BGP route preference question

[David Schwartz]

I'm drawing a blank.

This router has always compare meds and missing as worst on. BGP
maximum paths is 4

router#show ip bgp X.X.X.X
BGP routing table entry for X.X.X.X/18, version X
Paths: (3 available, best #1)
Advertised to non peer-group peers:
X.X.X.X
A B C
Q from R (V)
Origin IGP, metric 2224, localpref 100, valid, external, best, ref
2
Dampinfo: penalty 412, flapped 1 times in 00:04:24
A B C
S from T (U)
Origin IGP, metric 2240, localpref 100, valid, external, ref 2
Dampinfo: penalty 617, flapped 4 times in 00:39:49
D E F
W from X (Y)
Origin IGP, metric 0, localpref 100, valid, external, ref 2

Now, what I don't understand is why the "best" route is the one with
the higher metric. If 'bgp always-compare-med' is enabled, shouldn't the
oldest route with the lowest metric, same localpref, and same AS path
length win? The '0' metric is _not_ a missing metric, it's a zero
metric.

DS

[David Schwartz]

aaronw wrote:

> Not sure about this one... what if you take away one of the paths and
> then put it back, does it revert or does it stay at a 'new' path?
>
> aw

The best path had been flapping, so obviously taking that path away and
putting it back didn't help. The path that should have become the best
path had been stable, and there's no good way for me to test what would
happen if that path flapped.

DS