14:59:57 <jmorgan1> #startmeeting Infra WG Weekly Meeting
14:59:57 <collabot> Meeting started Mon Oct 30 14:59:57 2017 UTC.  The chair is jmorgan1. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:59:57 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:59:57 <collabot> The meeting name has been set to 'infra_wg_weekly_meeting'
15:00:04 <jmorgan1> #topic Roll call
15:00:08 <jmorgan1> #info Jack Morgan
15:02:02 <hw_wutianwei> #info tianwei Wu
15:02:22 <aricg> #info Aric Gardner
15:03:43 <jmorgan1> #topic Status Update
15:05:11 <jmorgan1> #info vPOD has been provided container4nfv from Huawei lab
15:05:13 <hw_wutianwei> #link https://gerrit.opnfv.org/gerrit/#/c/46283/
15:05:22 <hw_wutianwei> #info this is the patch
15:05:35 <jmorgan1> #info INFRA-185 can be closed
15:06:02 <jmorgan1> #info Nokia lab is providing hardware resources for Long duration stress testing (INFRA-154)
15:06:13 <CristinaPauna> #info Cristina Pauna (Enea)
15:06:26 <hw_wutianwei> #info about the vPOD for container4nfv - INFRA-185
15:07:53 <jmorgan1> #topic general purpose docker build capability
15:09:40 <jmorgan1> #info docker builds have been part of releases since Brahmaputra
15:12:21 <jmorgan1> #info docker builds have a script to build them - the desire is to have this script be owned by releng
15:12:24 <jose_lausuch> https://gerrit.opnfv.org/gerrit/#/c/46111/
15:12:45 <jmorgan1> #link https://gerrit.opnfv.org/gerrit/#/c/46111/
15:13:51 <jmorgan1> #info docker script will create a new doicker build per patch in gerrit
15:16:34 <Julien-zte> #info Julien
15:20:28 <jmorgan1> #info need to think about including docker builds with releases especially as they grow over time
15:21:11 <jmorgan1> #info the proposal is releng project own the script with test-wg helping support it
15:22:59 <jmorgan1> #info CristinaPauna mentions that there are those in the community who don't know about how to use this script
15:24:02 <jmorgan1> #info this means there someone from releng will need to help support community by documentation and user help
15:24:47 <jmorgan1> #info plus a process to how we can get changes made to the script
15:24:47 <Julien-zte> the script in releng is used to build and push image to the dockerhub
15:25:22 <Julien-zte> I anythink anyone can update and submit for review
15:25:40 <Julien-zte> sorry, the voice of mine is bad, nearly can not hear
15:35:15 <jmorgan1> #action dmcbride to start email thread on this topic to get releng project on board
15:35:30 <jmorgan1> #topic logging facilitites
15:38:22 <jmorgan1> #link https://github.com/opnfv/releng/blob/master/modules/opnfv/utils/ovs_logger.py
15:38:41 <jmorgan1> #info there is a python script to collect logs for ovs
15:43:57 <jmorgan1> #info there was discussion on how to solve this logging problem
15:44:38 <jmorgan1> #info like having the jump server collect the logs or have a script to collects logs when there is a failure
15:46:29 <jose_lausuch> https://git.opnfv.org/yardstick/tree/etc/yardstick/nodes/zte_pod1.yaml
15:46:48 <jose_lausuch> example of node file, with this information we can access any node with ssh/scp
15:46:54 <jmorgan1> #link https://git.opnfv.org/yardstick/tree/etc/yardstick/nodes/zte_pod1.yaml
15:47:08 <jmorgan1> #info example of node file, with this information we can access any node with ssh/scp
15:47:44 <CristinaPauna> #info https://gerrit.opnfv.org/gerrit/#/c/44133
15:48:16 <CristinaPauna> #info how logs are collected for armband
15:49:14 <jmorgan1> #action dmcbride to schedule a Plugfest discussion
15:51:32 <chigang_> CristinaPauna: Is it the log of Puppet manifest? or the log generated by OpenStack components such as nova , neutron?
15:51:53 <jmorgan1> #info Julien-zte proposes to use what Openstack does - use a stash log server (a central log server)
15:52:25 <jmorgan1> #topic POD allocation within OPNFV communtity
15:54:17 <CristinaPauna> chigang_: we push /var/log completly
15:54:38 <CristinaPauna> chigang_: it contains both installing logs and openstack logs
15:55:38 <jmorgan1> #info Julien-zte mentions that until we have dynamic CI still need two POD per installer
15:56:54 <jmorgan1> #info radez mentions that they still need both POD - daily jobs are running 24hrs a day for example for test frameworks
15:57:44 <chigang_> CristinaPauna: fine. but some installer put openstack components into different containers, so it is a little hard to collect the logs.
15:58:50 <jose_lausuch> chigang_: so it should be installer specific?
15:59:19 <Julien-zte> chigang_, yes, agree
15:59:22 <chigang_> jose_lausuch yes, I suggest installer collects its log it needs and put it in jumphost.
15:59:28 <Julien-zte> but the installer can deal with it, I think
15:59:41 <jmorgan1> #info timezone change for next meeting - see calendar invite
15:59:51 <Julien-zte> I just think only the installer guys are familiar with logs
15:59:52 <jose_lausuch> fuel does it already
16:00:54 <Julien-zte> hi jose_lausuch, do you mean Fuel or MCP
16:01:06 <Julien-zte> in the earlier Fuel, I think it was done.
16:01:13 <Julien-zte> For MCP, I don't know for now
16:01:30 <jmorgan1> #endmeeting