Speed-O-Meter goes between 60 and over 100%

Troubleshooting
,
1

**CMK version:Enterprise Edition 2.2.0p39 running on a VMware Virtual Host
**OS version:Ubuntu 22.04

Error message:
I am consistently getting bursts of different values for the speed-o-meter. I believe this is causing false positives. I randomly get unknown interfaces, failed pings and issues connecting to checkmk agents.

For example I get “No IP packet received for 15.522954 s (deadline is 15.000000 s)” AND then 1 minute later I get " Packet received via smart PING"

Below is my Global Settings and Core Statistics I dont have issues with CPU or Memory and Fetcher helpers dont seem to go over 20%.

Maximum concurrent active checks....	200
Maximum concurrent Real-Time checks....	1
Maximum concurrent Livestatus connections....	20
Maximum concurrent Checkmk fetchers....	26
Maximum concurrent Checkmk checkers....	5

Service checks:	67.75/s
Host checks and smart pings:	22.35/s
Livestatus connections:	0.17/s
Average active check latency:	0.00s
Average checker latency:	0.00s
Average fetcher latency:	0.00s
Active check helper usage:	0.0%
Fetcher helper usage:	9.8%
Checker helper usage:	2.7%
Livestatus usage:	0.0%

Im not sure where to start with troubleshooting this. Ive been trying to research this for awhile. Any help would be much appreciated.
I do not think its an issue with my network, but rather an issue within checkmk itself and the checks due to Speed-O-Meter fluctuating so much.

2

Not sure if it is related. I suddenly had some issues with failed pings a few weeks ago.

After changing Setup > Hosts > Host monitoring rules > Host Check Command rule from PING (active check with ICMP echo request) to Use a custom check plugin those issues seem to be gone now.

I used $USER1$/check_ping -H $HOSTADDRESS$ -w 400,40% -c 600,80% -t 30 -4 -p 5, but I had to add a symlink for ping first to make plugin work, since ping wasn’t in the expected location of the plugin: /usr/bin/ping.

3

This a Windows host? If so, check it’s local firewall and whether it allows ICMP echo request and reply.

4

I do have it to allow ICMP it only does this randomly. Other times it has the connection

5

If it happens randomly, you might want to investigate, if network equipment in between the affected host and the Checkmk server is overloaded. They tend to drop ICMP traffic in those cases to focus on more important traffic.

6

What is the best way to investigate this? I did have a thought that package could possibly be dropping due to too many of them

7

That is a networking question, and I am afraid that is outside my domain.