I have an SRX running 12.3X48-D75.4, I have 'no-traps' configured under the st0 interface, yet when I delete a subinterface of that int, a linkDown trap is sent for the deleted subint. Is there a way to prevent linkDown traps being send on deleted interfaces/subinterfaces?
Hoping you are doing great.
Could you kindly share the xls output of the events shown on space?
If you see that the ifindex is N/A it not necessarily indicates it corresponds to the st- interface.
Please let me know if you can narrow down answering the following questions:
1. Does this trap happen only on clusters without traffic?2. Is it only happening on clusters running 12.3X48-D75.4?
3. Can you temporary change the syslog message logging to "set system syslog file messages any any". After some time, after we have several new occurrences of these traps, collect the /var/log and log from the space server? The goal is to better analyze logs and see if maybe in the logs we will see correlating events to understand where the traps come from.
4. Can you also collect logs from the space on traps of similar kind, like interface down? Or maybe these are the only ones?
I will be waiting for your response.
Hello, thanks for the reply. I'm not running space so I cannot show that output. I'm looking at the trap and it clearly references the subint that was just deleted.
1. No cluster is invovled, just a standalone SRX
2. No cluster, but I have not tested on different JUNOS versions, I can do that.
3. I can do this, I can also provide the actual trap if you like
4. I'm not running space
"3. I can do this, I can also provide the actual trap if you like" yes..
Also please share complete SNMP configuration from the DUT
I did some more experimentation, and it looks like the 'units' under the st0 hierarchy don't inherit the 'no-traps' config. You have to have 'no-traps' configured under every unit to not have a trap send when that unit goes up/down or is created/deleted.