[c-nsp] MPLS TE reoptimization pains?

[Pete Templin]

Greetings,

We're using MPLS Traffic Engineering in a strategic fashion (not full 
mesh, but a mesh of all appropriate traffic patterns).  We recently 
upgraded core routers in several POPs from 7507 to 12008.  We're seeing 
strange behavior (routing loops, packet loss, etc.) for ~30 seconds at 
times when TE tunnels appear to be reoptimizing.  Current IOS is 
12.0(27)S3 on 7507s, 12.0(27)S5 on GSRs.

Anyone know of any issues we might be running into, and/or if there'd be 
any major harm done in tuning down the reoptimization period to a much 
lower value (8 hours?)?

Thanks,

Pete Templin