#fdio-meeting: FD.io CSIT Project Meeting

Meeting started by mackonstan at 14:05:29 UTC (full logs).

Meeting summary

    1. Tibor Frank (tifrank, 14:06:02)
    2. Dave Wallace (dwallacelf, 14:06:04)
    3. juraj linkeš (jlinkes, 14:06:14)

  2. Inputs from VPP (mackonstan, 14:06:18)
    1. Andrew - nothing to report (mackonstan, 14:07:04)

  3. Inputs from TSC (mackonstan, 14:07:12)
    1. Dave - nothing new from TSC (mackonstan, 14:07:36)

  4. E810 flow "offload" tests (mackonstan, 14:08:57)
    1. Kai - nothing outstanding AFAICT (mackonstan, 14:09:08)
    2. Maciek - to check status of vpp-device tests for dpdk offload flags. (mackonstan, 14:10:12)

  5. VPP and DPDK version (mackonstan, 14:10:21)
    1. Dave, on the last VPP call asked Jeff Shaw to help with tracking the dpdk bump patch (mackonstan, 14:10:58)
    2. 37840: dpdk: bump to dpdk 22.11 | https://gerrit.fd.io/r/c/vpp/+/37840 (mackonstan, 14:12:08)
    3. https://gerrit.fd.io/r/c/vpp/+/37840 (mackonstan, 14:12:20)
    4. Kai, team hit some issues with this patch (mackonstan, 14:12:33)
    5. Kai, issue is with cryptodev (mackonstan, 14:13:09)
    6. Kai, confirmed that Gabriel is fixing memory leak in cryptodev.c, and once fixed Xinyao will be updating dpdk to 22.11 and/or straight to 23.03 (mackonstan, 14:17:10)

  6. Previous Release (mackonstan, 14:18:14)
  7. Previous Release CSIT-2302 (mackonstan, 14:18:25)
  8. Next Release CSIT-2306 (mackonstan, 14:19:02)
    1. mackonstan, release schedule posted, aligned with VPP (mackonstan, 14:19:19)
    2. https://docs.fd.io/csit/rls2302/report/ (mackonstan, 14:19:49)

  9. Physical Infrastructure (mackonstan, 14:20:37)
    1. Peter, SPR supermicro server RMA - confirmation received that it may take up to 90 days (mackonstan, 14:22:15)
    2. no other issues with physical HW (mackonstan, 14:22:27)
    3. Peter, working on 2n3n testbeds support in CSIT code (2n physical testbed, wired in 3n setup, vpp per numa) (mackonstan, 14:32:58)
    4. Dave, we need SPR machines for device tests, as some VPP patches that require SPR HW accelerators (specifically DSA) are waiting for VPP device test validation before being merged into vpp master (mackonstan, 14:33:54)
    5. Dave, Maciek to discuss options how to address it with available HW, and order more SPR HW soon as per original plan of record for 2023 (mackonstan, 14:34:26)
    6. Maciek, still need to follow up on mlx cx6 NIC allocation to 3n-alt and icx testbeds (mackonstan, 14:35:38)

  10. Trending and Test Failures (mackonstan, 14:36:35)
    1. Peter, recovered ICX testbeds (tests were failing), there is something broken in the SW stack, need to t/shoot and RCA (mackonstan, 14:38:04)
    2. Peter, Maciek, rdma tests failing intermittently, need to t/shoot and RCA (mackonstan, 14:38:58)
    3. https://jira.fd.io/browse/CSIT-1877 (jlinkes, 14:42:28)
    4. https://gerrit.fd.io/r/c/csit/+/38550 (jlinkes, 14:42:32)
    5. Juraj, re slow boot of VMs on Taishan testbeds, some problem with core scheduling while VM booting, current fix is to pin the cpus while VM is booting (mackonstan, 14:42:36)
    6. Juraj, tested the patch on Taishan and 2n-clx testbeds (mackonstan, 14:43:14)


Meeting ended at 14:44:52 UTC (full logs).

Action items

  1. (none)


People present (lines said)

  1. mackonstan (34)
  2. collab-meetbot` (9)
  3. jlinkes (3)
  4. dwallacelf (1)
  5. tifrank (1)
  6. vrpolak (0)


Generated by MeetBot 0.1.4.