Community CI Meeting Minutes - December 8th, 2022

[Patrick Robb <probb@iol.unh.edu>]

[-- Attachment #1: Type: text/plain, Size: 5123 bytes --]

December 8, 2022

#####################################################################
Attendees

1. Patrick Robb
2. Min Zhao
3. Lincoln Lavoie
4. Juraj Linkeš
5. Ali Alnubani
6. Aaron Conole
7. Jeremy Spewock
8. Lijuan Tu

#####################################################################
Agenda

1. General Announcements
2. CI Status
3. DTS Improvements & Test Development
4. Any other business

#####################################################################
Minutes

=====================================================================
General Announcements

* 2022 Retrospective:
https://docs.google.com/document/d/1DMf-Lkl6QIJbe7wtX1aJz6mdHiITZjuUbufyCZBg-WY/edit
* 2023 Lab priorities:
https://docs.google.com/spreadsheets/d/1GW0hTbEXjA0ZmDLrQRhXwPFu8X-o8qH1sMBTdg2bDyE/edit#gid=0
* LTS 22.11.1 released
   * Should the 22.11 tag be updated to point to 22.11.1, right now, it
effectively points to 22.11.0? This would involve changing git history,
which the maintainers never do. Thomas followed up via email stating that
this was off the table.

=====================================================================
CI Status

---------------------------------------------------------------------
UNH-IOL Community Lab

* Periodic testing is now added for 22.11 LTS release, and for the
22.11-staging branch of dpdk-stable.
   * David notified us on Slack about an issue with the incorrect commit
urls being assigned to our tarballs acquired for dpdk-stable testing. This
does not affect testing accuracy, but it is annoying, and I will apply the
fix for that later today.
* Environment created for ZUC & Snow3G PMD unit tests, which have been
manually run successfully. The team is currently building a Jenkins
pipeline to get this online.
* After touching base with Dell support again about the RAID controller on
the Ampere test bed, they determined a replacement was in order and are
shipping on to the lab.
* Owen has brought Kmod testing against linux-next to production
* ABI testing rollout pending updates to lab code / scripts or resolution
of the conversation about 22.11 tag.

---------------------------------------------------------------------
Intel Lab

* The maintainer for next-broadcom may be inactive - a bugfix applied to
main two months ago was not pulled to next-broadcom, and Lijuan is unable
to get a response from the maintainer on this.
* What does the timeline look like for 19.11 EOL? There will be one more
(final) release next week.
* Intel CI Lab has some ideas for long term support testing - in the old
days only the latest revision of a branch would be tested - Intel is
considering also permanently testing the very first version of a release
(like 22.11.0). Some customers may always stick to using the initial
version of a release.

---------------------------------------------------------------------
Loongarch Lab

* There may be some problems with the test-report mail server. Some emails
from Loongson with test results have been blocked.
   * Ali: “It seems that the emails have been getting rejected by the
mailing list's server since Tuesday because loongson's mail server was
detected in a data set for addresses associated with spam. I just double
checked and it seems the address was removed from the spam list. Can you
please try to resend?”

---------------------------------------------------------------------
Github Actions

* Aaron has merged a change from David which adds some series dependencies
support to GHA-Robot
* Aaron having a meeting with Rashid Khan regarding how the tech board can
get responses back to UNH regarding the SOW

=====================================================================
DTS Improvements & Test Development

* Work continues on the hello world test case
* Further discussion needed regarding how DTS handles failures - currently
whenever a failure occurs anywhere in the process it is caught and
recorded, then DTS moves on. It is a problem when it’s unclear whether a
failure is caused by DTS or DPDK - DTS team wants to make the feedback from
failures more clear, and obviously just make DTS more stable in general.
   * Owen is interested in adding exit code to DTS for additional feedback
on failures. Juraj will follow up with Owen to figure out how exactly this
can be best implemented for CI testing purposes.
      * How can we deal with multiple failures at the same time? Can there
be a hierarchy of exit codes?
   * One idea is to place a higher priority on on failures from DPDK as
opposed to a DTS failure

=====================================================================
Any other business

* Next meetings: December 22, January 5, sticking to the normal schedule.
* December 28th DTS meeting will be canceled as Honappa is on sabbatical
until January 3rd.
* Tech board meeting is on January 14th for UNH feedback

-- 

Patrick Robb

Technical Service Manager

UNH InterOperability Laboratory

21 Madbury Rd, Suite 100, Durham, NH 03824

www.iol.unh.edu

[-- Attachment #2: Type: text/html, Size: 7403 bytes --]