Failed to detect Cloudera Manager Server.

[#1]

I am adding extra aws nodes using cloudera manager hosted on VM locally. However, scm server hostname is pointing to my broadband instead of the hostname where cloudera magager is running and giving error as shown in image below:

Below are entries of /etc/resolv.conf:

# Generated by NetworkManager

search cloudera
nameserver 192.168.0.1

[#2]

Hello,

I am pretty sure that the $SSH_CLIENT environment variable governs the IP address and then the following call is used to arrive at a FQDN for Cloudera Manager:

Then, "host -t PTR <IP>" is run to get the FQDN

For instance, in my testbed:

#echo $SSH_CLIENT 

10.17.102.53 65520 22

#host -t PTR 10.17.102.53

53.102.17.10.in-addr.arpa domain name pointer host-10-17-102-53.coe.cloudera.com.

The last value is chosen as the Cloudera Manager hostname.

I'm not sure if this helps you out, but if you can't figure this out, I recommend adding the host manually:

1. Install "cloudera-manager-agent" and "cloudera-manager-daemons"
2. Edit the /etc/cloudera-scm-agent/config.ini file and set the "server_host" to your Cloudera Manager host
3. start the agent with "service cloudera-scm-agent start"

[#3]

The CM server listens (for the agents) on port 7182 so you need to add that to your security group in AWS:

• In the EC2 dashboard, select Security Groups
• Check the box by the security group of your cluster
• Select Actions > Edit inbound rules
• Click Add Rule
• Input Port Range=7182 and Source=Anywhere (or some other appropriate source)

[#4]

Thanks so much, have same problem and was solved by adding 7182 to EC2 security group.