On Fri, 4 Dec 2015, Daniel Walton wrote:
I think here we need to mention that if the two paths are both eBGP and
compare router-id is NOT configured, then we prefer the older of the two. I
think that will help with explaining the router-id knob later.
Ack. Don't know how I didn't include that! Something like:
"@item 10 Already-selected external check
Where both routes were received from eBGP peers, then prefer the route
which is already selected. Note that this check is not applied if
@ref{bgp bestpath compare-routerid} is configured."
Note that the code doesn't obviously read that way:
if (! bgp_flag_check (bgp, BGP_FLAG_COMPARE_ROUTER_ID)
&& new_sort == BGP_PEER_EBGP
&& exist_sort == BGP_PEER_EBGP)
{
if (CHECK_FLAG (new->flags, BGP_INFO_SELECTED))
return 1;
if (CHECK_FLAG (exist->flags, BGP_INFO_SELECTED))
return 0;
}
It probably doesn't make a difference, because one can likely rely on
only route being SELECTED at a time, but I had to spend a few seconds
telling myself that. It'd be easier to read if that dependency was made
explicit and the code made self-documenting:
if (! bgp_flag_check (bgp, BGP_FLAG_COMPARE_ROUTER_ID)
&& new_sort == BGP_PEER_EBGP
&& exist_sort == BGP_PEER_EBGP)
&& (CHECK_FLAG (new->flags, BGP_INFO_SELECTED)
!= CHECK_FLAG (exist->flags, BGP_INFO_SELECTED)))
{
...
}
+@deffn {BGP} {bgp bestpath compare-routerid} {}
+@anchor{bgp bestpath compare-routerid}
+
+Ensure that when comparing routes where both are equal on most
metrics,
+including local-pref, AS_PATH length, IGP cost, MED, that the
tie is broken
+based on router-ID.
+
+If this option is enabled, then the external-age check, where
already
+selected eBGP routes are preferred, is skipped.
+
I think one sentence is mainly what need here....mabye change "where
already selected eBGP routes are preferred" to "where the oldest eBGP
route is preferred".
Or to:
"then the already-selected external check, where an older selected eBGP
route is preferred"
?
I.e. the oldest route need not be the selected one, so my original
tendency to call it an 'age' check seems a bit misleading. ;)
+If a route has an ORIGINATOR_ID attribute because it has been
reflected,
+that ORIGINATOR_ID will be used. Otherwise, the router-ID of
the peer the
+route was received from will be used.
I would chop the paragraph above since this is already mentioned in the
bestpath explanation for router-id.
I'd be more likely to chop it from the decision process thing, or just
keep both. A lot of readers will be using it as a command reference and
not reading it linearly - redundancy may benefit such readers.
Ack on everything else. Thanks for putting this together Paul
Welcome. :)
We possibly have slightly different philosophies on this topic perhaps,
but there's a good few details, trade-offs and nuances to it, and it's
nice to at least get the agreed ones down to and increase the shared
understanding.
MED... such a seemingly simple little attribute, but it's caused a lot of
work for people. ;)
There's some other tweaks I might make, based on other previous comments
of yours, will follow up.
regards,
--
Paul Jakma, HPE Networking, Advanced Technology Group
Fortune:
If it pours before seven, it has rained by eleven.
_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev
