[CCAMP] 答复: Webex meeting changed: Optical impairments draft

Fatai Zhang <zhangfatai@huawei.com> Fri, 06 November 2020 07:01 UTC

Return-Path: <zhangfatai@huawei.com>
X-Original-To: ccamp@ietfa.amsl.com
Delivered-To: ccamp@ietfa.amsl.com
Received: from localhost (localhost []) by ietfa.amsl.com (Postfix) with ESMTP id A27ED3A0E18; Thu, 5 Nov 2020 23:01:14 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.796
X-Spam-Status: No, score=-1.796 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Received: from mail.ietf.org ([]) by localhost (ietfa.amsl.com []) (amavisd-new, port 10024) with ESMTP id SZ4DBrjt2m3U; Thu, 5 Nov 2020 23:01:10 -0800 (PST)
Received: from frasgout.his.huawei.com (frasgout.his.huawei.com []) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id EE4313A07EA; Thu, 5 Nov 2020 23:01:09 -0800 (PST)
Received: from fraeml713-chm.china.huawei.com (unknown []) by frasgout.his.huawei.com (SkyGuard) with ESMTP id 4CSB8L6jCFz67HTm; Fri, 6 Nov 2020 14:59:30 +0800 (CST)
Received: from fraeml713-chm.china.huawei.com ( by fraeml713-chm.china.huawei.com ( with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.1913.5; Fri, 6 Nov 2020 08:01:06 +0100
Received: from DGGEML403-HUB.china.huawei.com ( by fraeml713-chm.china.huawei.com ( with Microsoft SMTP Server (version=TLS1_0, cipher=TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA) id 15.1.1913.5 via Frontend Transport; Fri, 6 Nov 2020 08:01:06 +0100
Received: from DGGEML530-MBX.china.huawei.com ([]) by DGGEML403-HUB.china.huawei.com ([fe80::74d9:c659:fbec:21fa%31]) with mapi id 14.03.0487.000; Fri, 6 Nov 2020 15:00:53 +0800
From: Fatai Zhang <zhangfatai@huawei.com>
To: Daniele Ceccarelli <daniele.ceccarelli@ericsson.com>, Italo Busi <Italo.Busi@huawei.com>, 'Gert Grammel' <ggrammel@juniper.net>, "ccamp-chairs@ietf.org" <ccamp-chairs@ietf.org>, "ccamp@ietf.org" <ccamp@ietf.org>
Thread-Topic: [CCAMP] Webex meeting changed: Optical impairments draft
Thread-Index: AQHWrhFEc+86QuZ2hkCgLsn8z48T9amvzTmAgAEg6QCAAaxfAIABnaoAgABFgICABjcPEA==
Date: Fri, 6 Nov 2020 07:00:52 +0000
Message-ID: <F82A4B6D50F9464B8EBA55651F541CF8ADF9867C@dggeml530-mbx.china.huawei.com>
References: <E3FA7667-4767-4DC6-9F08-210FB254F629@juniper.net> <af767563835443ce889b7b05afc5b87d@huawei.com> <379ed6ba-a3da-82a2-d867-148c8070da9a@nokia.com> <F52906B1-B3FB-4EB0-A2CA-D1B2A453C38A@juniper.net> <cc331482dc5748d79fd1f7b06de82ce6@huawei.com> <HE1PR07MB4156902CC74550758C284C04F0100@HE1PR07MB4156.eurprd07.prod.outlook.com>
In-Reply-To: <HE1PR07MB4156902CC74550758C284C04F0100@HE1PR07MB4156.eurprd07.prod.outlook.com>
Accept-Language: zh-CN, en-US
Content-Language: zh-CN
x-originating-ip: []
Content-Type: multipart/alternative; boundary="_000_F82A4B6D50F9464B8EBA55651F541CF8ADF9867Cdggeml530mbxchi_"
MIME-Version: 1.0
X-CFilter-Loop: Reflected
Archived-At: <https://mailarchive.ietf.org/arch/msg/ccamp/4VJRh3pYTbrpr-iNTtS86t0vTj8>
Subject: [CCAMP] =?utf-8?b?562U5aSNOiAgV2ViZXggbWVldGluZyBjaGFuZ2VkOiBP?= =?utf-8?q?ptical_impairments_draft?=
X-BeenThere: ccamp@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Discussion list for the CCAMP working group <ccamp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/ccamp>, <mailto:ccamp-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/ccamp/>
List-Post: <mailto:ccamp@ietf.org>
List-Help: <mailto:ccamp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ccamp>, <mailto:ccamp-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 06 Nov 2020 07:01:15 -0000

Hi all,

I think there are too many options (three!) , which introduces too much flexibility and complexity to this document and the real world.

If “only application codes” can work, why the others are necessary?

I think we can try if we could keep only one option and remove the other options to keep it simple , because we could treat e.g., “organization mode” as proprietary implementation since it is really proprietary  among some organization.


发件人: Daniele Ceccarelli [mailto:daniele.ceccarelli@ericsson.com]
发送时间: 2020年11月2日 23:51
收件人: Italo Busi <Italo.Busi@huawei.com>om>; 'Gert Grammel' <ggrammel@juniper.net>et>; ccamp-chairs@ietf.org; ccamp@ietf.org
抄送: Dieter Beller <Dieter.Beller@nokia.com>
主题: RE: [CCAMP] Webex meeting changed: Optical impairments draft


Sorry to jump in but I’m having hard times to understand what the issue is. What is wrong with the actual modes? I asked if it was possible to merge 2 of them and go down to 2 but it was proven not possible.

Moreover this is no longer just author’s agreement since the draft is a WG document.
What italo says below is correct. We decided that all the modes are optional, hence the model needs to be self-consistent and all permutations should be allowed.

But once again, it’s not clear to me what the issue is, could you please elaborate it a bit?


From: Italo Busi <Italo.Busi@huawei.com<mailto:Italo.Busi@huawei.com>>
Sent: den 2 november 2020 12:42
To: 'Gert Grammel' <ggrammel@juniper.net<mailto:ggrammel@juniper.net>>; ccamp-chairs@ietf.org<mailto:ccamp-chairs@ietf.org>; ccamp@ietf.org<mailto:ccamp@ietf.org>
Cc: Dieter Beller <Dieter.Beller@nokia.com<mailto:Dieter.Beller@nokia.com>>
Subject: RE: [CCAMP] Webex meeting changed: Optical impairments draft

Hi Gert,

Just an high level comment: the initial agreement between the co-authors was that to define three modes (application code, organization mode and explicit mode) all as being optional.

Therefore, IMHO, the YANG model for each mode should be self-consistent to give implementers the freedom to implement either:
-          Only application codes
-          Only organization modes
-          Only explicit modes
-          Both application codes and organization modes
-          Both application codes and explicit modes
-          Both organizational modes and explicit modes
-          All options (application code, organizational mode and explicit mode)

Please find some detailed comments of mine in line below.

My 2 cents


From: Gert Grammel [mailto:ggrammel@juniper.net]
Sent: domenica 1 novembre 2020 12:02
To: ccamp-chairs@ietf.org<mailto:ccamp-chairs@ietf.org>; ccamp@ietf.org<mailto:ccamp@ietf.org>
Cc: Dieter Beller <Dieter.Beller@nokia.com<mailto:Dieter.Beller@nokia.com>>; Italo Busi <Italo.Busi@huawei.com<mailto:Italo.Busi@huawei.com>>
Subject: Re: [CCAMP] Webex meeting changed: Optical impairments draft

It would be good to tune the heat of the down by a few centigrade. Why spending time writing definitions, getting agreement  and rewriting again in less than 24h?  The core of the discussion is where to hang parameters in a YANG-tree in a consistent manner. How exciting is this?

The minimum we should strive for in IETF is to come up with a consistent model. To do so, it is good engineering practice to first agree on the problem. That said, I am happy that apparently the Case-A and Case-B description laid out in my previous mail turned out not to be contentious.

About the issue of “organizational mode”:

  1.  Case-A: AFAIK the most common and Transponder Model today is OpenConfig, which uses a single identifier for the operational-mode. The use is simple because it can be used as a matching criteria to make sure the two transponders at each end interoperate (given the OLS is within spec).
  2.  Case-B: proposes a series of additional criteria that must be met by adding explicit parameters such as frequency range, power-range etc. So the simple matching rule is no more sufficient and additional parameters must be checked to assess compatibility.
I think it is a fair ask to hear from the CCAMP-WG if we are happy to ditch the OpenConfig way and  implement Case-B.

[[Italo Busi] ] I am not an expert of OpenConfig but Dieter has already confirmed that the latest definition in github (https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/issues/29#issuecomment-719503150<https://protect2.fireeye.com/v1/url?k=44077b9a-1b9c42d9-44073b01-861fcb972bfc-e503d80e6efd81dd&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fissues%2F29%23issuecomment-719503150>) is aligned with existing implementations (I assume, based on his working experience, he was also considering OpenConfig implementations).

Anyhow, OpenConfig experts can provide their comments on this definition, if they think it is not correct.

About the issue of “where to put the explicit parameters”:

  1.  Case-A: puts all explicit parameters in the explicit mode and provides a leafref to Organizational modes. Explicit min/max parameters can be liked to an Organizational mode by referencing the organizational mode from the explicit mode. Organizational modes continue to be used just like Application codes as a matching condition.

[[Italo Busi] ] The references between the explicit mode and the application code/organization modes have been added as informative to represent the fact that an explicit mode can support zero, one or more application codes and/or zero, one or more organization modes. It was not intended to provide additional information wrt the application code or organization mode definitions and I think it should not be used in that way to avoid creating inconsistencies (see comment below).

  1.  Case-B: puts a subset of the parameters also defined in the explicit mode into the Organizational mode. By doing so we now have duplicated parameters in operational and explicit mode.
Structurally this creates a conflict, because it is not clear which definition takes precedence in a matching condition. It is this structural issue we need to address to move forward.

[[Italo Busi] ] I think you have spotted a potential technical issue with the current definition of the leafref between an explicit mode and an application code or organization mode but this is independent from the fact that explitic parameters are defined or not for the application mode.
This inconsistency could apply also to parameters that are implicitly defined by the organization mode (or even by an application code). For example, the YANG model allows an explicit mode which supports one type of FEC to claim support of an application code or of an organization model which (implicitly) requires a different type of FEC.
If the client requests the interface to operate according to a given application code or organization mode, I think that the expectation is that the FEC defined for that application code or organization mode is being used.

I think to resolve this issue we should either clarify the rules an implementation MUST follow to avoid reporting inconsistent information or some precedence rules (considering both implicit and explicit parameters) or remove the leafref.

About the parameter set of the “Organizational mode”:

  1.  Case-A: allows to use any set of explicit parameters to add more information to organizational modes as a reference

[[Italo Busi] ] The YANG model for application codes and organization modes should be self-consistent to supports implementations that do not implement the explicit-mode. Moreover, using the leafref to provide more information would be a potential source of inconsistency (see comment above)

  1.  Base-B: allows a small subset of parameters to be used to further define the organizational mode. This set is fixed as described in case-B.
As the selection criteria for the set of parameters has not been disclosed, it is difficult to understand whether the list remains such short or is growing much longer over time. It would be good to get the sense of the WG on that aspect too.

[[Italo Busi] ] According to the latest definition in github (https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/issues/29#issuecomment-719503150<https://protect2.fireeye.com/v1/url?k=c4b9f9f1-9b22c0b2-c4b9b96a-861fcb972bfc-7ad4277e89b62b88&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fissues%2F29%23issuecomment-719503150>) the current list is complete.

Those who have implemented organization modes (including OpenConfig experts) can provide their comment if they think this list is not complete.



From: CCAMP <ccamp-bounces@ietf.org<mailto:ccamp-bounces@ietf.org>> on behalf of Dieter Beller <Dieter.Beller@nokia.com<mailto:Dieter.Beller@nokia.com>>
Organization: Nokia
Date: Saturday, October 31, 2020 at 10:28
To: Italo Busi <Italo.Busi@huawei.com<mailto:Italo.Busi@huawei.com>>, "ccamp-chairs@ietf.org<mailto:ccamp-chairs@ietf.org>" <ccamp-chairs@ietf.org<mailto:ccamp-chairs@ietf.org>>, CCAMP <ccamp@ietf.org<mailto:ccamp@ietf.org>>
Subject: Re: [CCAMP] Webex meeting changed: Optical impairments draft

[External Email. Be cautious of content]

Thanks, Italo, for your contribution to the technical discussion!

I would like to emphasize that the text on GitHub ( see https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/issues/29#issuecomment-719503150<https://protect2.fireeye.com/v1/url?k=2c758499-73eebdda-2c75c402-861fcb972bfc-3975d7433728ea76&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fissues%2F29%23issuecomment-719503150>) tries
to provide the technical justification for the modeling approach taken for the organizational mode option - here is the text, which was referenced in Italo's mail:

Organizational Mode:
Organizations like operator groups, industry fora, or equipment vendors can define organizational modes, which will allow these organizations to make use of advanced transceiver capabilities going beyond existing standardized application codes. Such an organizational mode is identified by the organization-identifier attribute defining the scope and an operational-mode that is meaningful within the scope of the organization. Hence, the two attributes must always be considered together. Two transceivers are inter-operable, if they have at least one (organization-identifier, operational-mode) pair in common and if the supported carrier frequency and power attributes have a matching range. This is a necessary condition for path computation in the context of organizational modes. An operational mode is a transceiver preset (a configuration with well-defined parameter values) subsuming several transceiver properties including:
• FEC type
• Modulation scheme
• Encoding (mapping of bit patterns to symbols in the constellation diagram)
• Baud rate (symbol rate)
• Carrier bandwidth (typically measured in GHz)

The major reason for these transceiver presets is the fact that the attribute values typically cannot be configured independently and are therefore advertised as supported operational mode capabilities.
It is the responsibility of the organization to assign operational modes and to ensure that operational modes are unique and not ambiguous within the scope of the organization.
In addition to the transceiver properties subsumed by the operational mode, optical power and carrier frequency related properties are modeled separately, i.e., outside of the operational mode. This modeling approach allows transponders using different transceiver variants (e.g. optical modules) with slightly different power and/or frequency range properties to interoperate without defining separate operational modes. Different optical modules (pluggables) from different suppliers typically have slightly different input and output power ranges or may have slightly different carrier frequency tuning ranges.
The received channel power and the received total power are two parameters that can be measured by the receiver and can be provided by the transceiver in order to allow a controller to determine the expected performance of the end-to-end service taking into account the optical impairments along the path.

Moreover, implementations exist that are based on the same modeling approach and hence it is already field proven!

On 30.10.2020 17:14, Italo Busi wrote:

Hi Gert,

Let me try to look at this issue from a different perspective.

I think the objective of our work is to define a YANG model for the optical specification methodologies which are used in the field to represent the capabilities of the transponders.

I do not think that it is in the scope of our work to discuss the optical specification methodologies other than understanding how they could be used for path computation (ensuring interoperability between the sender and receiver) and modelled using YANG.

As far as I have understood from past discussion, we have agreed to support three optical specification methodologies which are used by existing optical interfaces implementations and we have named them as application code (ITU-T G.698.2), organizational mode and explicit mode.

After some rounds of discussions and questions for clarification, the latest proposed text to describe the behavior of the organization mode has been prepared on github:


I think this updated description has improved the clarity of how the operational mode are implemented and, in particular that:

"Two transceivers are inter-operable, if they have at least one (organization-identifier, operational-mode) pair in common and if the supported carrier frequency and power attributes have a matching range."

As a consequence, I think that the YANG model defined in "Case A" below does not work (since it does not provide enough information to check that " the supported carrier frequency and power attributes have a matching range ") while the one defined in "Case B" works properly.

This text describing the operational mode behavior has been provided by people who have implemented it.

Therefore, I think we can update the draft with the changes discussed in the last few weeks incorporating the definitions capturing our current understanding of how these optical specification methodologies (application code, organizational mode and explicit mode) behave and the YANG model aligned with our current understanding (i.e., based on "Case B" below).

Since we are not yet in WG LC stage, we still have plenty of time to update both the description and the YANG model based on the feedbacks we receive from data plane experts.

My 2 cents

Italo (as co-author of the draft)

-----Original Message-----

From: Gert Grammel [mailto:ggrammel@juniper.net]

Sent: giovedì 29 ottobre 2020 17:34

To: ccamp-chairs@ietf.org<mailto:ccamp-chairs@ietf.org>; ccamp@ietf.org<mailto:ccamp@ietf.org>

Subject: Re: [CCAMP] Webex meeting changed: Optical impairments draft

We discussed today a difference in the interpretation of the model related to https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/issues/29<https://protect2.fireeye.com/v1/url?k=30d0fc94-6f4bc5d7-30d0bc0f-861fcb972bfc-5d5353826774237a&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fissues%2F29> and its interpretation. The different interpretation lead to different computation requirements.

The design team appreciates feedback which of both Cases is preferred and what consideration led to the preference

Case A:


https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/issues#issuecomment-677768528<https://protect2.fireeye.com/v1/url?k=af80a102-f01b9841-af80e199-861fcb972bfc-509d8509e721d1cb&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fissues%23issuecomment-677768528> from August 20 documented the structure of the model:

+--ro supported-modes* [mode-id]

  +--ro mode-id

  +--ro (mode)?

  |  +--:(G.698.2)

  |  |  +--ro standard-mode?                standard-mode

  |  +--:(organizational-mode)

  |  |  +--ro operational-mode?             operational-mode

  |  |  +--ro organization-identifier?      vendor-identifier

➢ |  +--:(explicit-mode)

➢ |    +--ro supported-standard-mode*? leafref

  |     +--ro modulation-types                   identityref

  |     ......

This snippet would allow to use standard mode or operational mode as a simple matching criteria for compatibility i.e. two transponders with Operational-mode=X and suitable line parameters would interoperate.

In case of “explicit” mode, the set of parameters provided, can support one or more application code/organizational mode, that means this set of parameters if applied to configure the transceiver, is aligned with one or more application code/organizational mode.

(Note that there are no additional parameters associated to the organizational mode in this case)

Case B:


https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-optical-impairment-topology-yang/blob/master/ietf-optical-impairment-topology.tree<https://protect2.fireeye.com/v1/url?k=dacf1e95-855427d6-dacf5e0e-861fcb972bfc-13a67d7f67153c4a&q=1&e=acdb6ee5-5f1d-46c2-9f88-0c7a7270ed91&u=https%3A%2F%2Fgithub.com%2Fietf-ccamp-wg%2Fdraft-ietf-ccamp-optical-impairment-topology-yang%2Fblob%2Fmaster%2Fietf-optical-impairment-topology.tree> contains additional parameters to the Operational Mode that are also replicated in the explicit mode. Those parameters are related to the min/max ranges which are allowed.

  +--ro supported-mode* [mode-id]

                +--ro mode-id                      string

                +--ro (mode)


                   |  +--ro standard-mode?         standard-mode


                   |  +--ro organizational-mode

                   |     +--ro operational-mode?

                   |     |       operational-mode

                   |     +--ro organization-identifier?

                   |     |       organization-identifier

                   |     +--ro min-central-frequency?

                   |     |       frequency-thz

                   |     +--ro max-central-frequency?

                   |     |       frequency-thz

                   |     +--ro minimum-channel-spacing?

                   |     |       frequency-ghz

                   |     +--ro tx-channel-power-min?      dbm-t

                   |     +--ro tx-channel-power-max?      dbm-t

                   |     +--ro rx-channel-power-min?      dbm-t

                   |     +--ro rx-channel-power-max?      dbm-t

                   |     +--ro rx-total-power-max?        dbm-t


                      +--ro explicit-mode

                         +--ro supported-modes

                         |  +--ro supported-application-codes*

                         |  |       -> ../../mode-id

                         |  +--ro supported-organizational-modes*

                         |          -> ../../mode-id

                         +--ro line-coding-bitrate?

                         |       identityref


                         +--ro min-central-frequency?

                         |       frequency-thz

                         +--ro max-central-frequency?

                         |       frequency-thz

                         +--ro minimum-channel-spacing?

                         |       frequency-ghz

                         +--ro tx-channel-power-min?

                         |       dbm-t

                         +--ro tx-channel-power-max?

                         |       dbm-t

                         +--ro rx-channel-power-min?

                         |       dbm-t

                         +--ro rx-channel-power-max?

                         |       dbm-t

                         +--ro rx-total-power-max?


In this case, interoperability can only be provided if the organizational code on each side match AND all other explicit parameters defined in the organizational-code have a matching range.

(Note that min/max parameters are associated to the organizational mode AND to the explicit mode in this case)

Juniper Business Use Only


CCAMP mailing list



Juniper Business Use Only