Cisco Cisco IOS Software Release 12.0(23)S
![Cisco](https://files.manualsbrain.com/attachments/7380d0050044647c30f5c24bbbf5d0c0b6d9bb84/common/fit/150/50/faa183d287233c52228cfea3dbc2a127fe780f60564fcb0955d9c3d1cd23/brand_logo.png)
MPLS Traffic Engineering (TE)—Link and Node Protection, with RSVP Hellos Support
Bandwidth Protection Considerations
88
Cisco IOS Release 12.0(23)S
link failures are in practice independent with high probability. This “independent failure assumption” in
combination with backup tunnels signaled without explicit bandwidth reservation enables efficient
bandwidth sharing that yields substantial bandwidth savings.
combination with backup tunnels signaled without explicit bandwidth reservation enables efficient
bandwidth sharing that yields substantial bandwidth savings.
Backup tunnels protecting the sub-pool traffic do now draw bandwidth from any pool. Primary traffic
using the global pool can use the entire global pool, and primary traffic using the sub-pool can use the
entire sub-pool. Yet, sub-pool traffic has a complete bandwidth guarantee if there is a single link failure.
using the global pool can use the entire global pool, and primary traffic using the sub-pool can use the
entire sub-pool. Yet, sub-pool traffic has a complete bandwidth guarantee if there is a single link failure.
A similar approach can be used for node and SRLG protection. However, the decision of where to put
the backup tunnels is more complicated because both node and SRLG failures effectively result in the
simultaneous failure of several links. Therefore, the backup tunnels protecting traffic traversing all
affected links cannot be computed independently of each other. The backup tunnels protecting groups of
links corresponding to different failures can still be computed independently of each other, which results
in similar bandwidth savings.
the backup tunnels is more complicated because both node and SRLG failures effectively result in the
simultaneous failure of several links. Therefore, the backup tunnels protecting traffic traversing all
affected links cannot be computed independently of each other. The backup tunnels protecting groups of
links corresponding to different failures can still be computed independently of each other, which results
in similar bandwidth savings.
Signaled Bandwidth versus Backup-Bandwidth
Backup-bandwidth is used locally (by the router that is the head-end of the backup tunnel) to determine
which, and how many, primary LSPs can be rerouted on a particular backup tunnel. The router ensures
that the combined bandwidth requirement of these LSPs does not exceed the backup-bandwidth.
which, and how many, primary LSPs can be rerouted on a particular backup tunnel. The router ensures
that the combined bandwidth requirement of these LSPs does not exceed the backup-bandwidth.
Therefore, even when the backup tunnel is signaled with zero bandwidth, the backup-bandwidth must be
configured with the value corresponding to the actual bandwidth requirement of the traffic protected by
this backup tunnel. Unlike the case when bandwidth requirements of the backup tunnels are explicitly
signaled, the value of the signaled bandwidth (which is zero) is not the same value as the
backup-bandwidth.
configured with the value corresponding to the actual bandwidth requirement of the traffic protected by
this backup tunnel. Unlike the case when bandwidth requirements of the backup tunnels are explicitly
signaled, the value of the signaled bandwidth (which is zero) is not the same value as the
backup-bandwidth.