[CentOS 7][Cacti 1.1.38] Problem with device status

[pberger]

Good morning,

I have a CentOS 7 device with Cacti 1.1.38 running on it.
Hereby my issue:
Settings: I have set to monitor on either ICMP or SNMP. Logs show no issues (except for a warning about the cron time being set to 1 minute while the poller takes 5 mins. No real problem there, will fix that soon)
Device 1: This device is down. ICMP does not arrive. Screen shows ICMP timed out. .. now the issue: Device status shows up (this is NOT the case)
Device 2: This device is up. ICMP does arrive, has no issue and updates regularly with the correct latency.

So.. the system does work, the pinging does work.. except when the pinging does not work, the device isnt shown as offline.
I may say.. this is the problem with ALL devices. If i change any IP to a non-existing/non-reachable one.. all still show status UP.

Attached screenshots that should make it clear
1 displays 2 devices with both status up.
2 displays a device without ping arriving even though 1 does not show down
3 shows the ping-mechanic does work when a device is online

1.jpg (38.75 KiB) Viewed 2005 times

---------------------

2.jpg (178.42 KiB) Viewed 2005 times

------------------------------------

3.jpg (175.77 KiB) Viewed 2005 times

[netniV]

That is because the OR detection is checking whether Ping OR Uptime responds, if it does, then it will be considered up.

[pberger]

That is because the OR detection is checking whether Ping OR Uptime responds, if it does, then it will be considered up.

Ah sorry for having the SNMP result blocked up in the picture.
At the device with the ICMP ping down, the SNMP also is not responding. So neither the ICMP nor the SNMP are responding but its still considered up.
So sadly doesnt fit that conclusion
Even setting it to just check the ping method, still doesnt count it as offline.
Basically i just cant get any of my devices to the offline status

[camerabob]

What is monitoring set to?

[netniV]

It may not yet go offline if you have set a number of events to occur before it's considered down.

[pachubw]

Same issue for me too -

I have a CentOS 7 device with Cacti 1.1.38 running on it.
I need to monitor the traffic on devices. For testing, I have added only 1 device with 1 Interface which have the traffic.

1.JPG (94.78 KiB) Viewed 1990 times

2.JPG (85.21 KiB) Viewed 1990 times

3.JPG (112.62 KiB) Viewed 1990 times

I was able to monitor the traffic before, suddenly it stopped and it started this error.


Please help me, if someone how to fix this issue.

[camerabob]

Before what? What has changed? Was there a yum update?

[pberger]

What is monitoring set to?

Which monitoring? Its just set to go off on offline SNMP or ping in the Availability/Reachability Options of the device. Shouldn't that be enough? (as far as im looking at my old 0.8.8a setup where it does work, looks to be set the same way)

It may not yet go offline if you have set a number of events to occur before it's considered down.

Not as far as im aware. Only set up the settings at the Availability/Reachability Options of the device. Cant see much else that differs from my old still working setup either.
Any place i should check to make sure?

[pberger]

What is monitoring set to?

Ah, think you meant this? Just standard.. only fully offline check, no latency alerts

2019-03-06 07_57_04-Console -_ Devices -_ (Edit).jpg (108.4 KiB) Viewed 1979 times

[pberger]

Kept searching for clues but still havent found any. The devices are responding correctly to SNMP and ping but just dont go offline when either of them fails :S

[pberger]

Seems like the graphs are not recording the data either even though the data is showing.
Graphs staying empty

[MrRat]

Can you post a screenshot that includes the snmp information at the top when the device is up?

[pberger]

Can you post a screenshot that includes the snmp information at the top when the device is up?

Hereby 2 examples. When a device is up it does show the SNMP and ICMP correctly. When its offline, it will show a timeout/connectionissue on both of the options however the device will still be on the "up" status

2019-03-11 13_56_08-Console -_ Apparaten -_ (Wijzig).jpg (158.12 KiB) Viewed 1909 times

2019-03-11 13_57_58-Console -_ Apparaten -_ (Wijzig).jpg (165.15 KiB) Viewed 1909 times

[MrRat]

You are hiding the pieces that I'm thinking are relevant. At one time I found that if there was an apostrophe in the snmp location that the devices status would never change. Can you remove any special characters from snmp location and contact just to see if that resolves your issue?

[MrRat]

I did try to reproduce my theory with 1.2.2 and cannot reproduce. I know the issue I saw was with 1.1.38. I’m not suggesting upgrade yet though. Wait for 1.2.3 in the coming days. The workaround was to remove the apostrophe from the snmp location on devices. I should have reported this but it appears that the issue has since been resolved.