SAWG MTG Minutes 2011-09-16
From NGDC Wiki
- Dan Kowal
- Anna Milan
- Phil Jones
- Heather Anderson (soon to be Brown)
- Deirdre Byrne
- Scott McCormick
- Jeremy Throwe
- Muhammad Rabi
SAWG Reflection: What purpose has this group served, should it continue?
- Provided input into sub. agreement formulation and use.
- Has been a mechanism to insure all 3 DCs and CLASS are having conversations. Provided a venue of communication about what’s going on at our level.
- Allows us to keep the broader picture in mind.
- Reports to the COWG as needed.
- Provides a venue to talk about process, how the DCs work with CLASS. Some issues needing attention:
- Process from "Request to Approval," clarity is needed.
- CLASS guidance on when ROM estimates ar done vs. formal ones?
- Upper management not interested in the details, but should be involved upfront, identifying resources, etc….
- Sharing perspectives from each DC has been beneficial.
- Unanimous agreement that the group should still continue.
- Keep to monthly meetings at established meeting time, 2nd Fri/mo. at 11:30 ET.
Data Center Discussion
- NODC replacement for Tess will be Gregg Foti, comes from the JPL PODAAC.
- Interested in working with more data streams than the Aquarius mission previously worked on; has background in writing user handbooks.
- Reorganization underway at NODC which may preclude further participation from Don Collins. Deirdre said that others including herself will participate depending upon the topic of discussion.
- NDE Ocean Color example: Deirdre may have Johnathan Blythe participate regarding Ocean Color products as they need to be updated. Existing SAs may be insufficient. New ones may be needed.
- Tom Schott believes that any legacy algorithms will only need an update to the SA that will reflect data volume increases.
- Deirdre inquired about who the science lead is? Phil is the technical lead.
- Phil suggested that this situation may require mare than a new SA, but an appraisal as well. He said Ocean Color (OC) is similar to the MIRS case in terms of how it fits in a current data set family in CLASS. If the updated OC for NDE requires a new family, there's a cost to this.
- Deirdre said that this situation requires some leadership from the DCs to sort out the funding issues with legacy products vs. their new replacements and decide what gets support. Someone needs to summarize a plan; she mentioned that this could be worked through Johnathan Blythe.
- Should the SAWG name be changed given the coverage of topics beyond the SA? We'll address this at the next meeting.
- Testing mtg. on Oct. 6.
- Archive/access PDR – Oct 5th. (Modified approach taken.)
- Anna wanted to know who the contacts are for CLASS collection-level metadata at each DC.
- NCDC: Phil; Drew Saunders (knowledgeable about NPP).
- NODC: John Relph (technical issues, wafs...); regular contact is still under discussion at NODC; will know in the next week or two.
- Collection Level Metadata (CLM) Management Issue
- Jeremy described how there's a potential issue for how the proposed GOES-R collection-level metadata creation process conflicts with the current process for how this metadata is currently managed by the DCs whereby they updated records by populated targeted WAFs for harvest by CLASS. The proposed model has GS generating it and delivering it to CLASS. Phil confirmed that the DEWG has discussed that the GS wouldn't send the metadata directly to the Data Centers. We'd have to go through GS.
- Anna suggested if the DCs pick up the "default" metadata record after it's been delivered by GS to CLASS would work.
- Phil: the DCs could set up a subscription and manage it.
- Jeremy: Would CLASS need to preserve all versions and make them available or only act as a "pass through" for the DCs to gain access to them. He said there is no set of requirements that cover this new way of managing these records.
- Scott: There are two issues on the floor:
- Is there changes to the interface such that CLM come through the CLASS Receipt Node?
- Do we get CLM from GS or through existing interface(s) - WAFS and keep multiple versions?
- Jeremy: There is no interface for getting the CLM directly nor for an arrangement to deliver the "default" version to the DC and then back to CLASS.
- Muhammad: There's been a tiger team for metadata.
- Phil: The tiger team has been tasked to focus solely on metadata development within the netCDF, and send quasi-static metadata in ISO.
- Jeremy: CLASS is concerned about this. There is no definition for how to subscribe to the Products (Production?) Distribution Segment.
- Anna: Need to have a conversation with GS or CLASS?
- GS was the answer.
- Jeremy: CLASS acting as a pass-through to the DCs is a new requirement. This scenario has been socialized through the tiger team, the DEWG. There's been discussion about CLASS "cobbling" together the CLM and merge with the granules to create an ISO-compliant metadata record.
- Phil: This idea has recently been presented.
- Jeremy: CLASS prefers getting the CLM through the WAFs.
- Phil: Pass-through w/o archiving? Version of the metadata recorded, but not vetted...
- Jeremy: There's a timing issue associated with this subject. At an upcoming DEWG Mtg, 9/21 (2pm ET), the tiger team will broadly socialize the results of their study. They will focus on examples one of them being the absolute minimum metadata to embed in the files and pushed to CLASS. DC participation is needed before this issue snowballs.
- Anna: The tiger team's focus was on the netCDF and ISO.
- Jeremy: Our focus can be brought up.
- Anna: Is a new tiger team needed to address this?
- Muhammad: Don't know if Harris is going with the embedded ISO metadata.
- Jeremy: Looks like we're heading down this path.
- Friday, October 14, 2011.