[Ofmfwg] Sunfish meeting notes for March 21, 2025

[Lee, Peter]

  1.
H3 CXL work
     *
Russ has access to the remote environment and is exploring the APIs. There seem to be some missing references. Russ will work on automating the discovery so he can find all the issues. Russ will share this list of issues with Welly.
     *
Next week, will work on getting an agent that can talk to the actual hardware to upload the inventory to Sunfish.
  2.
Documentation
     *
Skip
  3.
Github repository consolidation
     *
Discussion about consolidating Github repositories. A few people prefer fewer repositories for a number of reasons - simplicity, schema drifts and extra work involved to resolve them.
     *
The only repositories we are currently using today are sunfish_docs, sunfish_library_reference, and sunfish_server_reference.
     *
Russ will update the readmes in the reference repositories as not all of the readmes are current.
     *
Russ needs 3-4 weeks to merge in the various changes and get everything caught up. Once this work is done and all the mains are up to date, we can discuss about merging repositories.
  4.
Flux resources from Broker 0 to the compute brokers
     *
Mike is working on the following tasks. He wants to get the basics around Sunfish and Slurm prolog/epilog done next week so he can demo it. He will also put the code on the OFMF server.
        *
Mapping of the closest CDI resources to the computes
        *
Accounting of the compute node CDI resource use for each interactive and batch run
  5.
Code consolidation approach for Sunfish and the Sunfish NVMeoF Agent
     *
This is about the merging of the branches that Russ is working on. Mike will try to help by providing a second set of eyes.
     *
We are currently using the mockups under the OFMF workgroup repos. We need to move the mockups we are using to a directory under the sunfish_server_reference repository. That way, if the server reference is serving mockups, it has the mockups available. These mockups need to be maintain so they conform to any changes in the Redfish schemas.
  6.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/ofmfwg/attachments/20250321/53eaae54/attachment-0001.htm>