master/src/backend/access/hash/README 1 Hash Indexing 这个目录包含了 Postgres 的散列索引实现。其中大部分核心思想来自于 Margo Seltzer 和 Ozan Yigit 在 1991 年 1 月举行的冬季 USENIX 会议上的论文...
How To Use Journalctl to View and Manipulate Systemd Logs
1 Basic Log Viewing 2 Journal Filtering by Time 2.1 Displaying Logs from the Current Boot 2.2 Filtering by Message Interest 本文为摘录，原文为： https://www.digitalocean.com/community/tutorials/how-to-use-journalctl-to-view-and-manipulate-systemd-logs 1 Basic Log Viewing 要查看 journald 守护程序收集的日志，请使用 journalctl 命令。 当单独使用时，系统中的每个...
Hybrid Blockchain Database Systems: Design and Performance
1 ABSTRACT 2 INTRODUCTION 本文为摘录，原文为： attachments/pdf/e/p1092-loghin.pdf 1 ABSTRACT Abbrs: CFT: crash fault-tolerant BFT: byzantine fault-tolerant Byzantine Fault Tolerance (BFT) is a trait of decentralized, permissionless systems which are capable of successfully identifying and rejecting dishonest or faulty information. Byzantine fault tolerant systems have successfully solved the Byzantine Generals Problem and are robust against sybil attacks. 2 INTRODUCTION 学术界出...
iReader & Calibre
1 表现 2 原因 3 绕过方法 1 表现 通过 Calibre 向 iReader 中传输书籍失败，报 MTP 相关错误。 2 原因 不知道为什么，无法在设备存储空间中创建目录 3 绕过方法 用文件管理器提...
Linux Process States
1 The Linux Process States 1.1 Running or Runnable State (R) 1.2 Sleeping State: Interruptible (S) and Uninterruptible (D) 1.3 Stopped State (T) 1.4 Zombie State (Z) 2 Checking Process State 2.1 Displaying Process State Using ps 2.2 Using the top Command 2.3 The /proc Pseudo File 3 Summary 本文为摘录，原文为： https://www.baeldung.com/linux/process-states 1 The Linux Process States Running or Runnable (R) Uninterruptible...
Making RDBMSs Efficient on Graph Workloads Through Predefined Joins
1 ABSTRACT 本文为摘录，原文为： attachments/pdf/4/p1011-jin.pdf 1 ABSTRACT
Manu: A Cloud Native Vector Database Management System
1 ABSTRACT 2 INTRODUCTION 本文为摘录，原文为： attachments/pdf/8/p3548-yan.pdf 1 ABSTRACT 下一代向量数据库应该有的特性： 可长期演进 tunable consistency good elasticity high performance Manu 通过放宽数据模型和一致性约束来换取上述特性： 使用 WAL 和...
NBTree: a Lock-free PM-friendly Persistent B+-Tree for eADR-enabled PM Systems
本文为摘录，原文为： attachments/pdf/a/p1187-zhang.pdf PM stands for Persistent Memory， 持久内存，它是一种非易失性内存，即使断电也能保留数据。 eADR 保证了 CPU 缓存中的数据在电源故障时会被刷新...
New Wine in an Old Bottle: Data-Aware Hash Functions for Bloom Filter
1 INTRODUCTION 本文为摘录，原文为： attachments/pdf/4/p1924-bhattacharya.pdf FPR: Fault Positive Rate Partitioned Bloom Filter , 方法 将 BloomFilter 划分成 segments 每个 segment 使用简单的、基于投影的哈希函数， 函数通过数据计算得来 效果 减少误报 （两个数量级...
OceanBase 707M tpmC
1 Introduction 2 DESIGN OVERVIEW 2.1 Goals 2.2 Criteria of Design 3 架构图 3.1 Application Layer 3.2 Proxy Layer 3.3 Data Service Layer 3.4 Zones 3.5 表 3.6 Node 3.7 SQL Engine 3.8 Multi-tenancy 3.9 Resource Isolation. 3.10 Features 4 STORAGE ENGINE 4.1 LSM Tree-Based Architecture 4.2 Asymmetric Read and Write 4.3 Daily Incremental Major Compaction 4.4 Replica Type 5 TRANSACTION PROCESSING ENGINE 5.1 Partition and Paxos Group...
Origami: A High-Performance Mergesort Framework
1 ABSTRACT 2 INTRODUCTION 3 PIPELINE OVERVIEW 3.1 notation 3.2 better cache utilization, 3.3 4 Phases 本文为摘录，原文为： attachments/pdf/c/p259-arman.pdf 1 ABSTRACT Mergesort 优点： 不受数据倾斜影响 适于通过向量化执行来进行并行处理 适于多线程操作操作 Origami: 内存归...
Performance of 10 G Ethernet Using Commodity Hardware
1 INTRODUCTION 2 NETWORK PROCESSING IN THE LINUX KERNEL 2.1 Kernel Accounting 2.2 Packet Transmission 发送 2.3 Packet Reception 接收 2.4 Incorrect SoftIRQ Accounting 3 TESTBED AND KERNEL CONFIGURATION 4 UDP TRANSMISSION 5 TCP TRANSMISSION 6 TCP TRANSMISSION WITH ZERO-COPY 7 THE HARDWARE OFFLOAD FUNCTIONALITIES OF THE NETWORK ADAPTER 8 FIBRE CHANNEL TO 10 GIGABIT ETHERNET TESTS 9 CONCLUSIONS 本文为摘录，原文为...
Performance Tip of the Week
1 #9: Optimizations past their prime 1.1 Best practices 1.2 #21: Improving the efficiency of your regular expressions 本文为摘录，原文为： https://abseil.io/fast 1 #9: Optimizations past their prime 1.1 Best practices – 可能的情况下编写清晰、惯用的代码 Prefer writing clear, idiomatic code whenever possible. It is not only easier to read and...
pgaudit/pgaudit: PostgreSQL Audit Extension
1 Introduction 2 Why pgAudit 3 Usage Considerations 4 Settings 4.1 pgaudit.log 4.2 pgaudit.log_catalog 5 Session Audit Logging 6 Object Audit Logging 7 Format 本文为摘录，原文为： https://github.com/pgaudit/pgaudit#readme 1 Introduction PostgreSQL 审计扩展（pgAudit）通过标准的 PostgreSQL 日志记录功能提供详细的会话...
Prefix Filter: Better Than Bloom
1 Abstract 2 INTRODUCTION 本文为摘录，原文为： ../pdf/d/p1311-even.pdf 1 Abstract 2 INTRODUCTION What is a filter 属于近似查询 (approximate membership query), 用于判断独具是否在集合中。 如果输入在集合中，则必定输出 Yes 如果不在集合中，则输...
QueryFormer: A Tree Transformer Model for Query Plan Representation
1 INTRODUCTION 本文为摘录，原文为： attachments/pdf/0/p1658-zhao.pdf QueryFormer: learning-based query planer representation model with tree-structured Transformer architecture -integrate histograms from database into query plan encoding 1 INTRODUCTION Physical Query Plan As DAG (Directed Acyclic Graph) Figure 1: Example query and query plan node 表示操作 edge 表示方向 子节点先执行，执行结果给...
Ranked Enumeration of Join Queries with Projections
1 ABSTRACT 本文为摘录，原文为： attachments/pdf/6/p1024-deep.pdf 1 ABSTRACT
Red Hat Enterprise Linux 7 Performance Tuning Guide
1 TOOLS 1.1 top 1.2 ps 1.3 vmstat – Virtual Memory Statistics 1.4 sar – System Activity Repoter 1.5 perf 1.6 tubostat 1.7 iostat 1.8 irqbalance 1.9 ss – socket statistics 1.10 numastat 1.11 numad 本文为摘录，原文为： attachments/pdf/c/red_hat_enterprise_linux-7-performance_tuning_guide-en-us.pdf 1 TOOLS 1.1 top 1.2 ps 1.3 vmstat – Virtual Memory Statistics 1.4 sar – System Activity Repoter 1.5 perf 1.6 tubostat 1.7 iostat...