#fdio-meeting: FD.io CSIT Project Meeting

Meeting started by mackonstan at 13:58:58 UTC (full logs).

Meeting summary

    1. mackonstan (mackonstan, 13:59:04)
    2. Tibor Frank (tifrank, 14:00:31)

  2. Agenda bashing (mackonstan, 14:03:13)
  3. Inputs from projects - VPP (mackonstan, 14:03:50)
    1. Andrew, one week to rc2 (mackonstan, 14:04:23)
    2. Fan, added a fix to master, cherry picked into stable/2206, Andrew to review/merge (mackonstan, 14:05:08)

  4. Inputs from projects - TSC (mackonstan, 14:05:21)
  5. Inputs from projects - goVPP (mackonstan, 14:05:45)
    1. Maciek on behalf of Dave Wallace, goVPP has now regular project calls, see updated fd.io govpp wiki page for details (mackonstan, 14:06:24)

  6. E810 flow "offload" tests (mackonstan, 14:08:15)
    1. re 36119: Add ipsec policy test suites with flow director or rss accelerated | https://gerrit.fd.io/r/c/csit/+/36119 (mackonstan, 14:09:29)
    2. Fan, tests are failing (mackonstan, 14:09:38)
    3. Fan, need e810 driver and ddp comms package update - Peter and Yulong are working on it (mackonstan, 14:11:11)
    4. Peter, just confirmed that DDP version should be correct (mackonstan, 14:11:25)
    5. Fan, Maciek, for vpp_device tests, patch is in progress, 35743: dpdk: VPP own DPDK flags | https://gerrit.fd.io/r/c/vpp/+/35743 (mackonstan, 14:13:20)
    6. Fan, start email thread to resolve the proposed code and test design issue (mackonstan, 14:16:20)

  7. VPP and CSIT hot items (mackonstan, 14:17:03)
    1. re AMD performance results review, Sivaprasad, AMD team is in the process of setting up internal CSIT lab, will come back with updates (mackonstan, 14:19:28)

  8. Releases - 2206 (mackonstan, 14:19:58)
    1. rc1 testing, Tibor, all testing done, results in draft report in master (mackonstan, 14:20:33)
    2. https://s3-docs.fd.io/csit/master/report/ (mackonstan, 14:21:05)
    3. Tibor, added testbeds to report 3n-alt (mackonstan, 14:21:23)
    4. Tibor, observed a large difference between PDR and NDR (mackonstan, 14:21:39)
    5. Juraj, is following with Arm team on that issue (mackonstan, 14:22:18)
    6. Tibor, x710 10gbe links are down on 2n-clx, Peter is t/shooting (mackonstan, 14:22:41)
    7. Viliam, another issue known from 2202 csit-1803 all imix cbc failing on 3n-icx (mackonstan, 14:25:04)
    8. Viliam, csit-1827 - all avf crypto 3n-icx and 3n-skx failing for 1518b no traffic, imix excessive traffic loss, already caught in trending but was very sporadic (mackonstan, 14:27:29)
    9. Vratko confirmed it is indeed csit-1827 (mackonstan, 14:28:43)
    10. https://jira.fd.io/browse/CSIT-1827 (mackonstan, 14:28:50)
    11. Tibor, once rc2 artifacts are available, will be doing 5 iterative runs (mackonstan, 14:29:32)

  9. Physical Infrastructure (mackonstan, 14:34:14)
    1. Maciek, remaining ICX systems are arriving this week (mackonstan, 14:35:04)
    2. Sivaprasad, re new AMD hw, will ask Keesang to join the call and provide update (mackonstan, 14:35:55)
    3. Peter, re SNR onboarding, no progress, missing basic user manual info from Intel (mackonstan, 14:36:21)

  10. Virtual Infrastructure, CI/CD (mackonstan, 14:36:38)
  11. Environment, Framework (mackonstan, 14:36:58)
  12. Trending (mackonstan, 14:37:13)
    1. Peter, vpp-e810 and trex-cx5 link rings are not working, trex is unable to send traffic (mackonstan, 14:38:54)
    2. Vratko, another issue is tcp throughput tests, ticket csit-1830, showed up after trex upgrade (mackonstan, 14:40:28)
    3. Vratko, clearly a trex issue, struggling to find a workaround for small frame sizes (mackonstan, 14:43:57)


Meeting ended at 14:45:11 UTC (full logs).

Action items

  1. (none)


People present (lines said)

  1. mackonstan (42)
  2. collab-meetbot (8)
  3. tifrank (1)
  4. vrpolak (1)
  5. fan_zhang (0)


Generated by MeetBot 0.1.4.