- 博客(20)
- 收藏
- 关注
转载 Transaction isolation level
MVCC简介1.1 什么是MVCCMVCC是一种多版本并发控制机制。1.2 MVCC是为了解决什么问题?大多数的MYSQL事务型存储引擎,如,InnoDB,Falcon以及PBXT都不使用一种简单的行锁机制.事实上,他们都和MVCC–多版本并发控制来一起使用.大家都应该知道,锁机制可以控制并发操作,但是其系统开销较大,而MVCC可以在大多数情况下代替行级锁,使用MVCC,能降低其系统...
2019-07-19 10:40:38
120
原创 Spark - HA
Single point of failure recoverybased on FileSystem to recovery. only a master for develop mode.*) record Spark perform information to local system. if master down, to recovery master and read the ...
2019-07-17 17:37:00
103
原创 Spark Debug
https://confluence.jetbrains.com/display/SCA/Scala+Plugin+for+IntelliJ+IDEAhttps://www.jetbrains.com/help/idea/run-debug-and-test-scala.html
2019-07-17 16:50:34
132
原创 Scala Debug
https://www.jetbrains.com/help/idea/run-debug-and-test-scala.html
2019-07-17 16:47:55
539
原创 JAVA Out Of Memory
https://www.cnblogs.com/leodaxin/p/7477437.htmlhttps://www.cnblogs.com/JackDesperado/p/4798499.html
2019-07-17 15:51:53
103
原创 Spark
*Spark componentSQL and DataFramesSpark streamingMLib(machine learning)GraphX
2019-07-17 14:27:15
68
原创 scala高阶函数
#Scala APIhttps://docs.scala-lang.org/common functionmapforeachfilterzippartitionfindflattenflatMap
2019-07-01 23:56:37
69
原创 Scala
Scala Basicinstall and configuration runtime envtype of datavariable and constantfunctioncondition&looparguments of functionlazy valueexceptionarray mapping metagroupSpark Corecl...
2019-06-26 18:47:49
66
原创 Oracle 12c
11g example datahttps://blog.csdn.net/qq_36289559/article/details/88550622https://www.oracle.com/technetwork/database/enterprise-edition/downloads/112010-win64soft-094461.html
2019-06-26 10:22:36
58
原创 HDFS Federation
Cluster FeaturesFail OverLoad BalanceHDFS Federation Targetcache workload too high)promote namenode cache performanceviewFS - 文件视图系统VFS is lay on same hostserver with namenode.≈ F5/proxy...
2019-06-25 12:05:02
121
原创 Zookeeper implement HA for HDFS
HDFS namenode HA(fail over)configuration zookeeper address for hdfs - core-site<property><name>ha.zookeeper.quorum</name><value>hostname:2181</value></p...
2019-06-24 18:39:39
96
原创 Zookeeper
Zookeeper structuredownload resource - http://www.apache.org/dyn/closer.cgi/zookeeper/Rolesleaderleader electionfollowernote: at least, keep 3 zookeeper nodes.ModeStandaloneconfiguratio...
2019-06-24 17:17:16
80
原创 MySQL Installer 8.0.16
Prepare Installationdownload installation package from mysql offical websitehttps://dev.mysql.com/downloads/windows/(1) mysql installation package(2)mysql connector(3)mysql workbenchreq...
2019-06-21 12:16:17
940
原创 Pig notes
Install and configurationdownload installation package from - http://www.apache.org/dyn/closer.cgi/pigtar -zxvf pig-0.17.0.tar.gz -C localpathconfiguration environment variables vi /etc/b...
2019-06-20 16:56:40
148
原创 Hadoop - Proxy user - Superusers Acting On Behalf Of Other Users
http://hadoop.apache.org/docs/r2.7.1/hadoop-project-dist/hadoop-common/Superusers.html
2019-06-20 12:44:48
220
转载 [Hive]The different between Bucket and Partition Table
建立一个分桶表,并尝试直接上传一个数据create table student4(sno int,sname string,sex string,sage int, sdept string) clustered by(sno) into 3 buckets row format delimited fields terminated by ‘,’;set hive.enforce.bucke...
2019-06-19 11:07:01
129
转载 Hive Orderby SortBy DistriuteBy ClusterBy
一:order byorder by会对输入做全局排序,因此只有一个Reducer(多个Reducer无法保证全局有序),然而只有一个Reducer,会导致当输入规模较大时,消耗较长的计算时间。关于order by的详细介绍请参考这篇文章:Hive Order by操作。二:sort bysort by不是全局排序,其在数据进入reducer前完成排序,因此,如果用sort by进行排序,并...
2019-06-19 10:51:23
612
原创 Install Gradle Eclipse plugin
Gradle@AaronInstall Gradle Eclipse pluginInstall package. https://github.com/eclipse/buildship/blob/master/docs/user/Installation.mdchose a version, such as belowinstall manuallyhttps://gradl...
2019-05-24 03:56:14
156
空空如也
空空如也
TA创建的收藏夹 TA关注的收藏夹
TA关注的人