Your data. Anywhere you go.

New Relic for iOS or Android


Download on the App Store    Android App on Google play


New Relic Insights App for iOS


Download on the App Store


Learn more

Close icon

Relic Solution: A Common Reason for a Host Failing to Show Up in The Infrastructure Dashboard


#1

You just upgraded to Infrastructure, and you’re excited to start seeing host data in your Infrastructure dashboard. But every time you try to start the New Relic Infrastructure Agent (NRIA), it runs for about 3 minutes then stops, and no data is showing up for your host in the dashboard. What’s going on?

One of the most common reasons for this is that the NRIA is picking up a hostname of localhost. The NRIA does not work with hosts named localhost. The backend will fail to process data and the agent itself will shut down after about 3 minutes.

One of the differences between Servers and Infrastructure is that Servers used the host’s hostname, while Infrastructure actually looks at the Fully Qualified Domain Name (FQDN) of a host, and uses that as the hostname. You can have a unique hostname set up, but if your FQDN is still set to localhost, it will cause the NRIA to fail.

How can you check this on your host?

For Linux hosts:

From a terminal, enter hostname -f and check the output.

For Windows hosts:

This information is located under System Properties -> Computer Name next to Domain, and under Active Directory Users and Computers.

OK, now what?

If you see that localhost is the FQDN on the host you’re trying to install the NRIA on, change it to a unique name, then restart the NRIA. Another way around this is to set the display_name: attribute in your newrelic-infra.yml config file. This will override any FQDN that exists on the host and cause the NRIA to pick up on whatever display_name: you’ve set instead. Note that you will have to restart the NRIA to get it to pick up on this change.

Hopefully this helps get you up and running with New Relic Infrastructure!


Data not reflecting on NewRelic-Infra dashboard
Infrastructure not finding the correct hostname
Hosts showing as undefined in Infrastructure
#2

For the log-savvy, you can identify this issue by generating some verbose logs from your Infrastructure agent (read here on how to do that).

Once you have a few minutes of verbose-level logs to look at, search for identity and you should find a line that looks like the following:

level=debug msg="Updating identity" new=[localhost.localdomain] old=[]

If you see that the Infrastructure agent is picking up localhost as the host’s identity, then you know that you can fix the problem by setting the FQDN to anything other than localhost.


#3

Since this appears as a link in the troubleshooting documentation I thought I would add this information since it doesn’t appear to be here and can cause problems.

If you have cloned a linux machine and the ssh fingerprint is still the old one it will not communicate with newrelic until it is regenerated.

if you run “ssh-keygen -l -f /etc/ssh/ssh_host_rsa_key” and the host mentioned isn’t correct it will cause problems. you can regenerate your keys or simply delete the files and restart, they will be recreated.


#4

Thank you for sharing that on this thread, Greg :smiley: