[Starlingx-discuss] Minutes: StarlingX Networking Meeting - 07/25

[Khalil, Ghada]

Meeting minutes/agenda are captured at:
https://etherpad.openstack.org/p/stx-networking

Bugs - stx.2.0 gating 
High: 6
- https://bugs.launchpad.net/starlingx/+bug/1817936  - Austin, code review in progress, but there are concerns with the proposed fix    >> Matt recommends we change this to Medium given the VIM recovers and there is no notable system impact. Frank put this as high due to concerns about stability. Leave for now and evaluate again a little later.
- https://bugs.launchpad.net/starlingx/+bug/1835807  - Joseph, reproduced.  Still investigating the cause.
- https://bugs.launchpad.net/starlingx/+bug/1835965  - YaoLe, Confirmed by Paulina, this bug is an issue with steps. So this bug  marked as invalid.   >> Recommend to update the wiki with the steps
- https://bugs.launchpad.net/starlingx/+bug/1836252  - Joseph, unsure if each occurrence is the same issue.  Need a complete set of logs to investigate each occurence.  >> Logs requested from the reporter
- https://bugs.launchpad.net/starlingx/+bug/1836682  - Chenjie, steps are wrong and pci alias is not configured. Need submit a patch to override the pci alias in nova.conf.  >> There is code already that configures the pci aliases. Matt to addd reference in LP and Chenjie to investigate further why this code is not working.
- https://bugs.launchpad.net/starlingx/+bug/1836969  - Unassigned, Matt working on proposal, then need to assign to a developer  >> On the fence whether this is high or medium. AIO duplex-direct is not functional w/ IPv6.

Medium: 9
- https://bugs.launchpad.net/starlingx/+bug/1817593  - Teresa, not being worked. Can potentially defer to stx.3.0 given this is for a specific config; can be worked around by explicitly defining the cluster network
- https://bugs.launchpad.net/starlingx/+bug/1818118  - Joseph, not reproduced yet  >> There are other reports that live migration is not working at all
- https://bugs.launchpad.net/starlingx/+bug/1821026  - fpxie  >> looks to be actively being worked by the assignee.   Seems consistent with config-drive. May need to consider changing this to high.
- https://bugs.launchpad.net/starlingx/+bug/1822396  - Joseph, reproduced and understood.  stx-openstack application not reapplied (and new overrides not generated) after unlock if still in applying state from previous unlock. >> This should be re-assigned to the containers team. There is currently work in progress to optimize the application apply. Ghada to re-assign
- https://bugs.launchpad.net/starlingx/+bug/1832047  - Cheng, no one could reproduce the bug. The reporter had also only seen it happened once.  >> New logs added. Cheng will review the logs. Matt to check if there was feedback from EricM on pmon behavior
- https://bugs.launchpad.net/starlingx/+bug/1832892  - Steve, on vacation, will investigate in August
- https://bugs.launchpad.net/starlingx/+bug/1834234  - Unassigned, Need to assign. Can potentially defer to stx.3.0 and note this config as a limitation for IPv6. Workaround is to use a vlan for the cluster network. >> Agreed to move to stx.3.0
- https://bugs.launchpad.net/starlingx/+bug/1834556  - Marvin, reproduced. Still investigating the cause.   >> Is this a duplicate of https://bugs.launchpad.net/starlingx/+bug/1832697  which was recently fixed?
- https://bugs.launchpad.net/starlingx/+bug/1836972  - Steve, on vacation, will investigate in August

Low: 1
- https://bugs.launchpad.net/starlingx/+bug/1830082  - Teresa, lower priority, not being worked. Can potentially defer to stx.3.0 given a configuration step was missed in this scenario.

Undecided / New: 3
- https://bugs.launchpad.net/starlingx/+bug/1830286  - Elio  >> Still believe this is a configuration issue, so expected to close as Invalid.  Elio will send a further update this week.
- https://bugs.launchpad.net/starlingx/+bug/1835575  - New >> Assigned; will start investigation/reproduction
- https://bugs.launchpad.net/starlingx/+bug/1837759  - New >> Re-assigned to the distro.openstack team

stx.2.0 Networking Test Status
- Networking Regression
       - Tracker: https://docs.google.com/spreadsheets/d/1MoBroFimeQjsJvCC_N_DSLYMQsNZVMzTwAVc5tPy_t0/edit#gid=838066175 
       - Still working on TCs related to sriov & pci-pt
       - Not running NUMA mesh TCs that nobody knows what the steps are and what they are intended for
       - Not able to run any QAT device passthru TCs due to lack of hardware. Ghada believes Ada ordered some and they were used in testing the QAT device upversion. Elio to follow-up

stx.3.0
- OVS-DPDK Containerization - Prime: Cheng
       - Spec is under review - comments received from Matt
- TSN - Prime: Huifeng
       - Spec: https://review.opendev.org/#/c/666768/  merged on July 22
       - Not expecting code changes; likely integration/validation effort only.  Matt to send questions related to proposed test head/application to be used for the validation