Support Questions

Find answers, ask questions, and share your expertise
Announcements
Celebrating as our community reaches 100,000 members! Thank you!

Not able to start zookeeper after CDH 4.1 Installation

avatar
Expert Contributor

I created 3 VMs called nn1, nn2, and jt1 and assigned static IP addresses to each. I ensured that all VMs are able to ping one another.

I also edited the hosts file on each VM to that nn1, nn2 and jt1 have entries against their IP addresses.

 

Then I installed jdk-6u45-linux-x64-rpm.bin on each of the VMs.

 

Next i executed this command on all 3 VMs

 

wget http://
archive.cloudera.com/cdh4/redhat/6/x86_64/cdh/cloudera-cdh4.repo
cp cloudera-cdh4.repo /etc/yum.repos.d/
rpm --import http://archive.cloudera.com/cdh4/redhat/6/x86_64/cdh/RPM-
GPG-KEY-cloudera

 

Then I installed the following on all 3 nodes

 

yum install hadoop-hdfs-namenode
yum install hadoop-hdfs-journalnode
yum install zookeeper-server

 

I installed this only on nn1 and nn2

 

yum install hadoop-hdfs-zkfc

 


Finally I edit the /etc/zookeper/conf/zoo.cfg and added 3 lines (on all 3 machines)

 

server.1=nn1:2888:3888
server.2=nn2:2888:3888
server.3=jt1:2888:3888

 

I also executed this command on each machine

 

firewall-cmd --zone=dmz --add-port=2888/tcp --permanent
firewall-cmd --zone=dmz --add-port=3888/tcp --permanent
firewall-cmd --reload

 

But I keep seeing these errors in the /var/log/zookeeper/zookeeper.log file of each machine

 

WARN [WorkerSender[myid=1]:QuorumCnxManager@368] - Cannot open channel to 3 at election address jt1/192.168.1.32:3888 java.net.NoRouteToHostExceptoin: No route to host at java.net.PlainSocketImpl.socketConnect(Native Method)

 

Can you please tell me if I have left out a step in my cdh 4.1 configuration and also how can I troubleshoot this.

Also... I can see the word "WARN" here so is it just a warning? how can I know if my zookeeper is running successfully (perhaps with some warnings).

 

Can you please help me and sorry if this is FAQ.

1 ACCEPTED SOLUTION

avatar
Master Collaborator

ah you are on a different version of netcat

 

try

 

nc 192.168.1.32 3888

 

 

View solution in original post

4 REPLIES 4

avatar
Master Collaborator

 

no route to host suggests you are getting blocked still, using ping as a network test usually isn't sufficient as it uses the ICMP protocol and doesn't have a port per se.

 

a better test would be:

 

nc -z 192.168.1.32 3888 

avatar
Expert Contributor

for me it says 

 

nc: invalid option -- 'z'

Ncat: try --help or main(1) ncat for more information, usage options and help.

QUITTING

avatar
Master Collaborator

ah you are on a different version of netcat

 

try

 

nc 192.168.1.32 3888

 

 

avatar
Expert Contributor

I was able to successfully nc into all machines except one.

 

I went and re-issued these commands on that machine and that resolved the issue.

 

servicectl stop firewalld.service

servicectl disable firewalld.service

 

Thank you so much!!!!!