https://x.medemede.cn/categories/hadoopMedeMedeHalo v2.24.2zh-cnhttps://i.loli.net/2020/11/12/o6rsqgEJ78KtdhI.jpghttps://x.medemede.cn/categories/hadoopWed, 20 May 2026 21:35:05 GMThttps://x.medemede.cn/archives/haddophostname%E4%B8%8D%E8%83%BD%E6%9C%89%E7%89%B9%E6%AE%8A%E5%AD%97%E7%AC%A6]]>/archives/haddophostname%E4%B8%8D%E8%83%BD%E6%9C%89%E7%89%B9%E6%AE%8A%E5%AD%97%E7%AC%A6Saber踩坑LinuxHadoopSun, 29 Mar 2020 15:31:16 GMThttps://x.medemede.cn/archives/docker%E5%AE%B9%E5%99%A8ssh%E4%BA%92%E8%81%94准备 - Container 1 : Ubuntu 18.04, Name:hadoop01 - Container 2 : Ubuntu 18.04, Name:hadoop02]]>/archives/docker%E5%AE%B9%E5%99%A8ssh%E4%BA%92%E8%81%94SaberLinuxHadoopWed, 25 Mar 2020 05:46:11 GMThttps://x.medemede.cn/archives/1550166403289配置Yarn在配置完成NameNodeHA的基础上，只需再配置一下yarn-site.xml即可。&]]>/archives/1550166403289SaberJAVAHadoopThu, 14 Feb 2019 17:54:00 GMThttps://x.medemede.cn/archives/1550159816422先配置好Haoop集群配置ZookeeperHaoopHA的自动故障转移需要Zookeeper的支持]]>/archives/1550159816422SaberJAVALinuxHadoopThu, 14 Feb 2019 16:15:00 GMThttps://x.medemede.cn/archives/1550154253961ZooKeeper的操作较为简单，下面是一个简单的客户端操作实例客户端监听服务器上下线的简单实现：G]]>/archives/1550154253961SaberJAVAHadoopThu, 14 Feb 2019 14:32:00 GMThttps://x.medemede.cn/archives/1549878225530集群规划在hadoop101、hadoop102、hadoop103三个节点上部署Zookeeper]]>/archives/1549878225530SaberLinuxHadoopMon, 11 Feb 2019 09:55:00 GMThttps://x.medemede.cn/archives/15497125482651.Writable序列化序列化就是把内存中的对象，转换成字节序列（或其他数据传输协议）以便于存储（]]>/archives/1549712548265SaberJAVAHadoopSat, 9 Feb 2019 12:24:00 GMThttps://x.medemede.cn/archives/1549683494970案例：MapReduceDemoREADME中给出了对应知识点Mapper阶段用户自定义的Mappe]]>/archives/1549683494970SaberJAVAHadoopSat, 9 Feb 2019 03:38:00 GMThttps://x.medemede.cn/archives/1549327361109默认回收站默认值fs.trash.interval=0，0表示禁用回收站，可以设置删除文件的存活时间]]>/archives/1549327361109SaberJAVAHadoopTue, 5 Feb 2019 00:42:45 GMThttps://x.medemede.cn/archives/1549326430317快照相当于对目录做一个备份。并不会立即复制所有文件，而是指向同一个文件。当写入发生时，才会产生新文件]]>/archives/1549326430317SaberJAVAHadoopTue, 5 Feb 2019 00:36:24 GMThttps://x.medemede.cn/archives/1549326327878理论概述每个文件均按块存储，每个块的元数据存储在namenode的内存中，因此hadoop存储小文件]]>/archives/1549326327878SaberJAVAHadoopTue, 5 Feb 2019 00:25:00 GMThttps://x.medemede.cn/archives/1549326096765scp实现两个远程主机之间的文件复制scp-rhello.txtroot@hadoop103:/he]]>/archives/1549326096765SaberJAVAHadoopTue, 5 Feb 2019 00:21:44 GMThttps://x.medemede.cn/archives/1549267021993和namenode不同的是，datanode的存储目录是初始阶段自动创建的，不需要额外格式化。查看D]]>/archives/1549267021993SaberJAVAHadoopMon, 4 Feb 2019 07:57:34 GMThttps://x.medemede.cn/archives/1549266291195工作机制一个数据块在datanode上以文件形式存储在磁盘上，包括两个文件，一个是数据本身，一个是元]]>/archives/1549266291195SaberJAVAHadoopMon, 4 Feb 2019 07:50:06 GMThttps://x.medemede.cn/archives/1549249276517服役新数据节点环境准备克隆一台虚拟机修改ip地址和主机名称修改xcall和xsync文件，增加新增节]]>/archives/1549249276517SaberJAVAHadoopMon, 4 Feb 2019 03:01:00 GMThttps://x.medemede.cn/archives/1549247861909查看NameNode版本号在{HADOOP_HOME}/data/tmp/dfs/name/curr]]>/archives/1549247861909SaberJAVAHadoopMon, 4 Feb 2019 02:37:48 GMThttps://x.medemede.cn/archives/1549196656846Namenode启动时，首先将映像文件（fsimage）载入内存，并执行编辑日志（edits）中的各]]>/archives/1549196656846SaberJAVAHadoopSun, 3 Feb 2019 12:28:11 GMThttps://x.medemede.cn/archives/1549194015178SecondaryNameNode用来监控HDFS状态的辅助后台程序，每隔一段时间获取HDFS元数据]]>/archives/1549194015178SaberJAVAHadoopSun, 3 Feb 2019 11:53:00 GMThttps://x.medemede.cn/archives/1549186374579fsimage&editsNameNode被格式化之后，将在{Hadoop_HOME}/da]]>/archives/1549186374579SaberJAVAHadoopSun, 3 Feb 2019 09:41:00 GMThttps://x.medemede.cn/archives/1549185806501第一阶段：namenode启动1.第一次启动namenode格式化后，创建fsimage和edits]]>/archives/1549185806501SaberJAVAHadoopSun, 3 Feb 2019 09:29:00 GMT