Wednesday, 2017-06-21

*** wanghao_ has joined #openstack-mogan00:36
*** wanghao has quit IRC00:39
zhenguomorning mogan!01:00
*** fredli__ has joined #openstack-mogan01:01
zhenguoliusheng: hi, please have a look at this
liushengzhenguo: oh01:10
liushengzhenguo: ok01:10
openstackgerritwanghao proposed openstack/mogan master: Refactor exception raise by using mogan exception
wanghao_zhenguo: ping01:18
*** liujiong has joined #openstack-mogan01:18
wanghao_zhenguo: I found some exception which is not defined in, so I added them in my patch:
zhenguowanghao_: maybe for them we can just raise wsme Client Error, hah01:21
wanghao_zhenguo: I'm not sure that's ok,  you know,  I find that we just catch the exception,  but there is nowhere to raise them,  so I think it's a bug too....01:23
wanghao_zhenguo: even we raise a wsme client error in catch block,  but the code can't got there since no place to raise some exception.01:26
zhenguowanghao_: yeah, in catch block we should not raise client error, but on client side, if there's something wrong like the parameter is not suitable, we can just raise wsme error01:27
wanghao_that could be01:28
wanghao_but I didn't delete those exception,  because I'm not sure if they're useful,  so I just defined them.01:31
wanghao_maybe we check them later.01:31
zhenguowanghao_: ok01:31
zhenguowanghao_: thansk for working on that01:32
wanghao_zhenguo: np,  just find it when I refactor exception raise.01:32
zhenguowanghao_: hah01:32
zhenguoliusheng: seems this should wait for the patch adding floating ip to server nics landed01:37
liushengzhenguo: which patch ?01:38
zhenguoliusheng: this one
zhenguoliusheng: aha, just saw the TODO there01:39
liushengzhenguo: oh, this patch also update the tempest added in the previous patch01:40
liushengzhenguo: yes, the TODO will be done in 2ed patch. lol01:40
zhenguoliusheng: lol. if so, it's okay01:40
*** fredli__ has quit IRC02:02
openstackgerritVu Cong Tuan proposed openstack/mogan master: Stop using deprecated 'message' attribute in Exception
openstackgerritwanghao proposed openstack/mogan-specs master: Manage existing nodes in Mogan
openstackgerritMerged openstack/mogan master: Update api docs to add missing parameters
zhenguoliusheng: how's the placement POC going?02:47
liushengzhenguo: still doing some coding research :(02:48
zhenguoliusheng: hah02:48
liushengzhenguo: do you have any suggestion ?02:49
zhenguoliusheng: at first step, we can just create resource provider for each baremtetal node there02:49
zhenguoliusheng: which will not affect our current scheuling02:49
liushengzhenguo: yes, I am trying to do this02:50
zhenguoliusheng: ok, thanks02:50
zhenguoliusheng: the flavor refactoring work will be done soon, then we can test02:50
liushengzhenguo: np :D, ok02:50
openstackgerritZhenguo Niu proposed openstack/mogan master: Adds `resources` field for flavors.
shaohe_fengzhenguo: hello, I'm coming back05:37
zhenguoshaohe_feng: hah, waiting for your review...06:03
*** wanghao_ has quit IRC06:07
openstackgerritZhenguo Niu proposed openstack/mogan-specs master: New flavor for baremetal servers
zhenguoliusheng: hi, for the flavor resources, I don't plan to change our current scheduler to filter it, but wait for placement ready, then we will sent resources fitler directly to placement06:10
zhenguoliusheng: so, the new flavor task is almost done with the resources patch landed06:10
liushengzhenguo: yes, that make sense06:11
zhenguoliusheng: and after we migrate to placement, seems we lose the ability to filter az06:11
zhenguoliusheng: as we don't have place to save az info in inventory06:11
liushengzhenguo: :( we will skip all the filters of currectly scheduler ?06:12
zhenguoliusheng: but we can set a trait to the resource provider then filter based on that, and list az will list resource providers from placement06:12
zhenguoliusheng: yes06:12
liushengzhenguo: ok, will try06:12
zhenguoliusheng: only sent the fitlers to placement06:12
liushengzhenguo: may the availability zone filter will be added into placement06:13
zhenguoliusheng: no06:13
zhenguoliusheng: we just sent traits info to the placement to let it filter resource providers with that trait for us06:14
zhenguoliusheng: it's more general06:14
zhenguoliusheng: traits can be anyting06:14
liushengzhenguo: yes, I know that maybe a way, I mean may placement guys add some specific filter like az in placement06:15
zhenguoliusheng: there's not a filter notion in placement06:15
zhenguoliusheng: they just make some sql query from db,06:15
zhenguoliusheng: so, it's more quickly06:15
liushengzhenguo: ...yes, I know that. hah06:16
zhenguoliusheng: hah06:16
zhenguoliusheng: and for aggregates, seems there are already some openstack CLI to help manage it06:17
zhenguoliusheng: we may not need to do extra work for that, only some way to pass aggregate info from flavors06:18
liushengzhenguo: osc CLI of placement ?06:18
zhenguoliusheng: yes06:18
liushengzhenguo: that seems is aggregate in Nova, not in placement06:20
zhenguoliusheng: seems yes06:20
zhenguoliusheng: just saw that from
zhenguoliusheng: but I don't find the CLI like openstack resource-provider ...06:20
liushengzhenguo: just have a try by typing "openstack --debug aggregate list"06:21
zhenguoliuisheng: yes, that's a host aggreate06:22
openstackgerritZhenguo Niu proposed openstack/mogan-specs master: New flavor for baremetal servers
*** wanghao has joined #openstack-mogan06:36
*** wanghao has quit IRC07:37
*** wanghao has joined #openstack-mogan07:39
openstackgerritwanghao proposed openstack/mogan master: Refactor exception raise by using mogan exception
*** zhuli has joined #openstack-mogan08:43
zhenguozhuli: hi, welcome here :D08:46
*** wanghao has quit IRC08:53
*** wanghao has joined #openstack-mogan08:54
*** wanghao has quit IRC09:31
* zhenguo away09:58
*** liujiong has quit IRC10:16
*** litao__ has quit IRC12:04
*** litao__ has joined #openstack-mogan14:13
*** harlowja has joined #openstack-mogan17:07
*** openstackgerrit has quit IRC20:03

Generated by 2.15.3 by Marius Gedminas - find it at!