DPDK patches and discussions
 help / color / mirror / Atom feed
From: Patrick Robb <probb@iol.unh.edu>
To: ci@dpdk.org
Cc: dev <dev@dpdk.org>
Subject: Community CI Meeting Minutes - April 3, 2025
Date: Mon, 7 Apr 2025 11:19:35 -0400	[thread overview]
Message-ID: <CAJvnSUCrpSU0+exv_P6aqPev8cW7mARAdRSOmrLXsX378+K6Ew@mail.gmail.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 5943 bytes --]

#####################################################################
April 3, 2025
Attendees
1. Patrick Robb
2. Paul Szczepanek
3. Luca Vizzarro
4. Aaron Conole
5. Manit Mahajan

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

=====================================================================
General Announcements
* Can we reschedule this meeting to account for daylight saving time?
   * Will move the call 1 hour earlier going forward to account for
daylight saving time. This brings the meeting to 13:00 UTC.

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

---------------------------------------------------------------------
UNH-IOL Community Lab
* New DTS:
   * We are missing some DTS results (for Intel XL710) from this week due
to a dpdk tarball naming conflict on the DUT. This failed “silently” from
our perspective because we were not checking the DTS exit code and flagging
non-zero codes in our CI alert system. We are adding that now and Dean will
put in retests for the XL710 later today.
* Minimum Kernel Version Test:
   * The UNH Lab is running build and unit tests off of a VM which is set
to kernel 4.19 - the minimum tested kernel version according to DPDK docs.
* QAT Crypto Test:
   * Cody at UNH has setup a dpdk-crypto-perf test (from legacy DTS) on the
QAT crypto device which came with one of the Intel Xeon CPUs added to the
lab in 2024.
   * We will share the performance metrics with Bruce Richardson/Intel and
then enable reporting for this test
* Baremetal testing maintenance:
   * We are moving around some NICs to newer systems, updating drivers, and
other normal maintenance that we do on baremetal systems at the start of a
release dev cycle
* We are now reporting the dts-check-format label to patchwork

---------------------------------------------------------------------
Intel Lab
* None

---------------------------------------------------------------------
Github Actions
* Aaron has a pre-req patch ready to go for the recheck count. This is for
the PW-CI project.

---------------------------------------------------------------------
AWS
* Patrick will reach out to Shai for an update

---------------------------------------------------------------------
Loongarch Lab
* None

=====================================================================
DTS Improvements & Test Development
* Smoke Tests Testsuite bugfixes:
   * There are two issues, one pertaining to a keyerror on the tests_config
dictionary and the other pertaining to the regex which is used to capture
the driver NICs are bound to from dpdk-devbind.py -s
* DTS Packet Sniffer update:
   * No updates, but this is reviewed from Dean and Patrick
* Performance TG update:
   * DTS Configuration and TestSuite API notes:
      * 1. Test_run.yaml:
         * app location
         * trex configuration
      * TestSuite Class facing functions:
         * define a packet X with scapy
         * call TestSuite.assess_performance_by_packet(packet X, duration Y)
         * This returns an instance of a new performance stats dataclass
   * TG internals notes:
      * Imports a series of TREX stateless API components
      * A set of one STLCLient, Stream and PacketBuilder are created.
      * Once created, the STLStream is started using the given duration.
When completed, stats are gathered via stdout and put into the
PerformanceStats dataclass.
* DTS Roadmap
   * 1. TG abstraction rework (to facilitate performance TGs), testsuite
API additions for perf workloads, and TREX class implementation
      * Luca raised that it may be best to leave the perf TGs as child
classes of the current tg abstract class and simply not implement the
capturing methods
      * Luca also raises that his asyncsniffer patchseries adds some new
methods of collecting packets, such as collecting until a stop condition is
met (like seeing packet X). This may split out packet send and packet
capture from send_packet_and_capture(), making them independent.
   * 2. Write testsuite perf_single_core_fwd
   * 3. Write testsuite rte_flow
   * 4. Port VF support
   * 5. Docs improvements for new users
      * Luca has indicated that Thomas Wilks will contribute to this
   * 6. Rework packet send and receive functions in the Testsuite API
towards supporting IP networking in an L3 environment
   * 7. DPDK Testsuite API: This will provide a stable API which will not
be broken going forward (except at agreed upon intervals)
      * Moving existing testsuite functions to the context hierarchy
   * 8. RSS Support Testsuites
      * Work for this is basically done - reviews needed though:
https://patchwork.dpdk.org/project/dpdk/list/?series=34713
   * 9. Update TestResult module in order to give a more human readable
summary of the testsuites
   * 10. Improved interactive shell data collection and logging.
      * Add a file which exclusively logs the shell interactions
   * 11. Add dedicated file manager to handle passing files and artifacts
between the test engine and nodes in a path agnostic manner
      * This will be a component of the new DPDK DTS API
   * 12. Flexible port binding
      * Give the opportunity to test suites to use all or part of the
supplied port topology with the kernel drivers (instead of DPDK). Could be
achieved with a decorator
      * This will require the Topology class to be aware of what’s truly
available for DPDK and provide only the right ports for it
   * 13. Add  pseudo-shell which handles background services.
      * https://patchwork.dpdk.org/project/dpdk/list/?series=34865
   * 14. L4 port packet matching:
https://patchwork.dpdk.org/project/dpdk/list/?series=34805

=====================================================================
Any other business
* Next Meeting Apr 17, 2025

[-- Attachment #2: Type: text/html, Size: 6479 bytes --]

                 reply	other threads:[~2025-04-07 15:23 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAJvnSUCrpSU0+exv_P6aqPev8cW7mARAdRSOmrLXsX378+K6Ew@mail.gmail.com \
    --to=probb@iol.unh.edu \
    --cc=ci@dpdk.org \
    --cc=dev@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).