Documentation

PerfIf-Check

Checks transfer and errors per ethernet interface.

Usage

$ check_netapp_pro.pl PerfIf -H <host> [-z recv_data|send_data|recv_errors|...] [...] [--help]

Description

This plugin checks various performance counters of a NetApps ethernet interface.

PerfIf will not work on cluster-mode filers with DataONTAP >= 8.2.x. Please use the object lif for the getter and the check PerfLif instead.

A list of supported counters is printed below at --counter.

Hint: Calling the getter with the switch --discover=counters will explain each available counter in detail.

Simple Examples

[S1] ./check_netapp_pro.pl PerfIf -H filer --explore=counters

List all available counters on the target system.

[S2] ./check_netapp_pro.pl PerfIf -H filer

Checks all interfaces using the default thresholds.

NETAPP_PRO PERFIF CRITICAL - 11 ifnets checked, 2 critical and 1 warning.
No SVC-ACK relevant changes in reason.
e0a (cluster1-01:kernel:e0a): 3476.4B/s (CRITICAL)
e0a (cluster1-02:kernel:e0a): 2296.9B/s (CRITICAL)
e0b (cluster1-02:kernel:e0b): 1365.8B/s (WARNING)
e0b (cluster1-01:kernel:e0b): 426.7B/s
e0c (cluster1-01:kernel:e0c): 198.0B/s
e0d (cluster1-01:kernel:e0d): 18.2B/s
e0c (cluster1-02:kernel:e0c): 16.9B/s
g0a (cluster1-01:kernel:g0a): 0.0B/s
losk (cluster1-01:kernel:losk): 0.0B/s
e0d (cluster1-02:kernel:e0d): 0.0B/s
losk (cluster1-02:kernel:losk): 0.0B/s
 | e0a=3476B/s;1024;2048;0; e0b=427B/s;1024;2048;0; e0c=198B/s;1024;2048;0; e0d=18B/s;1024;2048;0; g0a=0B/s;1024;2048;0; losk=0B/s;1024;2048;0; e0a=2297B/s;1024;2048;0; e0b=1366B/s;1024;2048;0; e0c=17B/s;1024;2048;0; e0d=0B/s;1024;2048;0; losk=0B/s;1024;2048;0;

[S3] ./check_netapp_pro.pl PerfIf -H filer -s cluster1-01

Checks all interfaces on cluster1-01.

NETAPP_PRO PERFIF CRITICAL - 6 ifnets checked, 1 critical and 0 warning.
No SVC-ACK relevant changes in reason.
e0a (cluster1-01:kernel:e0a): 207661.4B/s recv_data (CRITICAL)
e0b (cluster1-01:kernel:e0b): 479.3B/s recv_data
e0c (cluster1-01:kernel:e0c): 193.7B/s recv_data
e0d (cluster1-01:kernel:e0d): 0.0B/s recv_data
g0a (cluster1-01:kernel:g0a): 0.0B/s recv_data
losk (cluster1-01:kernel:losk): 0.0B/s recv_data
 | e0a=207661B/s;1024;2048;0; e0b=479B/s;1024;2048;0; e0c=194B/s;1024;2048;0; e0d=0B/s;1024;2048;0; g0a=0B/s;1024;2048;0; losk=0B/s;1024;2048;0;

[S4] ./check_netapp_pro.pl PerfIf -H filer -s cluster1-01 -x losk

Checks all interfaces on cluster1-01, except of losk.

NETAPP_PRO PERFIF CRITICAL - 5 ifnets checked, 1 critical and 0 warning.
No SVC-ACK relevant changes in reason.
e0a (cluster1-01:kernel:e0a): 207661.4B/s recv_data (CRITICAL)
e0b (cluster1-01:kernel:e0b): 479.3B/s recv_data
e0c (cluster1-01:kernel:e0c): 193.7B/s recv_data
e0d (cluster1-01:kernel:e0d): 0.0B/s recv_data
g0a (cluster1-01:kernel:g0a): 0.0B/s recv_data
 | e0a=207661B/s;1024;2048;0; e0b=479B/s;1024;2048;0; e0c=194B/s;1024;2048;0; e0d=0B/s;1024;2048;0; g0a=0B/s;1024;2048;0;

[S5] ./check_netapp_pro.pl PerfIf -H filer -z recv_data -w 2000 -c 3000 -I e0c

Check the data received on interface e0c - warn if data rate is greater than 2000 Byte per second, exit critical if data rate is greater than 3000.

[S6] ./check_netapp_pro.pl PerfIf -H filer -z recv_errors -w 20 -c 50

Warn if more than 20 errors per second when receiving packets.

Advanced Examples

Hint: The input for advanced examples is shortend. Replace the '...' below with ‑H <your filers ip or host-name>

[A1] ./check_netapp_pro.pl PerfIf ... -z recv_errors -w 20 -c 50 -I ^e\d

Check all interfaces, whose name starts with 'e' followed by a number. Eg. e0a, e0b, e1a, ... Please look into the advanced examples in the Usage-module for even more possibilities of the --include and --exclude switches.