================================== #opnfv-meeting: Weekly TSC meeting ================================== Meeting started by tallgren at 13:01:06 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/opnfv-meeting/2017/opnfv-meeting.2017-07-25-13.01.log.html . Meeting summary --------------- * Uli (uli-k, 13:01:24) * hongbo (hongbo756987458, 13:01:28) * Frank Brockners (frankbrockners, 13:01:33) * Bin Hu (bh526r, 13:01:46) * Luke Hinds (proxy for Fatih) (lhinds, 13:02:02) * Tim Irnich (timirnich, 13:02:18) * Rossella Sblendido (rossella__, 13:02:34) * Bryan Sullivan (bryan_att, 13:02:50) * Xavier Costa (XavierCosta, 13:02:54) * approval of previous meeting minutes (rpaik, 13:03:31) * no feedback on minutes, thus previous minutes are approved (rpaik, 13:03:43) * agenda bashing (rpaik, 13:03:50) * LINK: https://wiki.opnfv.org/display/meetings/TSC#TSC-July25,2017 today's agenda (rpaik, 13:04:05) * Jose Lausuch (Morgan's proxy) (jose_lausuch, 13:04:28) * no other topics suggested (rpaik, 13:05:23) * Request for OPNFV infra resources for project Bamboo PNDA instance (rpaik, 13:05:54) * donaldsh notes that the aim is to deploy an infrastructure for Bamboo PNDA analytics (rpaik, 13:06:44) * fuqiao (fuqiao, 13:07:31) * correlation of data from tests is one of the value that the project is trying to provide (rpaik, 13:07:51) * LINK: https://github.com/pndaproject/pnda-guide/blob/develop/provisioning/aws/PREPARE.md#required-resources pico requirements (rpaik, 13:08:52) * VM deployment will work so BM is not needed (rpaik, 13:09:31) * donaldh notes that the pico spec above is based on AWS (rpaik, 13:11:22) * Julien (Julien-zte, 13:12:08) * running on GCE is also an option (rpaik, 13:13:30) * Is there a link to a clear description of what we will obtain for this expense, e.g. ala what you would have to get by your boss if you wanted to pay for this directly? (bryan_att, 13:14:52) * any request to the BOD needs to have such an explanation and all the TSC should agree that the cost serves a globally-useful goal for OPNFV, i.e. we will get ongoing value as an org (not just a single project) (bryan_att, 13:16:18) * rpaik notes that there's funding in TSC Initiatives bucket to try this out in GCE for a few quarters (rpaik, 13:16:31) * LINK: https://github.com/pndaproject/pnda-guide/blob/develop/provisioning/aws/PREPARE.md#required-resources (donaldh, 13:18:20) * donaldh notes that there community interest in deeper analysis of test results data (rpaik, 13:21:06) * there can be a checkpoint in 5-6 months to ensure that there's value to OPNFV (rpaik, 13:21:54) * I want the TSC to be diligent on understanding and reacting with consensus based upon that understanding. What are the goals of this data mining resource? It has t be more than just standing up another dashboard. There needs to be specific valuable things that we hope to learn from this. Otherwise in the interim it should be stood up on resources provided by the project team. (bryan_att, 13:24:51) * ACTION: donaldh to describe value to be provided via email/wiki and work with aricg to get a GCE quote before a decision at the TSC next week (rpaik, 13:25:09) * Euphrates update (rpaik, 13:25:41) * dmcbride notes that MS5 is due on July 28th where feature projects need to complete scenario integration/feature freeze (rpaik, 13:26:43) * MS6 is due on Aug 11th where test cases/preliminary documentation is due (rpaik, 13:27:41) * Stable branch freeze is open from August 11th to September 1st (tallgren, 13:28:18) * LINK: https://wiki.opnfv.org/display/SWREL/Documentation+Compliance+for+Euphrates (dmcbride, 13:28:38) * there's also a new wiki form where exception requests can be submitted (rpaik, 13:29:32) * LINK: https://wiki.opnfv.org/display/SWREL/Milestone+Exception+Requests+for+Euphrates (dmcbride, 13:29:55) * there's a one week grace period for an exception request, so for people who miss MS5. So exception request for MS5 needs to be in by August 4th (rpaik, 13:31:26) * LF IT/infra update (rpaik, 13:31:51) * Upcoming Java service restarts to address recently announce vulerability. Tenativly scheduled for this Saturday the 29th. Should be less than 30 minutes of downtime. (bramwelt, 13:35:31) * basic is also looking into tech-discuss subscription to the status page and make sure it's still subscribed (rpaik, 13:35:34) * Testing working group proposal on test strategy evolution (rpaik, 13:36:06) * gabriel_yuyang notes that there was a discussion in the Infra WG and a request for a POD for stress testing (rpaik, 13:37:26) * gabriel_yuyang create a jira ticket in Infra WG (gabriel_yuyang, 13:38:14) * jose_lausuch notes that round-robin testing across all 5 installers maybe a challenge (rpaik, 13:38:54) * If we could get 1 POD, then we could start discussion about installers that could be tested (gabriel_yuyang, 13:39:02) * new test cases will be implemented for Euphrates in addition some cases developed for Danube (rpaik, 13:39:57) * is there a link to the scenarios that will be tested using OSA as the installer? We need to be sure this does more than validate what OpenStack as a project should itself be validating (resilience of that generic platform). (bryan_att, 13:40:27) * we need to focus on test cases that stress what is unique to OPFNV - e.g. integration of cloud and SDNCs, detailed networking features e.g. FD.io / DPDK, etc (bryan_att, 13:41:24) * gabriel_yuyang and jose_lausuch note that work can start once a pod can be secured (rpaik, 13:42:05) * jose_lausuch notes that the team will start with simple scenario (e.g. nosdn-nofeature) and will build from there (rpaik, 13:44:24) * bryan_att notes that he desires moving aggressively to do stress testing beyond OpenStack (rpaik, 13:47:05) * I think we need to set broader and higher goals in OPNFV, even to start. Just focusing on OpenStack does not add enough value for OPNFV IMO. OpenStack has *thousands* of devs and they can't do resiliency testing? (bryan_att, 13:48:24) * there's consensus that the ultimate goal is to stress test OPNFV components (rpaik, 13:50:19) * projects health review (rpaik, 13:50:41) * LINK: https://wiki.opnfv.org/download/attachments/2926284/Q2%272017%20OPNFV%20Project%20Health%20Metrics%20v1.xlsx?version=1&modificationDate=1500935803574&api=v2 data for Q2'17 (rpaik, 13:50:57) * Re the last opic: In the first phase the issues in OpenStack that are investigated any deeper than a bug fix request should be those that are truly show-stoppers, i.e. blockers for more complex scenarios. Otherwise we risk getting distracted by issues that take focus off the real goal - resiliency of an integration scenario of components. (bryan_att, 13:52:11) * discussion on tracking upstream work (rpaik, 13:55:41) * ACTION: rpaik to follow-up with frank_brockners on tracking fd.io activities (rpaik, 13:59:29) Meeting ended at 14:02:49 UTC. Action items, by person ----------------------- * donaldh * donaldh to describe value to be provided via email/wiki and work with aricg to get a GCE quote before a decision at the TSC next week * rpaik * rpaik to follow-up with frank_brockners on tracking fd.io activities People present (lines said) --------------------------- * rpaik (37) * bryan_att (10) * tallgren (5) * collabot (4) * gabriel_yuyang (2) * donaldh (2) * dmcbride (2) * hongbo756987458 (1) * fuqiao (1) * bh526r (1) * frankbrockners (1) * lhinds (1) * Julien-zte (1) * bramwelt (1) * timirnich (1) * rossella__ (1) * uli-k (1) * XavierCosta (1) * jose_lausuch (1) Generated by `MeetBot`_ 0.1.4