This article explains the behavior observed on the GRE interface/port number allocation when bandwidth changes are done on tunnel services.

Example: Tunnel services are enabled without any manual bandwidth configuration as below:

lab@MX> show configuration chassis    
fpc 0 {
    pic 0 {
        tunnel-services;
    }
}

The gr interface below was created with the interface convention as gr-0/0/0 (Port number 0 was created automatically on the PIC 0 of the FPC slot 0):

root@R1_re> show interfaces terse | match gr- 
gr-0/0/0                up    up

The bandwidth is manually defined as 1g:

root@R1_re# set chassis fpc 0 pic 0 tunnel-services bandwidth 1g

root@R1_re> show interfaces terse gr-0/0/0 
error: device gr-0/0/0 not found

The interface which was previously created with port number as 0 is not found. Any services running over the gr-0/0/0 will be impacted.

The bandwidth configured for the tunnel services defines the port number allocated for the services.

When the Bandwidth is specified as 1g, the port number is always  10 . So, it would be  gr-0/0/10 . For the remaining bandwidth values it would be port number 0.

New port number has been allocated for the tunnel interface as shown below:

root@R1_re> show interfaces terse | match gr- 
gr-0/0/10               up    up

Whenever changes are done for the bandwidth on the tunnel interface, make sure to do it during the scheduled MW to avoid the impact.