Geoff,
On Nov 13, 2008, at 12:23 PM, Geoff Huston wrote:
I'm a little confused however by some of the details. It appears
that BMP is a method to tap into a peer session and the local BGP
speaker can be configured to send a copy of all received updates
from a BGP peer to a monitor point using BMP, including the time
when these updates were received by the local BGP speaker.
However in section 3 the text states: "If there are state changes in
the interim for that prefix, it is acceptable that the router
generate the final state of that prefix to the monitoring station. "
What I'm confused about is: what timestamp should be used in such a
case?
The timestamp should reflect the time the final state transition
occurred. So for example if the router saw:
time = 1, state = 10/8 via 1.1.1.1
time = 2, state = 10/8 withdrawn
time = 3, state = 10/8 via 2.2.2.2
And generated the BMP message 10/8 via 2.2.2.2, it would be
timestamped with time = 3.
And does this "final state" reflect the final state in the logical
Adj-RIB-In for this BGP peer, or the final state according to the
local Loc-RIB?
The former (Adj-RIB-In) unless the implementation is unable to support
Adj-RIB-In and uses Loc-RIB instead as discussed in Section 3
paragraph 2.
--John
_______________________________________________
GROW mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/grow
