execute command failed

When you say that you have the correct permission, can you do some 

ls -l /the-path-of-scripts/centreon_linux_snmp.pl

The command is executed with the centreon-engine user, so maybe there is cases where there is issues.
Aside of that, if you can copy the command that is failing on centreon and use it in this (replace the text in brackets)

sudo -u centreon-engine [your-command]

That way you can execute the command with the permissions used by the engine. I suppose that as you migrated some cache or associated file may have changed, and centreon doesn’t show stderr in the output on the page.

root@srv ~ $ ls -l /usr/lib/centreon/plugins/centreon_linux_snmp.pl
-rwxrwxr-x 1 root root 775694 17 nov.  07:46 /usr/lib/centreon/plugins/centreon_linux_snmp.pl
root@srv ~ $ sudo -u centreon-engine /usr/lib/centreon/plugins//centreon_linux_snmp.pl --plugin=os::linux::snmp::plugin --mode=cpu --hostname=x.x.x.x --snmp-version='2c' --snmp-community='test'  --warning-average='80' --critical-average='90'
OK: 4 CPU(s) average usage is 27.25 % | 'total_cpu_avg'=27.25%;0:80;0:90;0;100 'cpu_0'=27.00%;;;0;100 'cpu_1'=30.00%;;;0;100 'cpu_2'=22.00%;;;0;100 'cpu_3'=30.00%;;;0;100
 

Then i see nothing wrong there, and you have that specific command that fails to be executed on the centreon.

Quick question, are you using pollers ? Maybe the script centreon_linux_snmp.pl is not present on the specific poller where the service is configured for.

Yes, I am using a remote poller, not the central one, and all the commands above were executed from this poller. Furthermore, on the web interface, I can clearly see that the check is issued on the poller and not the central one.

I have no idea why then… Maybe some internal error to the centreon engine, or some resource MACRO only setup on the central poller and not the remote. I’m clueless from there. Sorry