Resolved Monitoring 360 - Data Missing Alert

[Greg Sevastos]

Server operating system version

CloudLinux 8.10 (Vladimir Aksyonov)

Plesk version and microupdate number

18.0.66 #1

Hi,

I’m getting pretty frequent alerts sent to my email with the following information:

The alert for data missing alert on [host] has been opened. The threshold has been triggered for 17 minutes.

Other monitoring tools haven’t picked up anything and every seems to be running fine.

Any pointers to diagnose/fix the issue with Plesk + Monitoring 360?

Thanks.

 

[Sebahat.hadzhi]

Could you please check /var/log/syslog and confirm if there were any incidents? If there aren't there's likely a connectivity issue between the server and the 360 Monitoring Backend. One of the possible reasons is misconfiguration of the monitoring agents. Could you please double-check /etc/agent360-token.ini and make sure it matches the values specified here:

https://support.plesk.com/hc/en-us/articles/19601781026583-Server-shows-no-data-received-or-is-not-showing-in-360-Monitoring

 

[Greg Sevastos]

Hi Sebahat,

The tokens in /etc/agent360-token.ini are correct. I'll keep an eye on the syslog to see if anything presents itself.

Thanks.

 

[Greg Sevastos]

Hi again,

agent360.log is showing the following errors:

2025-01-10 08:54:20,466  ERROR    plugin_exception
Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/agent360/agent360.py", line 503, in _execution
    payload = plugin.run(self.config)
  File "/usr/local/lib/python3.6/site-packages/agent360/plugins/iostat.py", line 100, in run
    disks = diskstats_parse()
  File "/usr/local/lib/python3.6/site-packages/agent360/plugins/iostat.py", line 39, in diskstats_parse
    lines = open(file_path, 'r').readlines()
PermissionError: [Errno 1] Operation not permitted: '/proc/diskstats'

 

[Sebahat.hadzhi]

Thank you for the update. I was able to find a similar case where our engineers observed that if agent360 is not running as root resulted in triggering such false-positive notifications. You should be able to confirm that in Monitoring > Process. If the user is agent360 or anything different than root, please try the applying the instructions from the following guide:

Run the Monitoring Agent as the root User

By default, agent360 runs under its own user agent360 . This prevents the monitoring agent from having access to certain data, such as process-level disk I/O. This document explains how to run agent360 as the root user to unlock these statistics. If the /etc/systemd/system/agent360.service file...

docs.360monitoring.com

 

[Greg Sevastos]

Hi,

I was looking through the docs and checked the settings as outlined below, as I’m running Cloudlinux: CloudLinux and CageFS

The wrong group id was set for the value fs.proc_super_gid (not sure when this changed on my system) - agent360 was also missing from the file: /etc/cagefs/exclude/systemuserlist

I followed the instructions again and it appears the error has stopped. Will monitor logs and see if the issues arises again now.

Thanks for the doc link!

 

[Greg Sevastos]

Hi Sebahat,

It appears that running through the CloudLinux and CageFS settings again as per the docs has resolved the issue.

Thanks for your help.