#### UPDATE, BUG confirmed see last post ############

I am trying to use the Netwatch tool but am finding the timeout value confusing.

I am running 6.32.2 ROS

I have the interval set to 00:00:10
if I set the Timeout Value to 10ms the netwatch status goes to UP
If I set the Timeout value to 9ms the netwatch status goes to down

This would imply to me that whenever netwatch tool polled the target the response was over 9ms but not over 10ms

this all makes sense so far

However why is it when I ping the target from the terminal window I get a max of 4ms over 10-20 seconds

I am confused as to what is going on here, and don't feel I can setup a netwatch alarm reliably until I work out why the difference in the ms value

Just guessing here...

Could it be when you ping you measure trip time from A to B, but what you refer in netwatch is timeout?

If time for a packet to travel from A to B is 4ms, for netwatch probe to get a packet from A to B (4ms) and the reply back from B to A (another 4ms) = 8ms. Too tight for 9ms timeout.

Just guessing here...

Could it be when you ping you measure trip time from A to B, but what you refer in netwatch is timeout?

If time for a packet to travel from A to B is 4ms, for netwatch probe to get a packet from A to B (4ms) and the reply back from B to A (another 4ms) = 8ms. Too tight for 9ms timeout.

I appreciate the reply but I don't this is what is happening here

I have added a netwach to something that is on the locally connected network to the router and when I ping it from the terminal I get 0ms results.

if I use netwatch the exact same 9ms threshold seems to exist
10ms = UP
9 ms = down

until I understand this I can't reliably use netwatch for anything because I don't trust the data it is giving me.

########BUG confirmed#####

Hello,

Thank you for your e-mail. There was indeed a bug regarding your issue. Looks like we managed to fix it, but the fix is not jet published thus not tested either. The fix will be available for testing in v6.34rc12 and latter. I will also inform you when it gets tested and whether or not it gets published in a bugfix version.

Basically, the problem right now is, the ping response gets rounded up in 10ms precision. If you have 4ms ping response, you have to set timeout to 10ms. If you have 21ms ping response, you have to set timeout to 30ms. And so on.

Regards,
Emils Z.