*** wanghao has quit IRC | 00:22 | |
*** wanghao has joined #openstack-mogan | 00:22 | |
zhenguo | morning mogan! | 01:13 |
---|---|---|
liusheng | morning | 01:15 |
zhenguo | liusheng: should we also use flavor name for server show in our CLI? | 01:15 |
liusheng | zhenguo: seems many patches merged yestoday. lol | 01:15 |
zhenguo | liusheng: hah | 01:15 |
zhenguo | liusheng: I proposed a new one for partitions | 01:15 |
liusheng | zhenguo: you are capable! | 01:16 |
zhenguo | liusheng: hah | 01:16 |
zhenguo | liusheng: seems I got the reason whey we only support whole disk image | 01:16 |
liusheng | zhenguo: why ? | 01:16 |
zhenguo | liusheng: as no partitions passed to the node | 01:16 |
liusheng | zhenguo: so maybe that will be solved after your proposal landed ? | 01:17 |
zhenguo | liusheng: need more test | 01:18 |
zhenguo | liusheng: https://github.com/openstack/mogan/blob/master/mogan/baremetal/ironic/driver.py#L147-L149 | 01:18 |
zhenguo | liusheng: currently we fetch the root_gb information form node local_gb property | 01:18 |
zhenguo | liusheng: but irnonic may not always provide that | 01:18 |
liusheng | zhenguo: so may need to ensure that property be set before provision ? | 01:20 |
zhenguo | liusheng: but not flexible | 01:20 |
zhenguo | liusheng: which means we also use the whole disk | 01:20 |
liusheng | zhenguo: how if we support partition | 01:22 |
zhenguo | liusheng: you can specify the root, swap, ephemeral partions size | 01:23 |
liusheng | zhenguo: does those info will be in flavor ? | 01:24 |
zhenguo | liusheng: no | 01:24 |
zhenguo | liusheng: you can specify any size | 01:24 |
zhenguo | liusheng: as the whole disk is for you, no like vm | 01:24 |
liusheng | zhenguo: oh, I found you patch, got it | 01:24 |
zhenguo | liusheng: hah | 01:25 |
liusheng | zhenguo: yes | 01:25 |
liusheng | zhenguo: just confirmed, OSC server show command show the flavor name and uuid, similar with image | 01:25 |
liusheng | zhenguo: may we can follow that | 01:25 |
zhenguo | liusheng: yes, better to follow that | 01:26 |
liusheng | zhenguo: ok, submit a patch | 01:26 |
zhenguo | liusheng: server list doesn't show flavor? | 01:26 |
liusheng | zhenguo: will submit | 01:26 |
zhenguo | liusheng: thanks | 01:26 |
liusheng | zhenguo: yes, list output only show name, but with --long parameter | 01:27 |
zhenguo | liusheng: got it | 01:27 |
zhenguo | liusheng: what image we used for server creation on our temepst? | 01:27 |
liusheng | zhenguo: seems the cirros -disk image | 01:29 |
zhenguo | liusheng: ok | 01:29 |
liusheng | zhenguo: in the [baremetal] in tempest.conf it already configure the partition image ,but seems we don't use that | 01:30 |
zhenguo | liusheng: let me check | 01:31 |
zhenguo | liusheng: nearly forget we will have a meeting today :D | 01:40 |
liusheng | zhenguo: hah | 01:40 |
openstackgerrit | Zhong Luyao proposed openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 01:47 |
Xinran | zhenguo, hello, so how can I sync db when I add a new column to server | 01:53 |
Xinran | zhenguo, seems need to restack? | 01:54 |
zhenguo | Xinran: if you just modify the initi db migration file, you can't sync db | 01:54 |
zhenguo | Xinran: yes | 01:54 |
zhenguo | Xinran: we will add new version of db changes from next cycle, sorry for the inconvenient | 01:54 |
Xinran | zhenguo, hah, np, I will try to restack | 01:55 |
zhenguo | Xinran: hah | 01:55 |
shaohe_feng | zhenguo: morning | 02:00 |
zhenguo | shaohe_feng: morning | 02:00 |
zhenguo | hi all, the weekly meeting will happen soon, please move to #openstack-meeting | 02:00 |
*** litao__ has joined #openstack-mogan | 02:02 | |
openstackgerrit | Zhong Luyao proposed openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 02:03 |
*** wong_ has joined #openstack-mogan | 02:52 | |
wong_ | hi | 02:52 |
liusheng | wong_: hi | 02:53 |
wong_ | I am working in ironic,my colleague tell me the mogan | 02:55 |
wong_ | what's the different between them? | 02:56 |
wong_ | I am chinese, if english is hard to type :) | 02:58 |
zhenguo | wong_: sorry we were in the weekly meeting :D | 03:00 |
zhenguo | wong_: ironic is quite like libvirt, but mogan is like nova | 03:00 |
wong_ | aha,my honor | 03:01 |
wong_ | your word is clear | 03:02 |
zhenguo | lol | 03:02 |
zhenguo | for more details, you can refer to our wiki | 03:03 |
* zhenguo brb | 03:03 | |
wong_ | I am looking it | 03:03 |
zhenguo | liusheng: seems we use partitions image for server creations in our temepst, the gate is red if I don't pass partitions info | 03:21 |
liusheng | zhenguo: :(, may need to use the partition info | 03:22 |
zhenguo | liusheng: yes, I will keep the current implementation to keep using local_gb as default root partions if there's no partions with partitoned images | 03:22 |
liusheng | zhenguo: so do you need to create another server for testing ? | 03:23 |
liusheng | zhenguo: you know for saving time, we only use 1 server in tempest | 03:23 |
zhenguo | liusheng: hah, so only keep the 1 server is enough | 03:24 |
zhenguo | liusheng: which uses the defult partition, or do you like ot send partitions there? | 03:24 |
liusheng | zhenguo: it is not a good way, but Mogan server creating is time consuming | 03:24 |
zhenguo | liusheng: yes | 03:24 |
zhenguo | liusheng: maybe add another job with runs in a separated vm | 03:25 |
liusheng | zhenguo: or maybe change to use partition, I think is partitoned server is advanced than general server, it can cover most cases. hah | 03:25 |
zhenguo | liusheng: hah ok | 03:26 |
zhenguo | liusheng: just found, seems we can also pass partition format to ironic, like ext4 | 03:26 |
liusheng | zhenguo: sounds greate, that is common use | 03:27 |
zhenguo | liusheng: hah | 03:27 |
zhenguo | liusheng: please have a look at this https://review.openstack.org/#/c/466168/ when you got time | 03:29 |
liusheng | zhenguo: ok, sure, will test | 03:30 |
zhenguo | liusheng: thanks | 03:30 |
*** wong_ has quit IRC | 05:39 | |
shaohe_feng | liusheng: my account is shaohef@126.com | 05:51 |
liusheng | shaohe_feng: seems need your username, not the email address | 06:06 |
openstackgerrit | Zhenguo Niu proposed openstack/mogan master: Add root disk partitions support https://review.openstack.org/499039 | 06:45 |
openstackgerrit | wanghao proposed openstack/python-moganclient master: Support to getting managebale servers https://review.openstack.org/499474 | 06:47 |
* zhenguo brb | 06:57 | |
openstackgerrit | liusheng proposed openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 07:04 |
openstackgerrit | Zhong Luyao proposed openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 07:11 |
openstackgerrit | Xinran WANG proposed openstack/mogan master: Return key name with server API https://review.openstack.org/499481 | 07:12 |
liusheng | luyao: oops, I thought you may busy in other things, so I uploaded a patchset of your patch | 07:14 |
luyao | liusheng: I didn't notice you update the patch. :D | 07:14 |
liusheng | luyao: hah | 07:14 |
luyao | liusheng: I saw it just now, thank you! | 07:15 |
liusheng | luyao: besides the place I commented, you'd better to add a exclusive group for the --net-id and --port-id paramters | 07:15 |
liusheng | luyao: because the command did check if both port-id and net-id specified, and it will fail | 07:16 |
liusheng | luyao: or specify both | 07:16 |
luyao | ok, got it | 07:17 |
openstackgerrit | wanghao proposed openstack/python-moganclient master: Support to getting managebale servers https://review.openstack.org/499474 | 07:21 |
openstackgerrit | Zhong Luyao proposed openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 07:25 |
zhenguo | liusheng: the parittions path passed the gate :D | 07:40 |
liusheng | zhenguo: hah, cool + awesome | 07:40 |
zhenguo | liusheng: lol | 07:40 |
zhenguo | liusheng: I will test with our bare metal env | 07:40 |
liusheng | zhenguo: ok | 07:40 |
zhenguo | liusheng: do you have an ready env to test https://review.openstack.org/#/c/498775/ | 07:41 |
liusheng | zhenguo: I can only test the client functionality | 07:43 |
liusheng | zhenguo: no conson support | 07:43 |
zhenguo | liusheng: for shellinabox is enough | 07:43 |
zhenguo | liusheng: no ned to access the console, just check if it's consistent with the API response | 07:44 |
liusheng | zhenguo: seems my current env also cannot support | 07:44 |
liusheng | zhenguo: ok, will check it | 07:44 |
zhenguo | liusheng: thanks, and please help to check the format with nova's :D | 07:44 |
liusheng | zhenguo: ok | 07:44 |
zhenguo | liusheng: list servers only return id, links, name https://developer.openstack.org/api-ref/compute/#list-servers | 08:12 |
openstackgerrit | Merged openstack/python-moganclient master: Add interface attach/detach support https://review.openstack.org/466168 | 08:15 |
*** openstackgerrit has quit IRC | 08:17 | |
* zhenguo brb | 08:24 | |
*** liusheng has quit IRC | 08:42 | |
*** openstackgerrit has joined #openstack-mogan | 08:45 | |
openstackgerrit | Tao Li proposed openstack/mogan master: Manage existing BMs: Part-2 https://review.openstack.org/481544 | 08:45 |
openstackgerrit | Tao Li proposed openstack/mogan master: Manage existing BMs: Part-2 https://review.openstack.org/481544 | 08:50 |
litao__ | zhenguo, liusheng,shaohe_feng : pls review my patch https://review.openstack.org/481544 | 08:52 |
zhenguo | litao__: ok, thanks for the update | 08:52 |
zhenguo | litao__: I found you import engine_api in engine_manager for listing az, as we can't support az for now, how about just get rid of az stuff? | 08:56 |
zhenguo | litao__: you can just leave az to None | 08:57 |
litao__ | zhenguo: we can't support it next release? | 08:58 |
litao__ | zhenguo: If then, I will drop it | 08:58 |
zhenguo | litao__: not sure, hah | 08:58 |
zhenguo | litao__: we should avoid import engine/api in engine/manager | 08:59 |
litao__ | zhenguo: OK, I will drop it currently | 08:59 |
zhenguo | litao__: ok thanks | 08:59 |
zhenguo | litao__: as pike will be a separated branch, we don't need to keep unused things | 09:00 |
litao__ | zhenguo: OK | 09:00 |
*** openstackgerrit has quit IRC | 09:03 | |
*** liusheng has joined #openstack-mogan | 09:07 | |
*** wanghao has quit IRC | 09:10 | |
zhenguo | liusheng: list servers only return id, links, name https://developer.openstack.org/api-ref/compute/#list-servers | 09:16 |
liusheng | zhenguo: we also have status and power_state | 09:18 |
zhenguo | liusheng: not sure why nova just return that | 09:18 |
*** openstackgerrit has joined #openstack-mogan | 09:19 | |
openstackgerrit | Zhenguo Niu proposed openstack/python-moganclient master: Change to use new console URI https://review.openstack.org/498775 | 09:19 |
zhenguo | liusheng: do we need to follow it? | 09:19 |
liusheng | zhenguo: not sure, seems the status is also the key info | 09:19 |
zhenguo | liusheng: if so, let's just keep it | 09:20 |
liusheng | zhenguo: ok, maybe in future we can change that according to real requirements | 09:21 |
zhenguo | liusheng: ok | 09:21 |
zhenguo | liusheng: please help to erase the disks of 10.81.8.4-8 when you got time, thanks! | 09:21 |
liusheng | zhenguo: I only use "openstack baremetal node delete $node" to delete.. | 09:23 |
liusheng | zhenguo: does that work ? | 09:24 |
zhenguo | liusheng: no | 09:24 |
liusheng | zhenguo: :( | 09:25 |
zhenguo | liusheng: do you remember we need to connect the BMC via our web to erase the disks? | 09:25 |
liusheng | zhenguo: we tried to use BMC to get the mac address | 09:26 |
zhenguo | liusheng: we also erase the original disk layout to make it work | 09:27 |
openstackgerrit | Tao Li proposed openstack/mogan master: Manage existing BMs: Part-2 https://review.openstack.org/481544 | 09:27 |
liusheng | zhenguo: seems we only use a image to install os | 09:28 |
zhenguo | liusheng: yes, you can try to install os again, but after erase the orginal disk layout you can quit, then we can use it | 09:28 |
liusheng | zhenguo: :( maybe we can try | 09:29 |
zhenguo | liusheng: hah, maybe you have forgotten, you just did that for node2 and node3 to make them work :D | 09:30 |
liusheng | zhenguo: not quite sure.. | 09:33 |
zhenguo | liusheng: I'm sure with that, lol | 09:33 |
liusheng | zhenguo: the node-2 and node3 is ok ? | 09:33 |
liusheng | zhenguo: hah | 09:33 |
zhenguo | liusheng: yes | 09:34 |
liusheng | zhenguo: if so, we can login to BMC and try to install OS | 09:34 |
zhenguo | liusheng: yes, | 09:34 |
zhenguo | liusheng: you can try one to see | 09:34 |
liusheng | zhenguo: please try node4 | 09:54 |
zhenguo | liusheng: I just restack the env, seems there are some problems, will try later | 09:54 |
openstackgerrit | Xinran WANG proposed openstack/mogan master: Return key name with server API https://review.openstack.org/499481 | 09:55 |
*** liusheng has quit IRC | 09:59 | |
*** liusheng has joined #openstack-mogan | 10:09 | |
*** litao__ has quit IRC | 11:55 | |
zhenguo | liusheng: just found when deleting a deployment from a node failed it will automatically move to Maintenance state, and not go to placement | 11:56 |
liusheng | zhenguo: it won't be tracked by placement ? | 11:58 |
zhenguo | liusheng: Maintenance node will not be tracked | 11:58 |
liusheng | zhenguo: oh, yes | 11:58 |
zhenguo | liusheng: so, maybe it's ok for us to follow nova? | 11:59 |
liusheng | zhenguo: Ironic will automatically turn into maintance state | 11:59 |
zhenguo | liusheng: ye | 11:59 |
liusheng | zhenguo: does this effect ? | 12:01 |
zhenguo | liusheng: we don't need to consider failed state anymore as it will move to maintenance | 12:02 |
liusheng | zhenguo: how if the node was manually recovered | 12:03 |
zhenguo | liusheng: not sure if we should consider that | 12:04 |
zhenguo | liusheng: you have already wiped the partitons on node4? | 12:09 |
liusheng | zhenguo: yes, node4 and node5 | 12:09 |
zhenguo | liusheng: ok, thanks | 12:10 |
liusheng | zhenguo: np | 12:10 |
liusheng | zhenguo: our env is ok now ? | 12:10 |
zhenguo | liusheng: yes | 12:11 |
liusheng | zhenguo: did you update the repos before run stack.sh ? | 12:11 |
zhenguo | liusheng: just updated mogan :D | 12:12 |
liusheng | zhenguo: :(, I want to know if you have met my problem, hah | 12:12 |
zhenguo | liusheng: hah | 12:13 |
zhenguo | liusheng: seems node4 still not work | 12:13 |
zhenguo | liusheng: but node2 and node3 do | 12:14 |
liusheng | zhenguo: ... node4 I am sure has been erased | 12:14 |
zhenguo | liusheng: let me try node5 | 12:15 |
liusheng | zhenguo: ok | 12:16 |
zhenguo | liusheng: I'm going to call it a day and head home, bye! | 12:28 |
liusheng | zhenguo: formating the disk with "use whole disk with LVM" right ? | 12:28 |
liusheng | zhenguo: hah | 12:28 |
zhenguo | liusheng: no | 12:28 |
zhenguo | liusheng: I just want to wipe the LVM :D | 12:28 |
liusheng | zhenguo: oh no.. that is my fault | 12:28 |
liusheng | zhenguo: will do again tomorrow | 12:29 |
zhenguo | liusheng: hah | 12:29 |
* zhenguo away | 12:30 | |
*** XueFengLiu has quit IRC | 18:12 | |
*** XueFengLiu has joined #openstack-mogan | 18:12 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!