Earlier  
Posted Nick Remark
#openstack-cyborg - 2019-09-19
03:42:26 chenke Great jobs thanks ShaoHe.
03:42:40 Sundar Also: "Move it to cyborg/accelerator/common/handler.py". Bind is really an operation on an ExtARQ. It logically belongs with objects/ext_arq.py. If you want to split that into separate source file, that is OK. But it can be a mix-in rather than a separate object/class, IMHO
03:43:37 Yumeng shaohe_feng: great! looking froward to the mock guide
03:44:12 shaohe_feng I have check nova's object code, Then I make this change.
03:44:28 wangzhh shaohe_feng, Thx for your effort.
03:44:38 shaohe_feng Sundar any details for how to split it?
03:45:23 Sundar shaohe_feng: I found this blog useful: http://www.qtrac.eu/pyclassmulti.html
03:46:00 Sundar It considers many ways to split a Python class into different source files, and finally recommends mix-ins
03:48:43 shaohe_feng glance it. seem it is a big change.
03:49:52 Sundar Hmmm... only the last part is the mix-in. That could be a small change. You can move your chosen methods into a separate file, put it in a mix-in, and inherit that mix-in into the ExtARQ object class
03:50:11 Sundar I can help as much as I can.
03:51:09 shaohe_feng good, then I can write a mock evn guide for test.
03:51:35 Sundar In that article, the last section "The Definitive Version?" alone is about mix-ins
03:51:40 Sundar OK, great
03:52:34 Sundar Anything else, Shaohe?
03:52:48 shaohe_feng no, that's all for me.
03:53:02 Sundar Thanks very much, once again.
03:53:06 Sundar #topic AoB
03:53:09 shaohe_feng let move the patch on
03:53:20 Sundar Python IPv6 jobs: https://review.opendev.org/#/c/682517/ Please review
03:53:52 Sundar Many patches hit merge conflict after recent merges
03:54:04 shaohe_feng it does not matter.
03:54:25 shaohe_feng we just improve our git skill
03:54:49 shaohe_feng other active project
03:55:18 Sundar We need one more review for https://review.opendev.org/#/c/680953/ from outside Intel.
03:55:28 shaohe_feng conflict is very common
03:55:51 Sundar Sure
03:56:03 Sundar Train schedule: https://releases.openstack.org/train/schedule.html RC1 candidate is next week!
03:56:18 Sundar Hope to get the critical patches in by that time.
03:56:49 Sundar After that, even bug fixes are not assured
03:57:48 Sundar BTW, Cyborg will get packaged as a RPM as part of OpenStack release: https://opendev.org/openstack/rpm-packaging/src/branch/master/openstack/cyborg
03:58:19 Sundar Anything else, guys?
03:58:28 shaohe_feng no
03:58:32 chenke no
03:58:52 Sundar Have a good day! Bye
03:58:56 Sundar #endmeeting
03:58:58 openstack Meeting ended Thu Sep 19 03:58:56 2019 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)
03:58:58 Yumeng bye
03:58:59 openstack Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-09-19-03.02.html
03:59:00 openstack Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-09-19-03.02.txt
03:59:01 chenke bye all.
03:59:01 shaohe_feng thank you
03:59:02 openstack Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-09-19-03.02.log.html
03:59:03 wangzhh bye
03:59:04 shaohe_feng bye
03:59:05 chenke thanks all.
03:59:21 s_shogo bye
03:59:26 xinranwang bye
04:12:40 openstackgerrit Sundar Nadathur proposed openstack/cyborg master: Fix arq api errors in delete and unbind https://review.opendev.org/682913
04:39:31 openstackgerrit Merged openstack/os-acc master: Removing project os-acc. https://review.opendev.org/682498
06:38:13 openstackgerrit chenchunxiu proposed openstack/cyborg master: Fix arq api errors in delete and unbind https://review.opendev.org/683035
06:40:38 openstackgerrit chenchunxiu proposed openstack/cyborg master: Fix arq api errors in delete and unbind https://review.opendev.org/683035
08:52:50 openstackgerrit chenker proposed openstack/cyborg master: Fix the hardcoding of user role using sdk_adapter approach https://review.opendev.org/682565
11:54:30 openstackgerrit YumengBao proposed openstack/cyborg master: conductor writes device_profile update to db https://review.opendev.org/679406
11:55:42 openstackgerrit YumengBao proposed openstack/cyborg master: conductor writes device_profile update to db https://review.opendev.org/679406
12:08:51 openstackgerrit chenker proposed openstack/cyborg master: Fix the hardcoding of user role using sdk_adapter approach https://review.opendev.org/682565
13:52:08 openstackgerrit Merged openstack/cyborg master: bug fixing: let FPGA driver report correct traits when no SRIOV enabled https://review.opendev.org/680953
#openstack-cyborg - 2019-09-20
03:03:41 openstackgerrit zhurong proposed openstack/cyborg-tempest-plugin master: Make cyborg-tempest job voting https://review.opendev.org/683271
03:35:09 Sundar Hi chenke
03:35:20 chenke Hi Sundar.
03:35:25 chenke How are you.
03:35:36 Sundar I added this section to cyborg.conf
03:35:54 chenke emm. the result raise exception?
03:35:56 Sundar [nova]
03:36:08 Sundar username = nova
03:36:19 chenke [nova]
03:37:05 Sundar I see in Cyborg api logs : Failed to create a working proxy for service compute: No valid endpoint was discoverable.
03:37:45 Sundar Should it be [nova] or [compute], based on some comments?
03:38:03 chenke should be nova.
03:38:31 chenke I had talked with effied yesterday.
03:38:33 Sundar Ok. I have restarted all 3 Cyborg services a few times
03:38:40 chenke Ok.
03:38:42 Sundar What else should I do?
03:39:37 chenke Since novaclient is not convenient for testing,
03:39:53 chenke First you can ensure placement_client is ok.
03:40:38 Sundar It is good. The RPs and traits get created. But I already had a placement section.
03:40:39 chenke Just wait the result now. see if still get an error.
03:41:13 chenke This nova section is indepentent with placement section.
03:41:13 Sundar The nova section in cyborg.conf is copy-paste of [placement], with user name alone changed
03:41:56 Sundar "project_name = service" -- should I change this?
03:42:22 chenke When you install the latest cyborg code after this patch merged. the username in nova section is different with placement.
03:42:48 chenke https://review.opendev.org/#/c/682565/8/devstack/lib/cyborg@a238
03:43:40 Sundar Ah, I should stack again
03:43:42 chenke The project_name is depending on the configuration of the devstack environment.
03:43:47 chenke yep.
03:43:54 Sundar Unstack and then stack -- will try that
03:44:22 Sundar Thanks, chenke :)
03:44:45 chenke Ok. Wait.
03:45:03 chenke Now the placement client is Ok, right?
03:47:25 chenke You can test nova_client placement_client like this:
03:47:45 chenke This is my test.py:
03:48:11 chenke from oslo_config import cfg
03:48:13 chenke nova_client
03:49:55 Sundar Yea, Placement client was ok from the beginning. Once the stacking completes, I'll try that
03:51:11 chenke Ok.
04:23:56 Sundar Hi chenke
04:24:01 Sundar I tried this:
04:24:02 Sundar from oslo_config import cfg
04:24:09 Sundar It crashes with:
04:24:25 Sundar Traceback (most recent call last):

Earlier   Later