Netonix Forums

I'm not sure this is really a software issue, but more of a suggestion if you will.

Currently any of the netonix switches will send an email if an unexpected link change occurs. But in the event someone is having trouble with 1G links downgrading to 100M it doesn't send the notification until the link resumes at 1G. This seems counter productive because it would be helpful to know the link dropped to 100M or that the link dropped altogether. Typically when a link falls back to 100M the log looks like this:

Code: Select all

Dec 3 13:34:40 Port: link state changed to 'down' on port 8
Dec 3 13:34:59 Port: link state changed to 'up' (100M-F) on port 8


Usually a quick cable test will cause the link to resume at 1G and that log looks like this:

Code: Select all

Dec 3 20:46:59 Port: link state changed to 'down' on port 8
Dec 3 20:47:06 Port: link state changed to 'up' (1G) on port 8
Dec 3 20:47:07 switch[648]: unexpected link change on port 8 (Port 8 Cora Junction PBE) from 100M-F to 1G
Dec 3 20:47:09 sSMTP[651]: Sent mail for ****@*****.com (221 2.0.0 mail.*******.com closing connection) uid=0 username=bleh outbytes=577


We have a few sites where there appears to be some EMI issues that causes this (typically FM transmitters) and getting an email about the issue after it's been corrected is nearly useless, but getting one when the link drops could be much more helpful.

(We actually use SNMP for monitoring this and correcting, but still worth mentioning.)

It will send a notification on all unexpected link changes. However, there is a cooldown period. Notifications will not be sent within 5 minutes of certain events:
- switch boot
- manual link speed change on the port
- port bounce (manual or due to ping watchdog)
- power management (ie PoE was disabled on the port due to high temp or insufficient power or something)

In your first example either the link was already at 100M (and thus there was no speed change) or the event occurred during the cooldown period.

None of those had occurred within 5 minutes on that port. Another port had a poe event due to voltage. But I'm 99.9% sure I've seen this occur in cases where none of those events had occurred anywhere in the switch.

I just tried logging into a radio and forcing it to 100M. The notice worked one time and didn't the next.

Here's another instance where these notifications could be improved. This has been ongoing for hours but wasn't noticed due to our snmp and other probes not seeing the changes when so close together.

Code: Select all

Dec 18 23:18:55 Port: link state changed to 'down' on port 3
Dec 18 23:18:57 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:04 Port: link state changed to 'down' on port 3
Dec 18 23:19:11 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:20 Port: link state changed to 'down' on port 3
Dec 18 23:19:22 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:23 Port: link state changed to 'down' on port 3
Dec 18 23:19:29 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:30 Port: link state changed to 'down' on port 3
Dec 18 23:19:32 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:34 Port: link state changed to 'down' on port 3
Dec 18 23:19:36 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:38 Port: link state changed to 'down' on port 3
Dec 18 23:19:40 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:41 Port: link state changed to 'down' on port 3
Dec 18 23:19:43 Port: link state changed to 'up' (100M-F) on port 3
Dec 18 23:19:43 Port: link state changed to 'down' on port 3
Dec 18 23:19:46 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:19:47 Port: link state changed to 'down' on port 3
Dec 18 23:19:49 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:20:12 Port: link state changed to 'down' on port 3
Dec 18 23:20:14 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:20:16 Port: link state changed to 'down' on port 3
Dec 18 23:20:18 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:22:34 Port: link state changed to 'down' on port 3
Dec 18 23:22:36 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:22:38 Port: link state changed to 'down' on port 3
Dec 18 23:22:40 Port: link state changed to 'up' (1G) on port 3
Dec 18 23:23:27 Port: link state changed to 'down' on port 3
Dec 18 23:23:29 Port: link state changed to 'up' (1G) on port 3


This is the exact problem I have been having but nothing has been done about it, im glad to see other users seeing the same issues I have been.

viewtopic.php?f=17&t=3258