On an ACX7509 equipped with a JNP‑FPC‑16C line card in Slot 1, two active 100G interfaces (et‑1/0/0 and et‑1/0/1) experienced a loss of all IP connectivity while the physical link state remained up. The disruption was caused by a hardware port‑group behavior associated with PR1898908, where a down port within the same port group can trigger a PHY reset during events such as RE failover or picd recovery. This reset impacts all ports in the group, including operational interfaces. A manual FPC restart restored connectivity.
Interfaces et‑1/0/0 and et‑1/0/1 stayed up/up but had no IP reachability
No ping on P2P link
ISIS adjacency down
RSVP sessions down
LDP neighbors down
MPLS forwarding impacted
Services restored only after manual FPC reboot
Workarround:
Mark all unused ports on the JNP‑FPC‑16C as unused to prevent down-state ports from causing PHY resets in the shared port group.
unused
Example:
set interfaces et-1/0/2 unused
set interfaces et-1/0/3 unused
set interfaces et-1/0/4 unused
Install the fix reported on PR: https://prsearch.juniper.net/problemreport/PR1854658