01:01:00 <zhiyuan> #startmeeting tricircle
01:01:01 <openstack> Meeting started Wed May  9 01:01:00 2018 UTC and is due to finish in 60 minutes.  The chair is zhiyuan. Information about MeetBot at http://wiki.debian.org/MeetBot.
01:01:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
01:01:05 <openstack> The meeting name has been set to 'tricircle'
01:01:43 <zhiyuan> now we are blocked by the ci issue
01:02:04 <song> yes I have resend the email
01:02:15 <song> but do not have anyone to reply now.
01:02:25 <zhiyuan> to the same topic?
01:02:53 <song> add ci and intra
01:04:26 <clarkb> this the no compute hosts issue? has anyone looked at the compute host logs?
01:05:07 <zhiyuan> let me check it again
01:05:54 <clarkb> also looks like both nodes run a full control plane? that may be confusing things if not intended
01:06:12 <zhiyuan> WARNING nova.compute.manager [None req-1dd21811-39c2-4a63-ad44-2c1f6d114667 None None] No compute node record found for host ubuntu-xenial-inap-mtl01-0003885152. If this is the first time this service is starting on this host, then you can ignore this warning.: ComputeHostNotFound_Remote: Compute host ubuntu-xenial-inap-mtl01-0003885152 could not be found
01:06:22 <zhiyuan> see warning
01:06:34 <zhiyuan> and this
01:06:37 <zhiyuan> WARNING nova.compute.resource_tracker [None req-1dd21811-39c2-4a63-ad44-2c1f6d114667 None None] No compute node record for ubuntu-xenial-inap-mtl01-0003885152:ubuntu-xenial-inap-mtl01-0003885152: ComputeHostNotFound_Remote: Compute host ubuntu-xenial-inap-mtl01-0003885152 could not be found.
01:07:07 <clarkb> that should be fine since it is the first time that cloud was ever spun up
01:09:47 <zhiyuan> then no error log found in the nova-compute log
01:10:38 <clarkb> I would also double check if the cells compute host registration command is run successfully. I forget the detaisl on that but know you ahve to enroll the compute hosts
01:11:26 <zhiyuan> is this command run in the devstack script?
01:11:43 <clarkb> it is part of devstack-gate because it has to run after all the devstacking is done
01:12:19 <clarkb> https://git.openstack.org/cgit/openstack-infra/devstack-gate/tree/devstack-vm-gate.sh#n248
01:14:51 <zhiyuan> ok, we will check it, thanks for your advise
01:15:17 <song> clarkb thanks for your help.
01:20:09 <zhiyuan> song, what about we do a fresh install locally, to see if we will meet the same problem
01:20:42 <zhiyuan> I am wondering if we need to do same change to our smoke test setup
01:20:50 <zhiyuan> s/same/some
01:21:01 <song> that is good idea I will do it today.
01:21:50 <zhiyuan> here is the log for discover-host
01:21:55 <zhiyuan> Found 2 cell mappings.
01:21:55 <zhiyuan> Skipping cell0 since it does not contain hosts.
01:21:55 <zhiyuan> Getting computes from cell 'cell1': 6e118c71-9008-4694-8aee-faa607944c5f
01:21:58 <zhiyuan> Found 0 unmapped computes in cell: 6e118c71-9008-4694-8aee-faa607944c5f/usr/local/lib/python2.7/dist-packages/psycopg2/__init__.py:144: UserWarning: The psycopg2 wheel package will be renamed from release 2.8; in order to keep installing from binary please use "pip install psycopg2-binary" instead. For details see: <http://initd.org/psycopg/docs/install.html#binary-install-from-pypi>.
01:22:04 <zhiyuan> """)
01:22:30 <zhiyuan> not sure if this is a expected result, need to check the discover-host script
01:22:38 <zhiyuan> s/a/an
01:24:10 <song> ok, I will reinstall a fresh locally to check that problem.
01:24:18 <zhangchi> sounds like need run "nova-manage cell_v2 discover_hosts"
01:25:19 <song> this may relate to the smoke fail?
01:25:33 <zhiyuan> yes, the discover-host script run nova-manage
01:25:38 <zhiyuan> https://github.com/openstack-dev/devstack/blob/master/tools/discover_hosts.sh
01:26:02 <song> that is good.
01:26:17 <zhiyuan> possibly, because nova-schedule complain that no host is found
01:26:59 <zhiyuan> "Found 0 unmapped computes in cell", seem like there is error when discovering hosts
01:27:39 <zhiyuan> here is the whole log: http://logs.openstack.org/59/566459/1/check/legacy-tricircle-dsvm-multiregion/f2e2510/logs/devstack-gate-discover-hosts.txt.gz
01:28:42 <Yipei> is it caused by the way we integrate cell v2 with tricircle?
01:29:21 <song> but in before we do not meet this ever.
01:30:08 <zhiyuan> Yipei, I don't think so, integration is disabled by default
01:31:29 <zhiyuan> if a fresh install locally has the same issue, we need to change our configuration and script
01:32:09 <Yipei> oh, yes, i forgot it
01:32:15 <song> yes, that maybe our test problem.
01:32:20 <zhiyuan> maybe we miss to adapt some change in devstack or nova
01:35:01 <zhiyuan> Yipei, any update for the new l3 networking model?
01:36:43 <Yipei> yes, zhuangzhuang has already joined in the feature, he continues with the first part and i will help him to finish it ASAP. i will go on with the second part about the routed network
01:39:53 <zhiyuan> first part means the creation of the gateway port without ip?
01:41:22 <Yipei> first part is creation of the external network with a local router, the second part is creation of the gateway port without ip for routed network
01:42:21 <zhiyuan> will this change affact our current process? or you use a configuration option to enable it?
01:43:28 <Yipei> i think it is better to use a configuration option to enable the model
01:43:43 <zhiyuan> I think so
01:44:03 <zhiyuan> by the way, could you update the spec to discuss about the detail?
01:44:32 <Yipei> ok, i will update the spec about the detail
01:44:46 <zhiyuan> thanks
01:44:53 <Yipei> you are welcome
01:45:09 <zhiyuan> other topics?
01:45:45 <song> no from me.
01:46:01 <xuzhuang> no
01:46:06 <Yipei> no
01:46:47 <zhiyuan> ok, thanks for join the meeting
01:46:52 <zhiyuan> #endmeeting