how to solve the SNMP fake alert?

[newcomer01]

Dear 4rum,

I am using Cacti v0.8.8b and having device settings as ICMP ping method, Ping and SNMP uptime as downed device detection, ping timeout value 400.

There are 50 networking device have configured like above settings but sometimes I received the SNMP alert vie email with subject "Host error DEVICE NAME (IP address) is DOWN" and with the reason: SNMP not performed due to setting or ping result, ICMP ping timed out from only 3 deives of them, while the device is up and running well and I saw its uptime is very long up to at that time (30 days), all remaining devices are normally.

Could you please help to fix why this fake alert was generate?

[newcomer01]

This issue is going on ...

Any help would be greatly appreciate.

[newcomer01]

hello ....


any suggestion please?

[Rno]

Did you use spine ?
I had the same issue and it was the permission of spine :
-rwsr-xr-x 1 root root 327872 Feb 2 08:58 /usr/local/spine/bin/spine

[newcomer01]

No! I do not use spine. I just use cacti with basic settings 2 years ago without any issue.

The image is my cacti error log.

[Rno]

Did you comply with this remark ?
ICMP requires that the Cacti Service ID have root privilages in Unix

When I had my issue, like yours it was due to the fact that my cacti process didn't have the root privilege, and 20 devices or so (I have over than 400), where having an issue once in a while with ping.

[newcomer01]

The remarks "ICMP requires that the Cacti Service ID have root privilege in Unix" is not yet applied to my cacti since I do not know about this. But my cacti is up and running properly 3 years ago without any issue as per mentioned subject.

My Cacti is installed on CentOS 6.4 and I am not friendly with this *nix distro. I will learn how to configure my Cacti service ID have a root privilege.

I am highly appreciate for your support.