Problems with node monitoring with ping only

[zscherlich]

Hello Support,

We have a number of servers that we are successfully monitoring with NetXMS, but they all have a NetXMS agent installed.
Now we want to monitor some servers only by ping, of all things this is causing us problems.

We create a new node and enable only "ICMP ping for all polls".
The node is created with an interface "unkown".
At the node we set "Use ICMP ping on primary IP address to determine node status".
A "Poll > Status" has the following output:

10.01.2022 17:14:32] **** Poll request sent to server ****
[10.01.2022 17:14:32] Poll request accepted, waiting for outstanding polling requests to complete...
[10.01.2022 17:14:32] Starting status poll for node stragazer
[10.01.2022 17:14:32]    Starting status poll on interface unknown
[10.01.2022 17:14:32]      Current interface status is UNKNOWN
[10.01.2022 17:14:32]      Starting ICMP ping
[10.01.2022 17:14:32]      Interface is UNKNOWN for 5 polls (1 poll required for status change)
[10.01.2022 17:14:32]      Interface status after poll is UNKNOWN
[10.01.2022 17:14:32]    Finished status poll on interface unknown
[10.01.2022 17:14:32] Checking primary IP address with ICMP ping
[10.01.2022 17:14:32]    Primary IP address is not responding to ICMP ping
…

A ping to the node from the server with the NetXMS server works of course.

What is missing in the configuration?
The NetXMS server has version 3.9.420.

Sincerely,

Frank

[zscherlich]

Hi all,

Doesn't anyone have a clue for me?

Sincerely,

Frank

[victor]

Hi,

on what OS server is running? If Linux, do you run it under root or under dedicated account?

Best regards,
Victor

[zscherlich]

Hi,

it's Linux with a dedicated account.

Sincerely,

Frank

[victor]

Then most likely netxmsd process cannot create raw sockets. You can grant it with command

sudo setcap cap_net_raw=ep /usr/bin/netxmsd

(use actual path to netxmsd)

Best regards,
Victor

[zscherlich]

Hi,

that is the solution, thank you very much!

Sincerely,

Frank