Core Commiter Weekly Interlock - Feb 6th,2017

Agenda:

Agenda for Core Committer meeting. What do CC members want to see in this meeting? We need to talk release every other CC meeting. The ones in between can talk technical issues however need the items sent to me by Friday EOD or early Monday morning so they can be put on the agenda.
UPDATE Create a confluence page at top of Core Committers meeting page a bulleted running list of agenda items. AI What is difference between this meeting and Arch Forum.
From the HW M&O meeting last week
Unified Data Model (Interim Milestones)
- Use WSMAN for looking at "consistent" data telemetry feeds
- Start looking for unified data structures such as:
    * Firmware Catalogue
    * IPMI Catalogue
    * BIOS Firmware Catalogue
- What would be the base line guidance for the structure?
- Still keep the raw set as an archive
- Take the two catalogues, do an OR operation and the have a consistent output
- Pollers (data grooming) would be a useful add here...
- AI: Start the architecture review on the "ask" for a canonical data set view
- AI: CC-> Do we just implement the "OR" mechanism? UPDATE push to Arch forum
1.1 API Deprecation to the agenda -
- Do we migrate the 1.1 API tests from CIT to FIT (part of phase 2)? Migration has actually been complete. I’d like to consider deprecation from a long term and maintenance perspective. UPDATE there is a story to add a warning to 1.1 API indicating that the API is going away. This will be in the next release 1.2.0. Need to determine actual release it will be completely gone. Proposal to send out notification of deprecation to community, with a goal of deprecating by 1.4.0 (March 10th release) Tom Cap how do we want to communicate (web portal, slack, google groups etc). Do we need to label the March 10th release 2.0.0 (API deprecation enough of a reason to change - what are our semantic versioning rules? Need to update all documentation to remove 1.1 API. (Read the docs etc - who is responsible to clean this up?) Create stories into the unassigned backlog.
- Do we include testing OS installation for both the 2.0 and 1.1 API (increasing test time and/or test resources)? I’d like to consider 2.0 support only. UPDATE agree to not support 1.1 API OS installation testing.
Round Table

AI Tom Capirchio do we want to pre-install SKU packs? See Peter Pan questions in Rac4034 1) in sku pack, will the debian packages being used ? CC need to make the call. if yes. both FIT/Travis/Jenkins need to take action to build them. 3) do we need SKU Pack pre-installed in released ova/vagrant/docker? ( this is another story OSE need to consider) Should we be building Debian packages at all?

UPDATE: Per TomC "we said no to having ski packs built into ova. It will create unnecessary bloat. A consumer would include the necessary skupacks to their environment"

Will update the RAC item as well

AI Former user (Deleted) Thomas Sullivan go through bugs and determine which bugs affect community and to talk about the triage process for open bugs. How do we triage? Who assigns the priority and owner? Do they originally get assigned to the unknown backlog and then get assigned to a specific group backlog? Also how many items are not in a group's backlog (i.e. in the unassigned backlog?)

UPDATE: Tom S sent out list of unassigned bugs The 1st 4 4125 and 4097 have been reported by QRB. Others are older.

RAC-4125 - Getting issue details... STATUS Redfish ESXi OS bootstrap failing with 500 error - in Stryker backlog but still unassigned marked as P2 - Amy Mullins

RAC-4097 - Getting issue details... STATUS get_driveid.js only getting 26 drives (/dev/sd[a-z]) - unassigned, no priority, no group assigned - Leo Zhang

RAC-4136 - Getting issue details... STATUS Post Merge Quality issue in Regression Test : ESXi 5.5 OS Install - unassigned, no group assigned, marked as a P2 - JIRA updated with comments Revert PR 497 and see if problem goes away, if so then need to reach out to Jean Pierre for resolution? Andrew Hou reach out to Rahman Muhammad as they have seen similar issues outside of RackHD

RAC-4137 - Getting issue details... STATUS Post Merge Quality issue in Regression Test : ESXi 6.0 OS Install - unassigned, no group assigned, marked as a P2 - JIRA updated with comments - Revert PR 497 and see if problem goes away, if so then need to reach out to Jean Pierre for resolution? Andrew Hou reach out to Rahman Muhammad as they have seen similar issues outside of RackHD

RAC-4000 - Getting issue details... STATUS Timeout for 50 nodes using node-simulator - unassigned, no priority, no group assigned - Leo Zhang

RAC-745 - Getting issue details... STATUS node tag and label feature - unassigned, no priority, no group assigned - closing not valid

RAC-490 - Getting issue details... STATUS Node will be discovered twice if two NICs are connected to control network - unassigned, no priority, no group assigned Problem was recent as a few months ago. - Leo Zhang.

AI Users guide and Readthedocs are different. Need to look at both as there is conflicting information.

AI Former user (Deleted) to work with Mustang team to get any generic bug fixes for RackHD that affect the mustang collateral - need to open up JIRA ticket and submit PR for the changes that are generic to RackHD.

UPDATE: Going through verification and will send out email with the JIRA tickets.

AI Thomas Sullivan to figure out how we shutdown Github issue creating and force users to use JIRA. Is there away to disable creating of github issue and have a redirect to JIRA?

UPDATE: ??