On Wed, Nov 19, 2008 at 11:40 AM, Geoff Huston <[EMAIL PROTECTED]> wrote:
>
> On 20/11/2008, at 5:16 AM, Stephen Stuart wrote:
>>>
>>> This should probably not be a common case, but we do see it from time to
>>> time. It would be useful for us to be able to have this data.
>>
>> Data regarding the state of configured peers are already available by
>> other means (the BGP 4 MIB described in RFC1657); correlating with
>> that data tells us that a peer has supplied no updates
>> (BGP4-MIB::bgpPeerInUpdates.A.B.C.D). If the update count is non-zero
>> and there are still no BMP messages indicating that the peer supplied
>> a prefix, then it's likely attributable to "state changes in the
>> interim" for that prefix.
>
> having access to a feed of BGP data for research purposes and SNMP access to
> the bgp speaker's MIB are invariably two entirely different concepts,

On the other hand, an application built to use a feed of BGP data for
decision support almost certainly has other sources of operational
data, such as BGP MIB data, readily available. At least that's the
case for me.

> unfortunately, so while it may be in the MIB that does not mean that a BMP
> client has a prayer of ever access that mib!
>
> So, yes, it would be useful to have this data inband within the BMP feed.

As noted previously, SR messages can signal the existence of a peer,
and the monitoring application can use that information to determine
that a peer exists which has not supplied updates.

Stephen
_______________________________________________
GROW mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/grow

Reply via email to