[j-nsp] v6 BGP Policy Bug?

[Richard A Steenbergen]

On Wed, Apr 15, 2009 at 11:16:56AM +0800, Mark Tinka wrote:
> That's right, which is what we see for the export policies 
> we applied to our IS-IS deployment. I guess the issue we're 
> seeing with wrong behaviour for the v4 BGP export policy 
> kind of threw us off with what to expect for v6.
> 
> Let me debug this. Will come back if I find anything 
> interesting.

Are you talking about an explicit "then accept" final term, or an
implicit default-action accept? Once upon a time I saw a nice table of
the protocol-specific default-action defaults, but I can't find it now
as Juniper's search is returning 99% xml results. You know I'd give good
money for a checkbox to block xml reference guides from search results
on the website, they tend to hog the results and make things impossible
to find at the most inopportune times.

On the subject of impossible dreams that would make troubleshooting
these kinds of things easier, it would be really handy if "test policy"
supported policy chains and logical-routers. Actually it would be nice
if you could change the output from something other than "brief", and
control the input with something more than a single prefix doing
"orlonger" matching too. I wonder why they didn't just do this as part
of "show route" so you get consistent features?

-- 
Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)