One Graph missing..wrong perfdata from the check? Icinga2(check_nwc_health), InfluxDB, Grafana

Hello :slight_smile:

I have a setup with about 800 hosts and few thousands services and normally everything works fine with my graphs…except one graph with the interface usage of one server.
I monitor the interface usage form my linux servers with ‘check_nwc_health’. The server which I monitor got an upgrade from Ubuntu 18.04 to Ubuntu 19.10, with 18.04 I had no problems (And the other servers which got an upgrade are still working fine). But after the upgrade I only see the the ‘lo’ interfaces. Even if i check the stored series in the InfluxDB I am not able to find the data, so it should be a problem between Icinga2(‘check_nwc_health’) and InfluxDB and not with graphana… at least this is my best guess

My graph:

If I execute the check manually on the bash I got the following output

root@IcingaServer:~# ‘/usr/lib/nagios/plugins/check_nwc_health’ ‘–community’ ‘xxxx’ ‘–hostname’ ‘XXX.XXX.XXX.XXX’ ‘–mode’ ‘interface-usage’
OK - interface lo usage is in:0.02% (2215.60bit/s) out:0.02% (2215.60bit/s), interface Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 2 (alias eno1) usage is in:0.73% (7293016.80bit/s) out:1.26% (12568806.80bit/s), interface Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 3 (alias eno2) usage is in:0.00% (0.00bit/s) out:0.00% (0.00bit/s) (down) | ‘lo_usage_in’=0.02%;80;90;0;100 ‘lo_usage_out’=0.02%;80;90;0;100 ‘lo_traffic_in’=2215.60;8000000;9000000;0;10000000 ‘lo_traffic_out’=2215.60;8000000;9000000;0;10000000 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 2_usage_in’=0.73%;80;90;0;100 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 2_usage_out’=1.26%;80;90;0;100 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 2_traffic_in’=7293016.80;800000000;900000000;0;1000000000 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 2_traffic_out’=12568806.80;800000000;900000000;0;1000000000 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 3_usage_in’=0%;80;90;0;100 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 3_usage_out’=0%;80;90;0;100 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 3_traffic_in’=0;0;0;0;0 ‘Broadcom Inc. and subsidiaries NetXtreme BCM5720 2-port Gigabit Ethernet PCIe 3_traffic_out’=0;0;0;0;0

I do not find any mistakes there, the perfdata looks correct for me…or do I miss something?

I hope someone has a idea where the problem is… If you need more data, pls ask and say what exactly you need. :smile:

I think that names are to long to store, check your icinga log if icinga discards those metrics.
Also check your snmp server settings why it shows such strange interface names and doesnt use the linux names for the interfaces.

1 Like

Hi Carsten,

thank you very much! :slight_smile: The name was too long. All our server are just show this kind of interface names and not the linux names for it.

For everyone who has the same problem:
Just edit “/etc/influxdb/influxdb.conf” and change “max-values-per-tag” to 0.

:slight_smile:

1 Like