Cloudera Manager. Failed to detect Cloudera Manager Server

问题

I have two PC's with CentOS 6.5

client86-101.aihs.net 80.94.86.101

client86-103.aihs.net 80.94.86.103

cloudera-manager-server installed on client86-101.aihs.net. I have the problem on detecting Cloudera Manager Server(3rd step on cluster installation.)

Issue trace:

BEGIN host -t PTR 80.94.86.101
101.86.94.80.in-addr.arpa domain name pointer client86-101.aihs.net.
END (0)
using client86-101.aihs.net as scm server hostname
BEGIN which python
END (0)
BEGIN python -c 'import socket; import sys; s = socket.socket(socket.AF_INET); s.settimeout(5.0); s.connect((sys.argv[1], int(sys.argv[2]))); s.close();' client86-101.aihs.net 7182
/usr/bin/python
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "<string>", line 1, in connect
socket.error: [Errno 111] Connection refused
END (1)
could not contact scm server at client86-101.aihs.net:7182, giving up
waiting for rollback request

I think the problem is in this domain names(lient86-101.aihs.net, client86-103.aihs.net). Because if you check this domain names ip-address it will get 80.94.86.165 for both names. But I don't know how I can fix this problem?

If you need some more data, please tell me.

Sorry for my english.

回答1:

It seems that port 7182 is blocked due to the secuirity group configurations. Try to allow 7182 TCP port. It is recommended to allow the ports in this link.

回答2:

If both DNS gives you the same IP, I confirm you've got a problem...

Check you're running a CentOS x86_64
Check the real IP of each node (ifconfig). If the IP is not the expected one, edit /etc/sysconfig/network-scripts/ifcfg- to set the expected fixed IP and restart the network service (/etc/init.d/network restart)
Check in each node of your cluster the file /etc/sysconfig/network the real hostname of your node (one should be client86-101.aihs.net and the other client86-103.aihs.net)
Check in each node of your cluster the file /etc/hosts to assign the correct IP Address to the correct hostname.
Check iptables is off (/etc/init.d/iptables stop && chkconfig iptables off)
Start (or restart) the cloudera-scm-server service in client86-101.aihs.net (service client86-101.aihs.net start (or restart))
Delete any existing cluster you made (main page of Cloudera Manager, on the left, the button on the right next to the cluster(s) name(s), delete)
You should be able to start again the deployment process of CDH on your cluster. Don't forget to specify the two nodes when you're defining your cluster, Cloudera Manager doesn't manage the node hosting the Cloudera Manager Server by default as a Hadoop node.

Note that Cloudera Manager is expecting such a definition of /etc/hosts to work :

127.0.0.1 localhost
80.94.86.101 client86-101.aihs.net client86-101
80.94.86.103 client86-103.aihs.net client86-103

and this configuration should be exactly the same for both nodes.

Hope this helps

来源：https://stackoverflow.com/questions/21912384/cloudera-manager-failed-to-detect-cloudera-manager-server

标签

Hadoop