Wireless devices are often down/up

Troubleshooting
1

Hello!

I’ve install check_mk a couple of weeks ago. Working fine, that is perfect. I have an issue where I got temperature sensor, connected to 2.4ghz network, and these devices always go up/down.

I thought maybe it was the threshold for icmp, so I set a rule in the folder where my device are, 4000 warning and 5000 critical, packet loss 80% warning, crit at 100%, number of packet 20, total of timeout 6.

I also set to check every 10 mins. When it goes down, the graph says: Host perfdata, last 396.6114ms, and it also show warning at 200 and crit at 500…

Metrics: rta=975.447ms;200.000;500.000;0; pl=0%;80;100;; rtmax=1088.230ms;;;; rtmin=857.098ms;;;;

Anyone can help me? I am not sure about the configuration. Note, any device that are using a network cable, I never get these thing.

Thanks in advance :slight_smile: !

2

Hi,

Well first of all it appears that the thresholds are not being applied.
I would check the parameters first and if the rules are being applied, try to see if you created the rule in the right folder where you have those hosts or try to move it to main directory.
If you are unsure, just follow the icon to parameters and post what it has.
Let me know.

3

Thanks for your reply. I have double check, they are in the right folder. See image Below

Imgur

Imgur

thanks again for your time :slight_smile:

4

Imgur

Parameter for the host, there no rule need to be trigger, but on the service PING

Imgur

5

Sorry about that but re reading I guess you are having issues on host down not on ping “service” try to review parameters for host check command !

EDIT: On previous versions I also found strange things and moving rules to main directory kind of fixed it. I didn’t deep dive enough lol

6

Ah maybe.

I have this:

Host Check Command…
Default Value PING (active check with ICMP echo request)

I tried to disable this rule, then just keep PING Services. Even wit ha negate rule, It will not say that a rule has been applied, even at the main directory.

Note: I’m on 1.6.0p6

edit: I tried to disable notification for Host Up down and keep services, but It might just stop sending email, not fixing the issue…

7

It will not work like that, you should give the same parameters that you made for the active check there and it will apply to host parameters and not service.

I don’t have access to my lab right now but later I’ll try out and take some prints. Are you on raw or enterprise edition ?

8

…I made a different solution for myself.

every AP has a label ap:nonalarm and i bend the output to always okay (with the corresponding rule) … (because I only look into the APs if there are messages from the WLCs)
on the WLC`s I have set up my actual monitoring and the alarms.

1 Like
9

Yes that’s a good approach indeed !

10

Is there a way to change the active check default parameter?

but for those IoT Devices, it looks to be working fine now. I have no active check but only PING, and I have like 98% less false email.

11

this can be fixed by:

https://checkmk.de/cms_notifications.html

point 6.9

2 Likes
12

Finally, what I did, I set a rule for host check, in t his folder, to use the status of service PING. Now, I almost have no problem. I changed also to 10 mins delay to ping, I don’t need those to be always checked.