[c-nsp] Best practices for Cat6500

[Phil Mayers]

On 01/11/10 15:37, Phil Mayers wrote:
> On 01/11/10 14:09, Phil Mayers wrote:
>
>>
>> That seems like a bit of a dangerous "feature" to introduce with no
>> warning, if that's what's really happening; what if you have a complex
>> multi-edit sequence to go through when rebuilding a route-map?
>>
>> I'm not seeing the behaviour - a quick test on an eBGP peer locally
>> doesn't do this on our SXI4a box. Can you show some BGP config?
>
> Bah, my mistake - now I see it. It happens when you *edit* a route-map
> on a peer, but not when you first add the route-map.
>
> This is seriously crummy; if you do this:
>
> route-map TEST 10
>     set community 65000:1
> router bgp 65000
>     neighbour 192.168.1.1 route-map TEST in
> route-map TEST 5
>     ! at this instant, the community disappears from the route!!!
>
> This was not how I thought IOS was defined to behave. Can anyone confirm
> this?

This is a bug, CSCtf64231, and SXI5 is now out I see, claiming to fix it:

"Inbound route-map change shouldn't be effective immediately"

Pfft, yeah! That's one way to put it!