zookeeper hadoop-ha znode错误导致ZKFailoverController无法启动

最新推荐文章于 2024-06-12 10:27:23 发布

原创最新推荐文章于 2024-06-12 10:27:23 发布 · 6k 阅读

4 ·

CC 4.0 BY-SA版权

文章标签：

#ambari #hdp #zkfc

本文介绍了解决Ambari启动ZKFailoverController失败的问题，通过日志定位到Parent znodedoes not exist错误，并给出了使用hdfszkfc-formatZK命令初始化ZooKeeper的具体步骤。

Ambari启动ZKFailoverController失败了，输入tail -f -n 100 hadoop-hdfs-zkfc-master.log

日志中显示：

org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller. Parent znode does not exist.
Run with -formatZK flag to initialize ZooKeeper.

按照提示，在失败的节点上输入:

hdfs zkfc -formatZK

重启zkfc就行了。

参考https://community.hortonworks.com/questions/50078/unable-to-start-zkfailovercontroller.html

确定要放弃本次机会？

福利倒计时

: :

立减 ¥

普通VIP年卡可用

立即使用

wszhtc

关注关注

0
点赞
踩
4

收藏

觉得还不错? 一键收藏
0
评论
分享

复制链接

分享到 QQ

分享到新浪微博

扫一扫
举报

举报

HDFS failover启动错误Parent znode does not exist

数据源的港湾

10-24

872

现象 HDFS配置了高可用，NameNode和DataNode状态看起来均正常，但两个failover进程异常。查看failover日志，报错信息如下， tail -100f /var/log/hadoop-hdfs/hadoop-cmf-hdfs-FAILOVERCONTROLLER-hxpernode01.esgyn.cn.log.out 2019-10-24 19:51:13,396 F...

Hadoop-HA高可用分布式集群总结

Biubiubiu!

03-14

1355

觉得有帮助的，请多多支持博主，点赞关注哦~ 文章目录Hadoop-HA高可用分布式集群一、HA概述二、HDFS-HA 工作机制1、HDFS-HA 工作要点2、HDFS-HA 自动故障转移工作机制三、总体配置及准备规划1、集群规划2、准备工作3、修改配置文件四、HDFS-HA 集群配置1、配置HDFS-HA集群2、配置 HDFS-HA 自动故障转移五、YARN-HA 配置1、YARN-HA 工作机制...

参与评论您还未登录，请先登录后发表或查看评论

ERROR org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller.

weixin_41877582的博客

11-04

2407

用jps查看java进程，发现没有DFSZKFailoverController进程格式化zk,在zookeeper上创建hadoop-ha节点（就是目录） hdfs zkfc -formatZK 启动zkfc（在所有namenode节点上） hadoop-daemon.sh start zkfc 刷新namenode，会自动选出active hdfs haadmin -failover nn1 nn2 https://blog.csdn.net/houzhizhen/article/details.

启动ZKFC时报错ERROR org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller. Paren

qq_53663722的博客

08-31

1043

执行以下命令之前需要先把所有机器的zookeeper启动起来。按照报错日志提示在该机器上执行命令，

ha.ZKFailoverController: Unable to start failover controller. Unable to connect to ZooKeeper quorum

zhglhy的专栏

06-02

1351

hadoop HA

HaHadoop 启动时：zk无法启动两个节点都是standby：Unable to start failover controller. Parent znode does not exist.

ShiningBlueOcean

03-20

865

解决：在namenode节点上重新格式化zk hdfs zkfc -formatZK 在两个namenode节点重新启动zk hadoop-daemon.sh start zkfc

25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:java.io.tmpdir=/tmp 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:java.compiler=<NA> 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:os.name=Linux 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:os.arch=amd64 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:os.version=3.10.0-1160.el7.x86_64 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:user.name=root 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:user.home=/root 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Client environment:user.dir=/usr/local/hadoop/sbin 25/12/02 16:37:59 INFO zookeeper.ZooKeeper: Initiating client connection, connectString=abc:2181,bcd:2181,cde:2181 sessionTimeout=5000 watcher=org.apache.hadoop.ha.ActiveStandbyElector$WatcherWithClientRef@478190fc 25/12/02 16:37:59 INFO zookeeper.ClientCnxn: Opening socket connection to server bcd/192.168.177.131:2181. Will not attempt to authenticate using SASL (unknown error) 25/12/02 16:37:59 INFO zookeeper.ClientCnxn: Socket connection established to bcd/192.168.177.131:2181, initiating session 25/12/02 16:37:59 INFO zookeeper.ClientCnxn: Session establishment complete on server bcd/192.168.177.131:2181, sessionid = 0x2000023ebf90000, negotiated timeout = 5000 =============================================== The configured parent znode /hadoop-ha/ns already exists. Are you sure you want to clear all failover information from ZooKeeper? WARNING: Before proceeding, ensure that all HDFS services and failover controllers are stopped! =============================================== Proceed formatting /hadoop-ha/ns? (Y or N) 25/12/02 16:37:59 INFO ha.ActiveStandbyElector: Session connected. y 25/12/02 16:38:07 INFO ha.ActiveStandbyElector: Recursively deleting /hadoop-ha/ns from ZK... 25/12/02 16:38:07 INFO ha.ActiveStandbyElector: Successfully deleted /hadoop-ha/ns from ZK. 25/12/02 16:38:07 INFO ha.ActiveStandbyElector: Successfully created /hadoop-ha/ns in ZK. 25/12/02 16:38:07 INFO zookeeper.ZooKeeper: Session: 0x2000023ebf90000 closed 25/12/02 16:38:07 INFO zookeeper.ClientCnxn: EventThread shut down

热门推荐

辰辰爸的博客

12-14

1万+

一、简介 DFSZKFailoverController是Hadoop-2.7.0中HDFS NameNode HA实现的中心组件，它负责整体的故障转移控制等。它是一个守护进程，通过main()方法启动，继承自ZKFailoverController。二、实现流程 1、启动通过main()方法启动，如下： /** * 进程启动的main()方法 */

ZKFC报错：ZK Failover Controller failed: Received stat error from Zookeeper. code:CONNECTIONLOSS

u010410697的博客

07-06

1934

Namenode HA模式下，一个节点掉了，另外的节点不能正常被切换成Active，看zkfc的日志发现出现zookeeper连接超时的异常。 2020-07-04 10:32:06,356 INFO org.apache.zookeeper.ClientCnxn: Unable to read additional data from server sessionid 0x1019acb4933001f, likely server has closed socket, closing socket co

Failover Controller 启动失败

szh_61的博客

08-14

2842

平台：CDH 日志报错： FATAL org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller. Parent znode does not exist. 网上建议方法：hdfs zkfc -formatZK，参考链接：https://community.hortonworks.com/con...

Unable to start failover controller. Parent znode does not exist.Run with -formatZK flag to initial

前方的路在刚开始

06-12

375

CDH出现了这个问题，hdfs报错因为我之前重新安装了zookeeper，导致zookeeper中的namenode信息丢失。然后之前的name node 挂了，看角色日志是这个问题。

hadoop HA异常：Unable to start failover controller. Parent znode does not exist.

B11050101的博客

12-03

2934

在namenode上执行./hadoop-daemon.sh start zkfc时，发现zkfc进程未启动，且报出如下错误：很明显，是由于Zookeeper上不存在namenode的临时节点。解决方法： 1、登录Zookeeper，执行./zkCli.sh进入Zookeeper终端，执行如下命令： ls / 结果发现，果然没有hadoop-ha节点。 2、登录namenode节点，并执行...

hadoop -Unable to start failover controller. Parent znode does not exist

u012581453的专栏

07-11

1063

如上图，k8s pod 起来又crash 掉，然后多次 kubectl delete pod +pod-name 重新拉起还是 crash 掉，查看日志。今天使用星环的TDH集群时，HDFS服务宕掉，在后台查看namenode 始终起不来。从日志中可知道是由于Zookeeper上不存在namenode的临时节点。干掉k8s中的namenod 的pod ，然后重新拉起，就可以了。进入可以看到没有format的ha 信息。登录zookeeper 的客户端。

zookeeper报错：Unable to start failover controller. Parent znode does not exist. Run with -formatZK fla

a13568hki的博客

12-15

1216

报错原因：我在做phoenix upsert 操作时，误操作，删除了来源表hbase的底层hdfs的数据，导致正在执行的phoenix任务挂掉，整个集群挂掉（由于hdfs的挂掉，导致的集群挂掉），如图,我已解决：这里Failover Controller挂掉，看日志，发现是zookeeper一直无法正常启动，报错如下：解决：格式化zookeeper hdfs zkfc -formatZK 分析：由于误删除了底层文件，导致正在被操作的hbase表发生变化，但是zookeeper已经记录了一部分元数据，

zkfc的znode不存在的问题

adzuxhs5024的博客

10-08

566

cd /soft/hadoop/logs/hadoop-centos-zkfc-s101.log发现： 2018-09-29 12:42:03,616 FATAL org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller. Parent znode does not exist. Run ...

启动Hadoop时DFSZKFailoverController没有启动

weixin_30394669的博客

08-10

2356

在启动Hadoop成功后，并没有报错信息，jps查看进程，发现DFSZKFailoverController没有启动成功，后来发现是因为防火墙的原因，关掉重试就OK了 systemctl stop firewalld 转载于:https://www.cnblogs.com/EnzoDin/p/7342000.html...

零基础搭建hadoop-zookeeper-hbase-hive集群全攻略

"本文档主要介绍了如何从零开始安装配置hadoop、zookeeper、hbase和hive集群，包括各个组件的安装、配置、测试以及常见问题的处理。文档中提供了详细的步骤和链接，适用于初学者按照步骤操作搭建集群。" ### 一、...