Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Service port types

Fair enough, Glen

I was focusing perhaps too much on the legacy implementations where the ONU would typically have a single XGMII-type service port, which would then have to split externally to the ONU silicon.

The new drawing is much clearer, I agree. One would not even need to deal with the dual media port types, because it would be enough (model-wise) to terminate one service port on a pluggable cage and another service port on RJ45 interface. The only thing this does not support is maintaining the same provisioning index, i.e., the pluggable cage and RJ45 look like two different ports. There was some beauty in being able to provision the port independently of what media is being used. With the commoditization of copper SFP modules perhaps it is not a problem at all anymore.

Marek

On 7/1/2021 12:25 PM, Glen Kramer wrote:

Hi Marek,

 

Can you explain this model more? It is confusing because the MAC client per 1904.4 architecture is exactly an embedded managed Ethernet switch.

 

 

The above picture is a more-detailed version of this DPoE picture

 

 

I think what you said below is that one of service ports (or SF in DPoE picture) on the user side of 802.1 switch will be connected to another embedded switch. What would be achieved by connecting the two switches in serial? What cannot be done with one switch that can be done with two switches?

 

Thanks,

Glen

 

 

From: Marek Hajduczenia <mxhajduczenia@xxxxxxxxx>
Sent: Thursday, July 1, 2021 6:17 AM
To: 'Glen Kramer' <000006d1020766de-dmarc-request@xxxxxxxxxxxxxxxxx>; stds-1904-4-TF@xxxxxxxxxxxxxxxxx
Subject: RE: Service port types

 

The only thing that comes to mind (and I hate to throw it in) is the case of a service port connecting to an embedded Ethernet switch. In this case, a single service port would be “split” into multiple UNI ports. The question on how to address individual UNIs is a separate challenge, since from the ONU perspective, they all look like a single service port. However, this is precisely what drove the addition of aPhyType attribute to address the case of a single service port terminating on a 1:2 Ethernet switch, where one UNI port is implemented as a pluggable cage and another UNI port as a standard RJ45 jack.

 

If we want to support such a scenario, we also need to consider if we want to be able to address individual UNI ports behind such an embedded Ethernet switch or not. It seems like a bit of a stretch for our scope of work, but on the other hand it is still part of an ONU 😊

 

Marek

 

From: stds-1904-4-tf@xxxxxxxxxxxxxxxxx <stds-1904-4-tf@xxxxxxxxxxxxxxxxx> On Behalf Of Glen Kramer
Sent: Wednesday, June 30, 2021 5:51 PM
To: stds-1904-4-TF@xxxxxxxxxxxxxxxxx
Subject: Service port types

 

Picking up the discussion from the last meeting on service port types.

 

These are the types currently defined in DPoE and SIEPON:

 

 

Q1: We decided to add another type for UNI_port. Also Ryan mentioned we need to add another type. What was this type?

 

Q2: Besides UNI ports, are there any service port types (i.e., eSAFE entities) that may exist as multiple instances in an ONU?

For example, can there be more than one eMTA or eDVA per ONU/CM?

 

Thanks,

Glen

To unsubscribe from the STDS-1904-4-TF list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-1904-4-TF&A=1

To unsubscribe from the STDS-1904-4-TF list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-1904-4-TF&A=1