Agenda
- Block Island Release is complete. Anything special need to be done for this 2.4.0 release that now has a lighthouse release name to it? Did anyone complete the release notes? We talked about automating release notes anything to report? AI Former user (Deleted) to talk with Tom Capirchio do we need anything special. AI Former user (Deleted) to talk with Thomas Sullivan do we have a automated way to create release notes.
- BIOS GUID was determined as the way to uniquely identify nodes, however slack dialogue indicates this might not be the best option (Switches, EMC nodes have same BIOS GUID etc). Need information from HW to make sure it is a new HW piece, that information could be different between the different device types. Primary key will be mongo UUID, hardware info would be the secondary key. AI Former user (Deleted) to summarize a design .
- Should Ansible Playbooks be supported/delivered as part of RackHD?
- Came up as a solution to retrieve Symphony driver versions. Code was developed a long time ago, not kept up to date. Efforts to bring it up to date and add it to RackHD deployment would be required.
- Including Kranti Uppala and Former user (Deleted) to the discussion
- Update Add an RI for this effort AI Amy Mullins to lead this effort.
- Requirements on resources on RackHD OVA (memory, disk, cpu)
- Again, came up in the context of Symphony. Current OVA is generated with 1 GB RAM and POST test includes only 3 nodes and ~ 12 pollers. Simple Symphony dev/test config has 4-5 nodes (~16-20 pollers) with some switches (not an exact number available, but will increase the poller (and therefore resources required))
- Previously articulated min RAM requirement was 4G .
- Tracked in this issue:
Jira Legacy server JIRA (rackhd.atlassian.net) serverId d7cc09d9-666d-3263-a71c-2a9ec3b8cd13 key RAC-4954 - Update - need to lock down specific requirements for CPU/Memory etc. Do we need to go to 8 GB in near future? Any metrics available to indicate the need? AI Leo Zhang are you aware of any metrics? Current plan is to go to 4 GB as stated in the above mentioned PR.
- Merge Guidelines do we have enough information to start process of other folks being able to merge?
- RackHD Code Review and Merge GuidelinesProcess AI All to review and have updates by next CC meeting. Each team has a CC that is responsible for training and overlooking the process for each team. AI Amy Mullins to provide guidelines from Symphony. Need to communicate the process of how to dub someone ready to be CC.
- Hound guidelines
- Not currently a PR gate, but are there any stylistic rules that should block PRs? Update - AI Former user (Deleted) to make recommendations of which hound checks are to be eliminated/ignored. James Turnquist to take that set of rules to flake8.
- See hound violations but the step "Pass" in Jenkins : https://github.com/RackHD/RackHD/pull/703
- Resource allocation for rackhdci Jenkins
- Andre has brought up a concern regarding the number of VMs now deployed to some of the ESXi hosts for the OVA post tests. Have we done an analysis of the resources required for each of the test and where they will be deployed? AI Former user (Deleted) to take this offline and figure out the analysis
- New SMI repositoriesSMI repositories - Update is to check in SMI as is and not put burden on entire team to review such a large PR for the entire SMI for the initial large PR. Going forward all SMI code will follow the same guidelines for RackHD. AI Michael Hepfer to make sure the permissions match the same as all other RackHD repos. Should be ok as it is inherited from the top RackHD repo.
- What is the merge process going to be for these new repositories?