Yes, the bandwidth command does limit the throughput of the tunnel interfaces (lt, gr, ip...).
Different lines cards have a maximum amount of bandwidth (packet processing capacity on the PFEs) that can be used for tunnel services. When you don't specify the bandwidth , tunnels can use up to that maximum.
For example, the maximum tunnel bandwidth for an MPC4E is 60G, and you can configure 1, 10, 20, 30 or 40G. If you don't configure the bandwidth, your tunnel interfaces can use up to 60G. When you do configure the bandwidth, you are limiting the tunnel interfaces to that value.
Regards, ------------------------------
Yasmin Lara
Juniper Ambassador
JNCIE-SP, JNCIE-ENT, JNCIE-DC, JNCIE-SEC
JNCDS-DC, JNCIA-DevOps, JNCIP-CLOUD, CCNP-ENT
------------------------------
Original Message:
Sent: 11-18-2020 03:32
From: Unknown User
Subject: What does tunnel-services bandwidth configuration statement do?
I have a question on tunnel services bandwidth. One thing I picked up is that the amount of bandwidth configured determines the virtual-interface numbering. The bandwidth that is specified determines the port number of the tunnel interfaces that will be created. When we specify a bandwidth of 1g, the port number is always 10. When we specify any other bandwidth, the port number is always 0.
Does, this statement also restrict the throughput of all services which require tunnel-services to the specified bandwidth? That would mean:
- PIM on FHR and RP
- GRE tunnels
- IPIP tunnels
- Logical tunnels lt-x/x/x
- Lsi interface for VPLS and VPWS