Problem with Plugins on new Icinga 2 server

Icinga 2
1

I have a new VM running Ubuntu 24.04, Icinga 2 clean install okay with director. Install of linuxfabrik ok without errors.

Had a couple of issues with setting up the hosts, but that seems to be mostly resolved now, they’re in the system and talking to the server. And I have 3 services already setup which are automatically trying to run on each server, a simple ping, uptimerobot & disk usage. They’re all running but they’re getting a “file not found” error for each one.

Command /usr/lib64/nagios/plugins/uptimerobot --timeout 30 --url https://status.linuxfabrik.io failed to execute: 2, “The system cannot find the file specified.”<Terminated with exit code 127 (0x7F).>
Command /usr/lib64/nagios/plugins/ping --count 5 --hostname 127.0.0.1 --interval 0.200000 --timeout 5 failed to execute: 2, “The system cannot find the file specified.”<Terminated with exit code 127 (0x7F).>
Command C:\Program Files\icinga2\sbin\linuxfabrik\disk-usage.exe --critical 95%USED --warning 90%USED failed to execute: 2, “The system cannot find the file specified.”<Terminated with exit code 127 (0x7F).>

Is this because the host install on each failed at the end with an error regarding the linuxfabrik-monitoring-plugins.msi file not found or could not be opened ???

image
image1422×996 325 KB

It was my understanding that the files should be getting downloaded from inside the installation script. (below)

[Net.ServicePointManager]::SecurityProtocol = 'tls12, tls11';
$ProgressPreference                         = 'SilentlyContinue';
[string]$ScriptFile                         = 'C:\Users\Public\IcingaForWindows.ps1';

Invoke-WebRequest `
    -UseBasicParsing `
    -Uri 'https://packages.icinga.com/IcingaForWindows/IcingaForWindows.ps1' `
    -OutFile $ScriptFile;

& $ScriptFile `
    -ModuleDirectory 'C:\Program Files\WindowsPowerShell\Modules\' `
    -InstallCommand '{"IfW-DirectorSelfServiceKey":{"Values":["f024e252bcd908284d6dbfaa464ffcb8196a0718"]},"IfW-DirectorUrl":{"Values":["http://10.32.8.49/icingaweb2/director/"]}
}'; 

$pkg = "linuxfabrik-monitoring-plugins.msi"; # https://download.linuxfabrik.ch/monitoring-plugins/lfmp-2.1.1-1.signed-packaged.windows.x86_64.zip

Write-Host "[Notice]:" -ForegroundColor Green -NoNewline
Write-Host " Starting installation of $pkg"

$process = Start-Process msiexec.exe `
    -ArgumentList "/i `"$pkg`" /norestart /qn" `
    -Wait `
    -PassThru

$exitCode = $process.ExitCode

# Map common MSI exit codes
$messages = @{
    0    = "Success"
    1603 = "Fatal error during installation"
    3010 = "Success, reboot required"
    1618 = "Another installation is already in progress"
    1619 = "MSI file not found or could not be opened"
    1620 = "Invalid or corrupt MSI package"
}

if ($messages.ContainsKey($exitCode)) {
    switch ($exitCode) {
        0 {
            Write-Host "[Notice]:" -ForegroundColor Green -NoNewline
            Write-Host " MSI exited with code $exitCode ($($messages[$exitCode]))"
        }
        3010 {
            Write-Host "[Warning]:" -ForegroundColor White -NoNewline
            Write-Host " MSI exited with code $exitCode ($($messages[$exitCode]))"
        }
        default {
            Write-Host "[Error]:" -ForegroundColor Yellow -NoNewline
            Write-Host " MSI exited with code $exitCode ($($messages[$exitCode]))"
        }
    }
} else {
    Write-Host "[Warning]:" -ForegroundColor White -NoNewline
    Write-Host " MSI exited with unmapped code $exitCode"
}
2

No, the script doesn’t download the zip automatically.

I already answered why, in the other thread: Problem with new Icinga2 Server & Windows Setup - #13 by rivad

3

So it looks like most of the issues have been cleared. I’ve got 4 plugins working on 2 of the 3 host servers but I’m having issues with the original host. This is the one where I tried installing the agent normally first. I uninstalled that agent and I’ve run the script @rivad gave me twice, which ran successfully, no errors. But I’m still getting “jaz-fsvm is not connected to npbvmjazicinga” errors on each of the 4 checks.

How do I troubleshoot this on one host? Is there a script or program that will wipe/clear the host server of all things Icinga, so I can try it from scratch? The other 2 servers that were run with just the script are working fine. I’m about to add a few more hosts to see if it keeps working, but I need that original host working, it’s one of the most important I need to monitor.

4

How did you configure the host on the icinga server side? Maybe deleting it there, rerun the script and deploy the updated director configuration.

5

I’ve tried that, twice. The script runs without errors, but it still says “not connected” on the icinga server side. I even restarted the icinga server itself. No change.

6

I there a reason in the logs?

7

Sigh, I hate when I’m stupid.

I focused entirely on the host and forgot to delete it from the server side.

Let me see if that works this time…

8

That seems to have done it… I’m getting a bunch of “overdue” messages, even though I’ve clicked on “check now” for each of the 4 checks, but it’s not giving me the “not connected” error anymore.

Thank you for the assistance. I literally couldn’t have done this without you!

9

Don’t worry, I also had my problems in the beginning until I got the hang of the whole cluster thingy.

I even recruited the @linuxfabrik to get me started with the director on a HA setup - had a single master setup before and did the config by hand.

10

Ok, this is weird… it looked like it was going to be fine, but now all 4 are back in the “service problems” category with the “not connected” error again.

I don’t get it. They’re both using FQDN, same as the other 2 hosts that are working fine. The only difference with this one is that it’s the one I was using at first and had setup the full Icinga Powershell stuff on with the Icinga 2 agent. I uninstalled the agent in Control Panel, I don’t know any way to clean out the other stuff. Your script runs without errors except two places where it tries to install a file that already exists. The host can ping the server, can access the site, it’s not that. And when I deleted the host from the server, then reran the script on the host, it re-added the host to the host list on the server, so CLEARLY it’s “connected”. I don’t know why the service checks are giving that error.

11

Did you put it into the /etc/icinga2/zones.conf? If so clean it out and restart the icinga2 service.

12

On the server itself? I don’t think so, but let me check. I did a bunch of weird things when I was first trying to get this running…

13

Ok, it’s not in zones.conf. Any other location I should check? They’re clearly talking to each other.

I have 6 hosts setup now and only that one is the one giving errors. But again, it’s the one I didn’t initially setup with your script. Ugh…

14

If you execute icinga2 -V you get all paths.
But I would enable the debug log first and try to figure out, what’s wrong with with the connection, the scheduling or what ever.