We tried cutting in another router and needed to change the loopback IP.

It appears that routes are getting rejected. The users are not able to reach internet.

• Customer changed the lo0 IP from a.b.c.d to a.b.c.e for MX480 which host ISP links & DNS
• They integrated MX304 with lo0 IP as a.b.c.d, as they wanted to cut over to this one from MX480
• But post the changes, the internet reachability got impacted for users behind the MX
• Later they blocked all uplinks of MX304 & isolated it
• But still the issue was seen with MX480 & all traffic to this specific MX480 seems to impacted from others

The below checks were performed on the issue. The MX480 was running old version & hence our expectation was to recover the traffic.

• We do not have route flaps in MX480 & there were no abnormal logs in device
• We were checking on this & one another PE, were they have connected a laptop for testing & the routes learned at both these PEs are fine
• The above PE is directly connected to this MX480
• The route did not have any errors on both side
• The PNH reachability is fine from both
• The traceroute do not even exit this device, and for monitor traffic also we do not see any info being passed
• The LSPs/LDP neighbors are up & the inet.3 table is populated correctly
• Hence customer went ahead with rollback to original config, but the issue was still seen.
• RE switchover did not help & we also tried RE reboots without success.
• The complete reboot of both REs at once cleared the issue & we are suspecting a rpd corruption here.