Update wis2node.adoc focusing on Registration, decomissioning of nodes & datasets, node performance metrics

[klsheets]

For Review: Updates to 2.2.1 Practices and Procedures, 2.2.2 Publishing data, metadata, discovery metadata and notification messages, and 2.2.3 Performance metrics.

All sections likely will need formatting reviews. Some things I noted while working are when do we use WIS vs WIS2 - specifically about nodes, but I think there are few other places where they are used interchangeably.

For performance metrics, the template asked specifically for WIS nodes and not global services. We focused more on the performance metrics for global services, I can add these if desired, but I stuck to nodes and drew more on what happens currently for GISCs and GISC watch for what I wrote.

[6a6d74]

Hi @klsheets ... thanks for committing these changes :)

Comments below:

General: you often refer to "the WMO" (e.g., lines, 9, 12). Better to refer to "WMO Secretariat" I think?

Line 9 (describing centre-id): the first token (ab) is based on an IANA Top Level Domain not ISO3166 2-letter code. Also, the third token (nodename) is optional. For example, a Member may only have a single WIS2 Node - in which case there's nothing to distinguish between at the organisation level.

Line 12 (describing decommissioning): Your proposal works for a NC. But is there an additional step for DCPC in contacting the "sponsor" (i.e., Regional Association or WMO Programme)? To be considered.

In case for both NC and DCPC I think it's worth stating that the operator shall ensure that obligations relating to data sharing within WIS continue to be met after the WIS Node is decommissioned, for example, by migrating these data sharing obligations to another WIS Node.

(in the case of DCPC, the other WIS Node may or may not be operated by the same Member, I guess)

Line 26 (discussing metadata): Data Publishers may work with their affiliated GISC to develop good metadata, but I thought that the GDC verified the metadata? Based on that verification, the GDC will publish a report indicating errors and/or potential improvements (based on discovery metadata KPIs). The GISC should work with the data publisher to remedy issues.

Also, I thought that the topic "recognition" was automated?

1. GDC verifies that the topic is valid.
2. By way of subscribing to notification messages about discovery metadata, the GBs automatically parse the (local) message broker address and topics to which it should subscribe
3. GCs have a blanket subscription to at least two GBs for everything on data/core and metadata/core - so they don't need to know about these updates.

@golfvert, @tomkralidis is this correct? ^^

In addition, the data publisher's affiliated GISC should do some systematic review of what's being published to make sure everything is working OK.

Section "Publication and topic selection": Should add something about using the "experimental" topic for getting started with pre-operational data exchange in test mode ... experimental sits under domain (level 8), e.g. ...weather/experimental. Data publishers can can extend the experimental branch with sub-topics as they deem appropriate. Data consumers must not assume that experimental topics will be durable (i.e., they may change or be removed).

Line 97: spelling ... methode >> method

Line 103: spelling ... q >> a

Line 103: I don't think there are implementations of option 3. Perhaps better to say that option 3 involves bilateral arrangement with another centre to publish metadata on their behalf, such as via their affiliated GISC.
... we could also note that WIS2 Portals may offer metadata publication services in future.

Line 136: spelling ... GISCC >> GISC

[klsheets]

Thank you @6a6d74 for the quick review and guidance,

WMO Secretariat makes sense to me, I was following some of the diagrams when I went with WMO, but in my own self edit I was flipping between WMO and Secretariat and changes all to WMO, so adding Secretariat makes sense to me as well.

I'll make that change and the typo/grammar connections you pointed out.

The things you point out regarding the the GDC and authenticating metadata and the role of GISCs are the items I struggled most to document when drafting these edits. I was largely following the workflows diagrams we reviewed in November, provided by Remy and Annexes to the meeting report. Which brings up and additional question - should those diagrams be linked or included as an Annex in the guide?

[6a6d74]

@klsheets - thanks for updating.

I think we'll need another round of review for how the process will work before we can write this in the documentation. I've raised a new issue (#69) to capture this action.

@tomkralidis - I recommend merging Kari's changes into the main branch as a baseline for further updates in the coming week or so.

[tomkralidis]

Suggest to merge once #67 is approved/merged, if possible, pending some updates from @golfvert

[kurt-hectic]

@klsheets . The broker endpoints are mandatory, I thus suggest the text to read "must provide". You may also consider to be more explicit that broker hostname, port(s) and supported protocols must be supplied to the WMO Secretariat as part of the Registration of a Node (in addition to other elements).
It could also be considered to link to a WMO community webpage which provides details about the registration process.

[klsheets]

@kurt-hectic @golfvert @6a6d74 I thought there was sensitivity to wording such as "must" in the guide. I know during the ET2AT meeting Fred, as a former CBS president, said we could use shall. Happy to word it as strongly as is permissible.

[kurt-hectic]

We must indeed write shall. My point was that this information is mandatory, not optional.

…

________________________________ From: klsheets ***@***.***> Sent: Wednesday, January 10, 2024 2:05:16 PM To: wmo-im/wis2-guide ***@***.***> Cc: Timo Proescholdt ***@***.***>; Mention ***@***.***> Subject: Re: [wmo-im/wis2-guide] Update wis2node.adoc focusing on Registration, decomissioning of nodes & datasets, node performance metrics (PR #62) @kurt-hectic<https://github.com/kurt-hectic> @golfvert<https://github.com/golfvert> @6a6d74<https://github.com/6a6d74> I thought there was sensitivity to wording such as "must" in the guide. I know during the ET2AT meeting Fred, as a former CBS president, said we could use shall. Happy to word it as strongly as is permissible. — Reply to this email directly, view it on GitHub<#62 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AABUXJLCVC3MPMFVCMV4GSLYN2GYZAVCNFSM6AAAAABBJWXXV6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBUHAYTKNBXGE>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[amilan17]

"Where ab is the the IANA Top Level Domain code, domain is the organization's domain name for the main website, and nodename (optional) is descriptive of the purpose of the WIS node (for example a node for a type of data or programme of the data e.g. climate, aviation, etc.) when the organization is hosting multiple WIS nodes."

Information about the centre-id rules will be in an appendix explaining the rules for centre-id and other topics in the Topic Hierarchy in the Manual to WIS2 .... Maybe add a reference to that here instead?

[amilan17]

Is this workflow for each new dataset? I would expect that once a WIS2. node has been approved that this metadata and data workflow would occur without a lot of oversight and regular monitoring should be sufficient.

"An organization ready to publish a new dataset should contact the WMO Secretariat with the approval of their PR (or designate) to The WMO Secretariat. The WMO Secretariat will contact the GISC for the organizaiton with the new (meta)data. The GISC will work with a Global Data Catalogue to verify the metadata."

[amilan17]

experimental topic may also be necessary when the sub-topics have not been approved yet