On Dec 1, 2014, at 7:42 PM, Tim Wicinski <tjw(_dot_)ietf(_at_)gmail(_dot_)com>
wrote:
On 12/1/14 7:38 PM, Dean Bogdanovic wrote:
On Dec 1, 2014, at 7:15 PM, "Thomas D. Nadeau"
<tnadeau(_at_)lucidvision(_dot_)com>
wrote:
On Dec 1, 2014:7:10 PM, at 7:10 PM, Randy Bush <randy(_at_)psg(_dot_)com>
wrote:
the “MIB Doctor” model we are using is not going to scale out to the
numbers of Yang models that are in need of advice or review, nor will
be scale in terms of progressing models through the IETF’s RFC
process. The fact is that we simply do not have enough Yang Doctors
to cover all of the models in question, despite our best efforts.
is this a sign that we do not have enough medical care or that we are
unleashing an unarchitected epidemic of overly device-specific snmp with
the syntax changed?
Speaking from my own personal opinion, it seems that operators are
finding this stuff useful and are
demanding that people build products with it.
I agree with you on this one. See it from multiple operators in Americas and
EMEA (haven't seen in APAC personally). Operators have been translating
their service configurations into YANG and are looking how to adapt their
service models to proprietary operator models. For them, having standard
config models, would make it much more easier to translate their service
models and let the vendors worry about translation from standard to
proprietary models.
Dean
What is the IETF when many of the newer networking companies do not find YANG
worth investing in?
As far I know Arista has their own proprietary schema language (as many other
companies have it too), that describes their data model. Arista uses that
schema to generate binding to be able to access all objects generated from the
data model from python.
If they use their proprietary or use IETF YANG to describe the data model is a
decision companies have to make by them self. There is a benefit to have a
standardized language that everyone is able to understand.
I'm thinking of Arista as a good example of a networking vendor that (i've
been told) feels supporting YANG makes them less agile than their APIs.
this is one part I don't understand. Why adding another language would make
them less agile?
Dean
tim