Re: [netconf] [core] some comments on netconf-adaptive-subscription

[Qin Wu <[e-mail]>]

-----邮件原件-----
发件人: Michael Richardson [mailto:[e-mail]] 
发送时间: 2021年6月26日 0:58
收件人: Qin Wu <[e-mail]>; [e-mail]; [e-mail]
主题: Re: [core] some comments on netconf-adaptive-subscription

Qin Wu <[e-mail]> wrote:
    mcr> I didn't finish my thought here.  Specifically, it seems that the CORE
    mcr> WG is doing a lot of stuff with CoAP Observes and SENML and all sorts
    mcr> of interesting triggers.

    Qin> data modeling language such as YANG can be common work for both CORE
    Qin> WG and NETCONF WG, But they may choose different transport protocol,
    Qin> such as NETCONF, RESTCONF, HTTP 2.0, COAP, MQTT, COMI Secondly, I
    Qin> think it will be nice to separate constrained device management from
    Qin> resource unconstrained network device management, Constraint device
    Qin> as IoT device can be temperature sensor, current, voltage sensor
    Qin> which can be seen as affiliated devices pertaining to resource
    Qin> unstrained network device.

temperature, current are critical values for routing equipment, aren't they?

[Qin]: My understanding is yes, there is hardware YANG model which can be used to collect these inventory data associated with the routing equipment
See RFC8348.

    Qin> In data center scenarios, we do see a lot of such affiliated devices or
    Qin> sensors such as air conditioner, water leakage alarm sensor deployed
    Qin> together with data center switch devices.  These IoT devices only
    Qin> use MQTT/COAP/HTTP to communicate with their IoT platform, for data
    Qin> center switch devices, you may still rely on network management
    Qin> protocols such as NETCONF, SNMP, gRPC, etc.  Do you see this
    Qin> differently? Michael?

It seems that we ought to seek convergence here.
That's what I'm saying.  That any split here is for historical reasons, and does not make sense going forward.

[Qin]: Tend to agree, unified data modeling or semantics interoperability may be the solution. One of example is to
Build digital twin data model and models sensor data and network device health data together.

--
Michael Richardson <[e-mail]>   . o O ( IPv6 IøT consulting )
           Sandelman Software Works Inc, Ottawa and Worldwide