| *** openstackstatus has quit IRC | 00:13 | |
| *** openstackstatus has joined #openstack-mogan | 00:15 | |
| *** ChanServ sets mode: +v openstackstatus | 00:15 | |
| *** wanghao has joined #openstack-mogan | 00:36 | |
| *** wanghao_ has joined #openstack-mogan | 00:40 | |
| *** wanghao has quit IRC | 00:41 | |
| *** liujiong has joined #openstack-mogan | 00:59 | |
| zhenguo | morning mogan! | 01:07 |
|---|---|---|
| zhenguo | liusheng: seems ironic gate is fixed | 01:31 |
| liusheng | zhenguo: let's recheck! | 01:33 |
| zhenguo | liusheng: hah | 01:33 |
| zhenguo | liusheng: https://review.openstack.org/#/c/476498/ | 01:34 |
| zhenguo | liusheng: it's not landed yet, not sure how they fixed it | 01:34 |
| liusheng | zhenguo: oh, seems, we also need to fix like this | 01:35 |
| zhenguo | liusheng: yes | 01:41 |
| zhenguo | liusheng: shaohe_feng told me that Xinran is also interested in Placement | 01:50 |
| liusheng | zhenguo: hah, welcome to work together | 01:52 |
| zhenguo | liusheng: hah | 01:52 |
| liusheng | zhenguo: do we enable nova in our gate ? | 01:53 |
| zhenguo | liusheng: yes, I remember you said we must enable nova | 01:53 |
| liusheng | zhenguo: hah | 01:53 |
| zhenguo | liusheng: not sure if it includes the placement service | 01:53 |
| liusheng | zhenguo: it should be, Nova need also need that | 01:54 |
| zhenguo | liusheng: hah, recheck works | 01:54 |
| liusheng | cool | 01:54 |
| openstackgerrit | liusheng proposed openstack/mogan master: WIP: import placement service https://review.openstack.org/476325 | 02:00 |
| openstackgerrit | Zhenguo Niu proposed openstack/mogan master: Adds `resources` field for flavors. https://review.openstack.org/475626 | 02:07 |
| zhenguo | liusheng: for ironic node resource class, do we need to force them use a pattern like CUSTOM_XXX | 02:09 |
| liusheng | zhenguo: 'CUSTOM_' + node['resource_class'].upper() | 02:10 |
| liusheng | zhenguo: seems Nova also does this | 02:10 |
| zhenguo | liusheng: for ironic node? | 02:10 |
| liusheng | zhenguo: I am not sure, but before I make mogan + placement work, I found there is a CUSTOM_SMALL resource class in nova's db | 02:11 |
| zhenguo | liusheng: hah | 02:11 |
| liusheng | zhenguo: all the 3 ironic nodes are "small" | 02:11 |
| zhenguo | liusheng: I remember nova don't create resouce_class for ironic yet, only use the standard resources class | 02:12 |
| liusheng | zhenguo: the current PoC can report nodes to placement | 02:12 |
| zhenguo | liusheng: cool | 02:12 |
| liusheng | zhenguo: not very sure | 02:12 |
| zhenguo | liusheng: hah, so we can set resource class, rp, inventories correctly now? | 02:13 |
| liusheng | zhenguo: yes | 02:13 |
| zhenguo | liusheng: cool | 02:13 |
| liusheng | zhenguo: but I didn't implement the queryclient now | 02:14 |
| zhenguo | liusheng: seems that part is easy, hah | 02:14 |
| zhenguo | liusheng: only a select destinations? | 02:14 |
| liusheng | zhenguo: do you think we need to refactor our scheduler, to user like: schedulerclient.select_destinations ? | 02:15 |
| zhenguo | liusheng: yes | 02:15 |
| liusheng | zhenguo: ok | 02:15 |
| zhenguo | liusheng: with placement we can have a simple scheduler | 02:15 |
| liusheng | zhenguo: hah, yes | 02:15 |
| * zhenguo brb | 02:16 | |
| zhenguo | liusheng: do we need to revert the node list API? lol | 02:31 |
| liusheng | zhenguo: maybe, for now, I didn't remove the compute_nodes table and also allow node info passed to db | 02:32 |
| liusheng | zhenguo: will clean that | 02:32 |
| zhenguo | liusheng: hah | 02:32 |
| zhenguo | liusheng: I think when you test for query placement, we need the new flavor landed to construct the resources/resources traits filter | 02:33 |
| liusheng | zhenguo: oh, yes | 02:34 |
| zhenguo | liusheng: it's not a review day, but I will try go get more people review the flavor part | 02:34 |
| liusheng | zhenguo: seems, gate still broken ? | 02:34 |
| zhenguo | liusheng: ? | 02:35 |
| liusheng | zhenguo: hah | 02:35 |
| liusheng | zhenguo: oh, it is not | 02:35 |
| zhenguo | liusheng: I find placement patch pass the gate | 02:35 |
| zhenguo | liusheng: really cool | 02:35 |
| liusheng | zhenguo: hah, yes | 02:35 |
| liusheng | zhenguo: just because it don't effect other things | 02:36 |
| zhenguo | liusheng: hah, but it reports resources to placement | 02:36 |
| liusheng | zhenguo: yes | 02:37 |
| zhenguo | liusheng: should resources be a required filed for flavors? | 02:43 |
| openstackgerrit | Zhenguo Niu proposed openstack/mogan master: Adds `resources` field for flavors. https://review.openstack.org/475626 | 02:46 |
| liusheng | zhenguo: yes | 03:00 |
| zhenguo | liusheng: ok, but that need to update the CLI first, otherwise the devstack will fail | 03:01 |
| zhenguo | liusheng: will do that later | 03:01 |
| liusheng | zhenguo: hah | 03:01 |
| zhenguo | liusheng: http://lists.openstack.org/pipermail/openstack-dev/2017-May/117705.html | 03:02 |
| zhenguo | liusheng: jay said we can leverage placement aggregates to get around nova host aggregate | 03:03 |
| zhenguo | liusheng: just like what we plan to do | 03:03 |
| liusheng | zhenguo: yes, we planed to do that | 03:05 |
| zhenguo | liusheng: maybe we can ask them to support set traits to aggregate | 03:08 |
| zhenguo | liusheng: if so, it really like the metadata to host aggregate | 03:08 |
| liusheng | zhenguo: yes, we can try' | 03:09 |
| zhenguo | liusheng: hah, things seems clear now | 03:09 |
| liusheng | zhenguo: not sure if there is already a plan of them | 03:09 |
| zhenguo | liusheng: seems they don't have such requirements | 03:10 |
| zhenguo | liusheng: for az do we need to set it for every rp, or wait for aggregate traits | 03:10 |
| liusheng | zhenguo: hah | 03:10 |
| liusheng | zhenguo: may we can wait, we also dont have a strong requirement, right ? | 03:11 |
| zhenguo | liusheng: so revert the az related stuff in mogan? | 03:11 |
| zhenguo | liusheng: now we have a default az, maybe it's ok to keep it but just remove things related to compute_node.az | 03:12 |
| liusheng | zhenguo: maybe just wait untill we landing placement ? | 03:12 |
| zhenguo | liusheng: maybe aggregate traits is just wrapping tow placement api | 03:13 |
| zhenguo | liusheng: list all rp in the aggregate and set traits to every rp | 03:13 |
| zhenguo | liusheng: at list it can work, nova's task is very slow | 03:13 |
| liusheng | zhenguo: yes, that is similar we use current interface directly | 03:22 |
| zhenguo | liusheng: yes, after finishing the work of migrrating to placement, we can discuss more about this | 03:23 |
| liusheng | zhenguo ok | 03:26 |
| openstackgerrit | Zhenguo Niu proposed openstack/mogan master: Adds `resource_traits` field for flavors https://review.openstack.org/476768 | 03:53 |
| openstackgerrit | Zhenguo Niu proposed openstack/mogan master: Adds `resource_traits` field for flavors https://review.openstack.org/476768 | 04:50 |
| *** liusheng has quit IRC | 05:17 | |
| *** liusheng has joined #openstack-mogan | 05:20 | |
| -openstackstatus- NOTICE: unknown issue with the git farm, everything broken - we're investigating | 06:03 | |
| *** ChanServ changes topic to "unknown issue with the git farm, everything broken - we're investigating" | 06:03 | |
| zhenguo | everything broken ... | 06:05 |
| wxy | what is git farm | 06:10 |
| zhenguo | not sure, hah | 06:19 |
| *** wanghao_ has quit IRC | 06:37 | |
| *** wanghao has joined #openstack-mogan | 06:38 | |
| *** wanghao_ has joined #openstack-mogan | 06:42 | |
| *** wanghao has quit IRC | 06:43 | |
| openstackgerrit | Zhenguo Niu proposed openstack/python-moganclient master: Add support for new flavor properties https://review.openstack.org/476804 | 07:00 |
| zhenguo | shaohe_feng, wanghao_: please help to look at the new flavor specs https://review.openstack.org/#/c/454113/ when you got time | 07:14 |
| wanghao_ | zhenguo: sure, it's always on my list | 07:15 |
| zhenguo | liusheng: hah | 07:15 |
| zhenguo | wanghao_: thanks | 07:15 |
| zhenguo | liusheng: can we split the placement into tow patches | 07:15 |
| zhenguo | liusheng: one is report and the other is query? | 07:15 |
| liusheng | zhenguo: sure, small patches are easy to review | 07:16 |
| zhenguo | liusheng: yes | 07:16 |
| -openstackstatus- NOTICE: ok git.openstack.org is working again, you can recheck failed jobs | 08:02 | |
| *** ChanServ changes topic to "ok git.openstack.org is working again, you can recheck failed jobs" | 08:02 | |
| zhenguo | hi all, place recheck your failed jobs, it's working again. | 08:11 |
| *** ChanServ changes topic to "Bugs: bugs.launchpad.net/mogan | Review: https://review.openstack.org/#/q/project:openstack/mogan,n,z" | 09:10 | |
| openstackgerrit | wanghao proposed openstack/mogan master: Refactor exception raise by using mogan exception https://review.openstack.org/475971 | 09:17 |
| openstackgerrit | Zhenguo Niu proposed openstack/python-moganclient master: Add support for new flavor properties https://review.openstack.org/476804 | 09:29 |
| *** wanghao_ has quit IRC | 09:31 | |
| zhenguo | liusheng: hey, are you working on splitting placement patches? | 09:37 |
| openstackgerrit | Tao Li proposed openstack/mogan master: Fix the stunk in *ing in powering servers. https://review.openstack.org/476862 | 09:43 |
| openstackgerrit | Merged openstack/mogan-specs master: New flavor for baremetal servers https://review.openstack.org/454113 | 09:52 |
| *** liujiong has quit IRC | 10:04 | |
| * zhenguo away | 10:05 | |
| *** openstackgerrit has quit IRC | 11:33 | |
| *** litao__ has joined #openstack-mogan | 14:03 | |
| *** liusheng has quit IRC | 15:01 | |
| *** liusheng has joined #openstack-mogan | 15:02 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!