Hello,
Could somebody clarify the following situation, is it bug or not?
I have OOB management through VME interface for VC constructed on QFX5120-32C?
Both em0 ports of the VC connected to the additional management switch and they status is UP. My VC is accessible.
Now I decide administratively shutdown port on additional management switch to the QFX5120-32C direction with role "Master". The role is important.
After that, I expect, that my VC should be accessible via VME interface through em0 interface of the QFX5120-32C with role "Backup", but is not true.
If I will change role via CLI between QFX5120-32C, VC begin again accessible.
>request chassis routing-engine master switch no-confirm
Junos: 20.4R3.8 limited
BEFORE shutdown:
> show interfaces terse | match em0
em0 up up
em0.0 up up eth-switch
> show interfaces terse | match vme
vme up up
vme.0 up up inet 10.128.2.2/24
> ping 10.128.2.1 source 10.128.2.2 count 1
PING 10.128.2.1 (10.128.2.1): 56 data bytes
64 bytes from 10.128.2.1: icmp_seq=0 ttl=64 time=41.471 ms
AFTER shutdown port to master role QFX5120-32C:
> show interfaces terse | match em0
em0 up down
em0.0 up down eth-switch
> show interfaces terse | match vme
vme up up
vme.0 up up inet 10.128.2.2/24
> ping 10.128.2.1 source 10.128.2.2 count 1
PING 10.128.2.1 (10.128.2.1): 56 data bytes
--- 10.128.2.1 ping statistics ---
1 packets transmitted, 0 packets received, 100% packet loss
AFTER switchover role Master/Backup on QFX5120-32C
> show interfaces terse | match em0
em0 up up
em0.0 up up eth-switch
> show interfaces terse | match vme
vme up up
vme.0 up up inet 10.128.2.2/24
> ping 10.128.2.1 source 10.128.2.2 count 1
PING 10.128.2.1 (10.128.2.1): 56 data bytes
64 bytes from 10.128.2.1: icmp_seq=0 ttl=64 time=4.811 ms
------------------------------
ALEXANDR GURBO
------------------------------