=================== #fdio-vpp: fdio-vpp =================== Meeting started by DaveBarach at 15:01:52 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/fdio-vpp/2019/fdio-vpp.2019-06-25-15.01.log.html . Meeting summary --------------- * CSIT (Maciek Reporting) (DaveBarach, 15:02:37) * CSIT-1904.1 maintenance testing done, report published (mackonstan, 15:02:46) * LINK: https://docs.fd.io/csit/rls1904_1/report/introduction/introduction.html (mackonstan, 15:02:55) * Continuing work on VAT to PAPI migration, going okish, some "hiccups" with VPP API changes, but "mouth covering" with CSIT PAPI fixes :) (mackonstan, 15:03:21) * Really really need closer VPP-CSIT CI alignment, current proposed solution is "VPP API flag day" for CSIT, but it's taking long time... (mackonstan, 15:03:51) * Already missing VAT abstraction, just kidding ;) (mackonstan, 15:04:03) * Re vpp-csit-verify-device-master jobs voting status: currently blocked by broken CSIT-VPP PAPI, all IP tests failing (mackonstan, 15:04:14) * Trending tests partially broken, all IP tests failing (mackonstan, 15:04:35) * All will be back to normal once CSIT PAPI "mouth covering" exercise complete, ETA today/tomorrow CET (mackonstan, 15:04:42) * LINK: https://gerrit.fd.io/r/c/20278/ (mackonstan, 15:04:49) * Adding more performance tests: VXLAN, IPsec scale up to 60k tunnels (mackonstan, 15:04:57) * On a path/mission to disable VIRL: perf, dev/func tests environment alignment, will enable running perf test code as functional dev tests (mackonstan, 15:05:11) * CSIT-1908 plan wiki updated (mackonstan, 15:05:47) * LINK: https://wiki.fd.io/view/CSIT/csit1908_plan (mackonstan, 15:05:56) * Sweetcomb (DaveBarach, 15:06:29) * leverage VOM library for state reconciliation in Sweetcomb (DaveBarach, 15:06:52) * Host Stack (Florin) (DaveBarach, 15:07:44) * TCP now supports tcp-output -> arbitrary next node (DaveBarach, 15:08:24) * Adding tcp delivery rate estimation (DaveBarach, 15:08:48) * needed for sophisticated congestion control algorithms (DaveBarach, 15:09:14) * used for BBR, other proprietary congestion control algos (DaveBarach, 15:09:38) * host stack: aggregate all applications into one place (DaveBarach, 15:10:07) * echo-server/client built into vpp (DaveBarach, 15:10:23) * move to "hoststack-plugins" (DaveBarach, 15:10:40) * Documentation (John D reporting) (DaveBarach, 15:15:09) * main page making progress, waiting for feedback (DaveBarach, 15:15:18) * waiting for LF to give us the DNS entries (DaveBarach, 15:15:34) * aarch64 (Honnappa reporting) (DaveBarach, 15:16:07) * reached out to Damjan for specific patch concerns (DaveBarach, 15:16:45) * working on c11 atomic based locks (DaveBarach, 15:17:09) * LF lab: please use Thunderx-1'S not macciato-bin (DaveBarach, 15:18:03) * 19.04.1 (Dave Wallace reporting) (DaveBarach, 15:20:58) * patches queued up, not planning to do another dot release until someone asks (DaveBarach, 15:21:12) * 19.08 starts consuming csit resources the first week of August, (DaveBarach, 15:21:39) * should decide in mid-July if we want to do another 19.04.X release (DaveBarach, 15:21:56) * 19.08 (Andrew Y reporting) (DaveBarach, 15:22:24) * .py packaging improvements, shout out to Aloys (DaveBarach, 15:22:51) * API changes (Dave Wallace reporting) (DaveBarach, 15:23:39) * CSIT will start checking for API changes (DaveBarach, 15:23:56) * need to be able to take specific tests offline (DaveBarach, 15:24:33) * will verify any api changes - if used messages actually change, (DaveBarach, 15:25:42) * patches will get -1 (DaveBarach, 15:25:51) * Table creation / feature enablement (Dave B) (DaveBarach, 15:32:12) * small devices don't need huge virtual space consumers (DaveBarach, 15:35:09) * LINK: https://gerrit.fd.io/r/20117 API Change patch to be merged soon (dwallacelf, 15:35:41) * PaulV to push API Change patch which moves default values from PAPI Executor to API files. (dwallacelf, 15:36:18) * fd.io community events (TSC request for info) (DaveBarach, 15:37:03) * previous events have been fd.io people talking to each other (DaveBarach, 15:38:09) * not so helpful in terms of growing the community (DaveBarach, 15:38:18) * one event per year community meeting might make sense (DaveBarach, 15:38:58) * [Jerome T comments above] (DaveBarach, 15:39:11) * Infra (Ed Kern reporting) (DaveBarach, 15:40:32) * a couple of jenkins issues last week, asked for root-cause (DaveBarach, 15:40:54) * checkstyle commit message checker minor issue (DaveBarach, 15:41:31) * Welcome swati to the infra team (DaveBarach, 15:41:56) * test-debug failures (Paul V) (DaveBarach, 15:44:34) * packet generator causing "random" failures (DaveBarach, 15:47:23) * dpdk drivers vs native driver interface creation vs. probing (DaveBarach, 15:49:31) Meeting ended at 15:56:01 UTC. People present (lines said) --------------------------- * DaveBarach (45) * mackonstan (13) * collabot` (3) * dwallacelf (2) Generated by `MeetBot`_ 0.1.4