============================================= #fdio-csit: 31-May FD.io CSIT Project Meeting ============================================= Meeting started by mackonstan1 at 13:58:18 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/fdio-csit/2017/fdio-csit.2017-05-31-13.58.log.html . Meeting summary --------------- * Jan Gelety (jgelety1, 14:00:25) * Agenda bashing (mackonstan1, 14:01:28) * Inputs from LF and FD.io projects (mackonstan1, 14:02:15) * FD.io CSIT operational status (mackonstan1, 14:02:26) * [http://fd.io Helpdesk #41298] Jobs are not triggered from gerrit - jobs working now apparently (mackonstan1, 14:04:08) * vpp-csit-verify-virl- - jobs are continuously failing (mackonstan1, 14:05:30) * Dave Wallace (dwallacelf, 14:05:33) * Peter reporting - issue seems to be related change of dpdk vpp pkg introduced by vpp crypto patch (mackonstan1, 14:07:02) * Peter Mikus (pmikus1, 14:07:24) * LINK: Job triggered with latest dpdk version 17.05-vpp4 https://jenkins.fd.io/job/vpp-csit-verify-virl-master/5684/ (pmikus1, 14:07:52) * Peter - using few patches to verify the fix (mackonstan1, 14:08:03) * Tom Herbert (tfherbert, 14:09:02) * Gerrit patch for updating and operational branch back to oper-170531 https://gerrit.fd.io/r/#/c/6953/2 (pmikus1, 14:09:12) * Gerrit patch for updating dpdk version https://gerrit.fd.io/r/#/c/6952/ (pmikus1, 14:10:05) * Why did this patch verify correctly (https://gerrit.fd.io/r/#/c/5847/) despite using incompatible dpdk pkg version - DPDK_STABLE_VER=17.05-vpp2_amd64 - (https://jenkins.fd.io/view/vpp/job/vpp-csit-verify-virl-master/5650/consoleFull) (mackonstan1, 14:13:45) * Need to revise the way vpp-csit-verify- and csit-vpp-verify jobs pick up dpkd pkg (mackonstan1, 14:15:00) * ACTION: edwarnicke - start a thread to resolve the dpdk pkg ver pegging between csit-dev and vpp-dev (mackonstan1, 14:18:30) * Centos7 VIRL tests - status (mackonstan1, 14:18:52) * Tom updating - focusing on Centos7 make test python library issues (mackonstan1, 14:19:25) * Tom device binding issue on TG still open (mackonstan1, 14:20:00) * perf tb1 status - still healthy ? (mackonstan1, 14:21:26) * Peter - working fine (mackonstan1, 14:21:51) * Test environment updates (mackonstan1, 14:22:17) * adding test equipment: Netgate (mackonstan1, 14:22:41) * Matt Smith reporting - was assisting with installation in LF labs (mackonstan1, 14:25:06) * details of connectivity are unclear (mackonstan1, 14:25:15) * can assist from Netgate side re onboarding Netgate rcc-2758 onto FD.io CSIT (mackonstan1, 14:25:41) * all boxes should be running Centos7.0 (mackonstan1, 14:25:53) * has stock Centos + few packages to support built-in QAT HW (mackonstan1, 14:27:11) * Tom - should be upgraded to Centos 7.3.. that's confirmed working fine on Netgate rcc-2758 (mackonstan1, 14:27:12) * once boxes onboarded management wise and running the right Centos ver, then can proceed with physical setup (2-node topology - 2 of Netgate rcc-2758 connected back-to-back over 2x 10GE) and then start testing … (mackonstan1, 14:30:01) * ACTION: - Maciek, Tom, Matt - drive progress with Netgate rcc devices - onboarding, cabling into 2-node topologies, then go ahead with func tests (mackonstan1, 14:34:08) * VIRL platform (mackonstan1, 14:34:28) * VIRL hosts upgrade - tb4-virl1 upgrade status (mackonstan1, 14:35:29) * issue - spurious NS/ND pkts seen - root caused to running radvd on virl host that was issuing RAs. (mackonstan1, 14:39:27) * 2nd issue - qemu starts failing - independent on load … happens sporadically - so keeping t/shooting. (mackonstan1, 14:40:08) * VIRL hosts upgrade - upgrading remaining hosts: tb4-virl2, tb4-virl3 - BLOCKED (mackonstan1, 14:41:14) * CSIT jobs reorgs (mackonstan1, 14:42:06) * need to update the csit jobs page - bring the most relevant info the top, and review validity of the jobs listings … and update accordingly (mackonstan1, 14:42:55) * ACTION: Tibor, Jan, Maciek to own the csit jobs wiki update (mackonstan1, 14:43:26) * LINK: https://wiki.fd.io/view/CSIT/Jobs (mackonstan1, 14:43:32) * csit1704 report (mackonstan1, 14:44:09) * csit1704 known issues followup (mackonstan1, 14:44:19) * Peter updating - recommendation from dpdk test team to use x16 slot and use only one 40ge port (mackonstan1, 14:45:10) * followup over email on csit-dev to drive closure of known issues with xl710 performance (mackonstan1, 14:45:58) * csit1707 plan (mackonstan1, 14:46:10) * LINK: https://jira.fd.io/secure/RapidBoard.jspa?rapidView=8 (mackonstan1, 14:58:35) Meeting ended at 15:03:29 UTC. People present (lines said) --------------------------- * mackonstan1 (43) * collabot` (4) * pmikus1 (4) * jgelety1 (1) * dwallacelf (1) * tfherbert (1) Generated by `MeetBot`_ 0.1.4