《Database Internals: A Deep-Dive into How Distributed Data System Work》读书笔记

• Overview

  Terminology can sometimes be ambiguous and hard to understand without a complete context.

  Distinction among the DBMS in terms of :

  • a storage medium:

    1. Memory

       In-memory DBMS (main memory DBMS) store data primarily in memory and use the disk for recovery and logging.

       Store their contents almost exclusively in RAM

    2. Disk based DBMS

       Disk based DBMS hold most of the data on disk and use memory for caching disk contents or as a temporary storage.

  • layout: Column vs. Row Oriented DBMS

  • three major categories:

    1. Online transaction processing databases (OLTP)
    2. Online analytical processing databases (OLAP)
    3. Hybrid transactional and analytical processing (HTAP)

  • another way:

    1. key-value stores
    2. relational databases
    3. document-oriented stores
    4. graph databases

• Client/Server

  Database management systems use a client/server model, where database system instances (nodes) take the role of servers, and application instances take the role of clients.

  Client requests arrive through the transpot subsystem.

  

  The storage engine has several components with dedicated responsibilities:

  1. Transaction manager
  2. Lock manager
  3. Access methods
  4. Buffer manager
  5. Recovery manager

• Horizontal Scaling vs. Scaling Vertically

  Scaling out: improving performance and increasing capacity by running multiple database instances acting as a single logical unit: Gamma Database Machine Project, Teradata, Greenplum, Parallel DB2.

  Scaling up by moving the database to a larger, more powerful machine.

• Database components

  1. A transport layer accepting request

  2. A query processor determing the most efficient way to run queries

  3. An execution engine carrying out the operations

  4. A storage engine(or databasa engine)

     A software component of a database management system responsible for storing, retrieving, and amanging data in memory and on disk, designed to capture a persistent, long-term memory of each node.

  Database management systems are applications built on top of storage engines, offering a schema, a query language, indexing, t