System

进度

笔记

基础知识

CQL

不支持join,like,非主键查询，支持需要allowing filter，二级索引，或者SASI，物化视图。

SASI：LIKE，SPARSE INDEX。 memory mapped B+ trees，不支持集合索引，普通二级索引支持。

函数：ttl，writetime，token

Clustered Index and Secondary Indexes：聚簇索引只有一个，优先primary key 没有就是unique key，两个都没有，innodb自动生成。看来唯一索引也是非聚簇的，唯一索引也能建多个，是非聚簇的，也能为空，能多个都是空，但是不能重复。 https://dev.mysql.com/doc/refman/5.7/en/innodb-index-types.html

数据建模：Chebotko method，query driven，primary key=partation key + clustering key，把表想象成一个分区表。【https://cassandra.apache.org/doc/latest/cassandra/data_modeling/index.html】

write-once immutable ordered data model

Release Note: https://gitbox.apache.org/repos/asf?p=cassandra.git;a=blob_plain;f=CHANGES.txt;hb=refs/tags/cassandra-4.0.0

cassandra是宽列行存储引擎。数据模型是列存（Wide Column Stores ），但是磁盘是行存。hbase和bigtable也是。面向列的磁盘存储一般在数仓较多，如clickhouse，druid，greenplum，还有关系型分为行存和列存。

https://en.wikipedia.org/wiki/List_of_column-oriented_DBMSes

Apache Parquet is an open source, column-oriented data file format designed for efficient data storage and retrieval

TiDB（行存） TiFlash（列存）

OLTP（行存） OLAP（列存）

先读bigtable(2006)，dynamo(2007)，cassandra(2009)的paper.

范式：one table serve many queries 反范式：one table serve one query

数据建模cassandra风格: 数据查询在一起，存储应该在一起，保持分区和结果集小，使用高基数的key,表是为了查询而建的，没有全表扫描

https://en.wikipedia.org/wiki/Apache_Cassandra

read path和write path发生了什么？

compaction类型minor，major，user defined，scrub 和策略：STCS，LCS，TWCS

多维hashmap

读取很快，简单的查询模型，数据在一起查询那么应该存在一起，使用高基数的key，

数据时间 ttl

cassandra datastax driver三种模式：statement，mapper，accessor + spring data cassandra

column families 也叫 tables

副本修复方式：read repair,hinted handoff

计数器可以有多个counter，但是不能出现非主键key。

Cassandra uses a Last-Write-Wins Element-Set conflict-free replicated data type for each CQL row。

SSTable Attached Secondary Index (SASI): LIKE

CODD 12条原则，5大范式。

除了主键外的每一列都有时间戳，那么设置更新时间就没有意义了，关系型每一个是一个完整的时间。

运维

compaction

gc_grace_seconds=10d

hints

max_hint_windowin_ms=3h

百科

workshop

https://github.com/DataStax-Academy/cassandra-workshop-series

person

[Avinash Lakshman]

[Prashant Malik]

Eben Hewitt

Jeffrey Carpenter Jeffrey Carpenter

Site

https://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-model-general.html

https://cwiki.apache.org/confluence/display/cassandra

Common-Error

Query

[Invalid query] message="ORDER BY is only supported when the partition key is restricted by an EQ or an IN."。

order by 操作必须在分区key是 == 或者 in的时候下才可以使用。不能排序分区键，因为需要遍历所有分片。
[Invalid query] message="Order by currently only support the ordering of columns following their declared order inthe PRIMARY KEY"。

orderby必须是主键指定的顺序。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot rename non PRIMARY KEY part createtime"。 不能重命名非PRIMARY KEY。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Invalid operation (studytimechange =studytimechange + 1000) for non counter column studytimechange" 累加操作必须在counter类型字段上面。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot mix counter and non counter columns in the same table"。 计数器列可以多个，但是必须非主键之外所有键。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Order by currently only support the ordering of columns following their declared order in the PRIMARY KEY" 排序字段只能是PRIMARY KEY定义的顺序。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot execute this query as it might involve data filtering and thus may have unpredictable performance. If you want to execute this query despite the performance unpredictability, use ALLOW FILTERING"。 原来查询：select * from answers_by_user where courseid=1; 现有查询：select * from answers_by_user where courseid=1 allow filtering; 其他方法：change your data model add an index, use another table use ALLOW FILTERING
InvalidRequest: Error from server: code=2200 [Invalid query] message="User-defined functions are disabled in cassandra.yaml - set enable_user_defined_functions=true to enable" UDF没有打开。UDF没有打开，那么UDA也用不了。因为UDA借助两个UDF。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Invalid non-frozen user-defined type for PRIMARY KEY component address"。 非冻结的UDT不能作为主键的一部分，而冻结的UDT可以作为主键的一部分。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Unsupported order by relation"。 orderby必须是集群键指定的顺序，不可以相交。建表默认是a asc，b asc，order by b asc,c desc; order by b desc,c asc; 均会失败，order by a asc，b asc 或 a desc，b desc均会成功。
InvalidRequest: Error from server: code=2200 [Invalid query] message="INSERT statements are not allowed on counter tables, use UPDATE instead" counter类型不能直接插入，需要直接UPDATE。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot directly modify a materialized view" 不能insert，delete物化视图，物化视图修改只能由基础表来派生。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot TRUNCATE materialized view directly; must truncate base table instead" 不能TRUNCATE物化视图。
InvalidRequest: Error from server: code=2200 [Invalid query] message="Cannot execute this query as it might involve data filtering and thus may have unpredictable performance. If you want to execute this query despite the performance unpredictability, use ALLOW FILTERING" 在执行((a,b),c)建表索引时候，查询条件只带了where a=xxx，会导致这个报错，因为分区key没有完整，应该是where a=xxx and b=xxx。。
副本因子只能配置在keyspace级别，也即复制策略在一个keyspace级别，或者一个数据中心级别，不能配置在table级别。
如果索引是((a,b),c)，那么查询至少是where a=xxx and b=yyy，而不能直接a=xxx。
如果索引是(a,b,c)，那么默认第一个是分区键，b，c是排序键。

开源项目的代码案例研究

cassandra-workshop-series

https://github.com/DataStax-Academy/cassandra-workshop-series

killrvideo-java

https://github.com/KillrVideo/killrvideo-java

主要是core和mapping，GRPC，大量CF，

workshop-spring-data-cassandra

https://github.com/datastaxdevs/workshop-spring-data-cassandra

workshop-ecommerce-app

https://github.com/datastaxdevs/workshop-ecommerce-app

基于SpringDataCassandra写的。

DataStax-Examples

https://github.com/DataStax-Examples/

cassandra-guide

https://github.com/jeffreyscarpenter/cassandra-guide

基于core写的。没有mapper。

参考

dbdb.io: https://dbdb.io/db/cassandra

https://github.com/xiaozhiliaoo/cassandra-practice

java客户端：https://docs.datastax.com/en/developer/java-driver/4.13/

spring-data-cassandra: https://docs.spring.io/spring-data/cassandra/docs/current/reference/html/#preface

scylladb:C++写的cassandra. https://www.scylladb.com/resources/introduction-to-apache-cassandra/

术语：https://docs.datastax.com/en/glossary/doc/glossary/glossaryTOC.html

https://cwiki.apache.org/confluence/display/CASSANDRA2/Home

https://cwiki.apache.org/confluence/display/CASSANDRA/Home

Merkle tree：http://people.eecs.berkeley.edu/~raluca/cs261-f15/Merkle.html

https://en.wikipedia.org/wiki/Ralph_Merkle https://ralphmerkle.com/

静态列的值所有行都一样，最新的会覆盖其他列的。

paper

Backup and Recovery Mechanisms of Cassandra Database: A Review

Gossip : http://awinterman.github.io/simple-scuttle/

Facebook’s Cassandra paper, annotated and compared to Apache Cassandra 2.0 : https://docs.datastax.com/en/articles/cassandra/cassandrathenandnow.html

新的机制。

读书笔记