check_snmp_int filling 0 valued rows PALO-ALTO 3220

[afernique]

Hi,
Sometimes in our /tmp/tmp_Icinga_int.[IP].161.[interface_name] file lines with 0 values appear

Expected Behavior

1677147427:130606416903670:49920642709205:0:0:0:0
1677147727:130633105024917:49926414084156:0:0:0:0
1677148148:130666857271937:49932368204656:0:0:0:0
1677148448:130690060146440:49937607659580:0:0:0:0

Current Behavior

1677147427:130606416903670:49920642709205:0:0:0:0
1677147727:130633105024917:49926414084156:0:0:0:0
1677148027:0:0:0:0:0:0
1677148148:130666857271937:49932368204656:0:0:0:0
1677148448:130690060146440:49937607659580:0:0:0:0

Possible Solution

We've modified the read_file function to bypass those lines when in value less than 1

# if ($#file_values >= ($items_number - 1)) {
# modifié le 23/02/2023 par A.FERNIQUE
# j'ajoute une condition si le compteur dans le fichier est 0 on saute
if (($#file_values >= ($items_number - 1))&&($file_values[1] >=1)) {

Steps to Reproduce (for bugs)

Dunno this only seems to happens on our palo-alto 3220 (rev 10.2.3-h2) So i guess it somehow sends bad values.

Context

monitoring solutions, huge throughput appearing on graph which are false

Your Environment

omd distrib on debian bullseye with snmp plugins for ease of use.

[drdisk]

Hi.

Thanks for your feedback. This seems to be a solutions which may work for you - but I don't think at the moment that this may be a solution for everyone. More interessting in the first step for me would be to search why you get this lines. The reason maybe outside of the check as you already assumed. So for a clean fix that should be fixed outside of the check.

Have a nice day,
Steffen