| russell_h | or I could add a way to passthrough a call to a driver, then and implement the search API-side | 00:00 |
|---|---|---|
| *** dkehn__ is now known as dkehn_ | 00:00 | |
| russell_h | the latter has some complications with how conductors work though. | 00:00 |
| NobodyCam | russell_h: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/pxe.py#L153 | 00:01 |
| russell_h | NobodyCam: right, if I understand correctly, that passes the node ID in as a kernel parameter | 00:02 |
| NobodyCam | yep | 00:02 |
| NobodyCam | if I recall correctly | 00:02 |
| russell_h | I'd rather not do that if possible | 00:02 |
| NobodyCam | ahh | 00:04 |
| russell_h | I'm leaning towards adding a vendor_passthru type method off the drivers API | 00:05 |
| russell_h | so I could do something like POST to /drivers/my_custom_driver/new_vendor_passthru/locate_node | 00:06 |
| NobodyCam | how would it work? | 00:06 |
| russell_h | then the API would pick a random conductor with that driver | 00:06 |
| *** dkehn has quit IRC | 00:06 | |
| russell_h | Is there a desire that the API not directly execute driver code? | 00:07 |
| *** dkehn has joined #openstack-ironic | 00:07 | |
| NobodyCam | what would you use to identify the node you are trying locate | 00:08 |
| *** dkehn_ has quit IRC | 00:09 | |
| russell_h | I don't want to specify that as part of my driver's API, I'd just submit an inventory of the server to it and it does its best to find or create a node | 00:09 |
| *** dkehn is now known as dkehn_ | 00:09 | |
| *** harlowja is now known as harlowja_away | 00:09 | |
| *** dkehn_ has quit IRC | 00:09 | |
| russell_h | but it'd be like looking at serial numbers and mac addresses | 00:09 |
| *** dkehn has joined #openstack-ironic | 00:10 | |
| *** dkehn_ has joined #openstack-ironic | 00:10 | |
| NobodyCam | hummm | 00:10 |
| russell_h | the nice thing about this is that all I need to do is power on a server, it PXE boots an image with only the Ironic API URL as a parameter and automatically adds itself to inventory | 00:11 |
| russell_h | I'm sure I'd wind up needing to work out.. other complications | 00:11 |
| NobodyCam | ya .. :) | 00:11 |
| NobodyCam | that auto add need to be configurable | 00:12 |
| NobodyCam | not everyone will want that | 00:12 |
| russell_h | right, well I plan to put all of this in a driver | 00:12 |
| russell_h | really all I need is a way to call my code without having to know a node ID | 00:13 |
| NobodyCam | i could see a vendor_passthru/locate_nodeuuid?mac=aa:bb:cc:dd:ee:ff type call | 00:13 |
| russell_h | which isn't really possible today without me doing something terrible | 00:13 |
| russell_h | right | 00:13 |
| russell_h | I mean, I could also just make it possible to filter on a GET to /nodes | 00:13 |
| russell_h | but then I either need to upstream the inventory stuff, or figure out how to do filtering in a driver | 00:14 |
| NobodyCam | devananda: what do you think about some thing like "vendor_passthru/locate_nodeuuid?mac=aa:bb:cc:dd:ee:ff" returns node.uuid | 00:14 |
| russell_h | I don't think either of those is really appropriate | 00:14 |
| russell_h | NobodyCam: the thing is, the current vendor_passthru hangs off of a node | 00:14 |
| russell_h | its like /nodes/<node_uuid>/vendor_passthru | 00:14 |
| russell_h | the reference I made to "me doing something terrible" is where I add a fake node with UUID 000-000-000 whatever | 00:15 |
| *** romcheg1 has quit IRC | 00:15 | |
| russell_h | but | 00:15 |
| russell_h | I want to think I have some degree of engineering integrity left | 00:16 |
| NobodyCam | lol | 00:16 |
| NobodyCam | :) | 00:16 |
| NobodyCam | what if /port/<MAC ADDR>/vendor_passthru/locate_nodeuuid | 00:18 |
| NobodyCam | were to exist? | 00:18 |
| russell_h | I mean | 00:18 |
| russell_h | wait, are ports identified by mac addr? | 00:18 |
| russell_h | if they are I can just do GET /ports/<mac_addr> and pull the node UUID from there | 00:19 |
| NobodyCam | something is saying that they currently are not | 00:19 |
| russell_h | hmm | 00:20 |
| russell_h | it seems like you really need the node_uuid and optionally a port_uuid | 00:20 |
| russell_h | I could add something like | 00:21 |
| russell_h | GET /ports?mac_address=<mac_address> | 00:21 |
| russell_h | to filter the list | 00:21 |
| NobodyCam | ya .. I could see such a think here https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/port.py | 00:21 |
| NobodyCam | s/think/thing/ | 00:22 |
| *** dkehn__ has joined #openstack-ironic | 00:23 | |
| russell_h | wait | 00:24 |
| russell_h | huh | 00:24 |
| russell_h | https://github.com/openstack/ironic/blob/master/ironic/db/sqlalchemy/api.py#L108 | 00:24 |
| russell_h | so dbapi.get_port(mac_address) | 00:25 |
| devananda | back | 00:25 |
| russell_h | would appear to work | 00:25 |
| NobodyCam | ahh | 00:25 |
| russell_h | (which seems bizarre actually) | 00:25 |
| russell_h | but, in any case, isn't exposed in the API | 00:26 |
| NobodyCam | ya | 00:26 |
| russell_h | there's also no index on mac address | 00:26 |
| *** dkehn__ has quit IRC | 00:26 | |
| *** dkehn_ has quit IRC | 00:26 | |
| *** dkehn has quit IRC | 00:26 | |
| *** dkehn__ has joined #openstack-ironic | 00:27 | |
| NobodyCam | Wb devananda | 00:27 |
| russell_h | I could also just stand up an elasticsearch cluster and have it index everything | 00:28 |
| russell_h | and have my deploy agent hit that | 00:28 |
| *** dkehn has joined #openstack-ironic | 00:28 | |
| NobodyCam | but wouldn't that tie the driver to your search cluster | 00:30 |
| NobodyCam | lol :-p | 00:30 |
| russell_h | ha, yeah... | 00:31 |
| devananda | russell_h: so port has an index on mac | 00:31 |
| devananda | russell_h: but let me start from the beginning | 00:31 |
| devananda | russell_h: you are asking about an auto-discovery feature, yes? | 00:31 |
| devananda | russell_h: whereby a new node, not yet in the DB, would power on, boot a default/discovery ramdisk, then POST some information to the Ironic API | 00:31 |
| devananda | without any authentication | 00:32 |
| devananda | and get added to the DB | 00:32 |
| devananda | russell_h: that about right? | 00:37 |
| russell_h | devananda: right | 00:37 |
| russell_h | devananda: and if its already in the DB I want to get back its ID | 00:37 |
| devananda | russell_h: ok, so two things | 00:38 |
| devananda | 1) enrollment | 00:38 |
| devananda | 2) search for existing node | 00:38 |
| devananda | (2) is possible -- but you need to be authenticated | 00:38 |
| devananda | (1) is terrible without authentication | 00:38 |
| devananda | if you're concerned about security, then allowing anonymous enrollment of hardware exposes a pretty big security risk, no? | 00:39 |
| russell_h | well.. it won't be entirely anonymous | 00:40 |
| devananda | but i'm side tracking on security again | 00:40 |
| russell_h | but yeah, there are definitely risks | 00:40 |
| russell_h | (this is going to be happening inside of a more-or-less trusted network) | 00:40 |
| devananda | k | 00:40 |
| devananda | so if we put some caveats about security around it, then yes, auto-discovery by DHCP BOOT is definitely possible | 00:41 |
| devananda | the DHCP server should respond differently to BOOT requests coming from known vs unknown MACs | 00:41 |
| devananda | eg, if the server is already known, serve this ramdisk. if it's unknown, serve that ramdisk | 00:42 |
| russell_h | why? | 00:42 |
| devananda | you don't need to re-run discovery on a node that's already been discovered (unless something changed and you /want/ to do that) | 00:42 |
| russell_h | I actually want to avoid any of that and just have only one ramdisk with the API URL as its only parameter | 00:42 |
| devananda | why? | 00:43 |
| russell_h | I want to avoid having to reconfigure DHCP a bunch | 00:45 |
| devananda | russell_h: why is that? | 00:46 |
| russell_h | like if I understand the PXE deploy driver, every time the hash ring gets rebalanced I'm going to have to make a neutron call for every port and rewrite a bunch of pxe configs | 00:46 |
| devananda | yep | 00:47 |
| devananda | "every time" == if a conductor comes on/goes offline | 00:47 |
| devananda | what is the problem you're trying to solve? | 00:48 |
| russell_h | less state in the conductors | 00:49 |
| russell_h | well, thats my goal | 00:50 |
| devananda | that's not a problem statement, that's a goal :) | 00:50 |
| devananda | i'm trying to see your motivation, but not getting it yet | 00:50 |
| russell_h | yeah, sorry, really distracted, trying to pretend like I'm paying attention in this meeting | 00:51 |
| devananda | :) | 00:51 |
| russell_h | devananda: let me stew on this overnight and come back to you with a more considered approach tomorrow | 00:51 |
| russell_h | I might be prematurely optimizing | 00:51 |
| russell_h | the short story is that I'm concerned about cascading failures when we lose a conductor | 00:52 |
| devananda | russell_h: sounds good. what you're asking isn't impossible - but i'd like to understand the reason behind it :) | 00:52 |
| devananda | ahh | 00:52 |
| devananda | the load from deploy is in transferring the user image | 00:53 |
| devananda | not the kernel& ramdisk, which are very small by comparison | 00:53 |
| devananda | user image is likely several GB. k & r probably no larger than 15MB | 00:53 |
| devananda | so in eg a power failure / recovery situation, we're only looking at PXE boot transferring the k & r -- not the user images | 00:54 |
| devananda | russell_h: fwiw, i'm sure we'll hit some scaling issues once ironic starts being deployed | 00:56 |
| devananda | russell_h: i'd like us to focus on getting _there_ before we look at what comes after ;) | 00:57 |
| russell_h | fair | 00:57 |
| NobodyCam | doh found my missing ssh key :( *face in palms* | 00:59 |
| NobodyCam | ok /me calls it a day too | 01:09 |
| *** dkehn_ has joined #openstack-ironic | 01:17 | |
| *** dkehn_ has quit IRC | 01:19 | |
| *** dkehn has quit IRC | 01:19 | |
| *** dkehn_ has joined #openstack-ironic | 01:20 | |
| *** dkehn__ has quit IRC | 01:20 | |
| *** dkehn has joined #openstack-ironic | 01:21 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Fix leaking DB details to API on error https://review.openstack.org/73120 | 01:27 |
| JoshNang | Whoops, grabbed the wrong bug ID there. | 01:29 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Fix leaking DB details to API on error https://review.openstack.org/73121 | 01:33 |
| *** rloo has quit IRC | 01:33 | |
| *** harlowja_away is now known as harlowja | 01:43 | |
| *** nosnos has joined #openstack-ironic | 02:00 | |
| openstackgerrit | Yongli He proposed a change to openstack/ironic: Use six.moves.urllib.parse instead of urlparse https://review.openstack.org/72889 | 02:02 |
| *** wusWWW has quit IRC | 02:09 | |
| *** wusWWW has joined #openstack-ironic | 02:10 | |
| *** mdurnosvistov_ has joined #openstack-ironic | 02:13 | |
| *** harlowja_ has joined #openstack-ironic | 02:13 | |
| *** aignatov_ has quit IRC | 02:13 | |
| *** mtaylor has joined #openstack-ironic | 02:13 | |
| *** mikal_ has joined #openstack-ironic | 02:14 | |
| *** mikal has quit IRC | 02:14 | |
| *** BLZbubba has quit IRC | 02:14 | |
| *** mordred has quit IRC | 02:15 | |
| *** SpamapS has quit IRC | 02:15 | |
| *** russell_h has quit IRC | 02:15 | |
| *** nosnos_ has joined #openstack-ironic | 02:15 | |
| *** davidlenwell_ has joined #openstack-ironic | 02:15 | |
| devananda | JoshNang: why two patch sets? | 02:16 |
| *** max_lobur has joined #openstack-ironic | 02:16 | |
| *** dkehn__ has joined #openstack-ironic | 02:16 | |
| *** wusWWW has quit IRC | 02:16 | |
| devananda | JoshNang: oh - you can amend the commit and post a new revision without creating a whole new patch set | 02:16 |
| devananda | JoshNang: https://wiki.openstack.org/wiki/GerritWorkflow#Updating_a_Change | 02:17 |
| *** max_lobur_afk has quit IRC | 02:18 | |
| *** dkehn_ has quit IRC | 02:18 | |
| *** mdurnosvistov has quit IRC | 02:19 | |
| *** dshulyak has quit IRC | 02:19 | |
| *** harlowja has quit IRC | 02:19 | |
| *** creiht has quit IRC | 02:19 | |
| *** pradipta has quit IRC | 02:19 | |
| *** nosnos has quit IRC | 02:22 | |
| *** davidlenwell has quit IRC | 02:22 | |
| *** pquerna has quit IRC | 02:22 | |
| *** hemna has quit IRC | 02:22 | |
| openstackgerrit | Yongli He proposed a change to openstack/ironic: Use six.moves.urllib.parse instead of urlparse https://review.openstack.org/72889 | 02:22 |
| *** BLZbubba has joined #openstack-ironic | 02:23 | |
| *** pquerna_ has joined #openstack-ironic | 02:23 | |
| *** aignatov_ has joined #openstack-ironic | 02:23 | |
| JoshNang | devananda: sorry about that. I had too many tabs open at the end of the day. Thanks for the link. Hopefully I won't need it next time :) | 02:23 |
| *** aignatov_ is now known as aignatov | 02:23 | |
| devananda | JoshNang: no worries | 02:24 |
| *** SpamapS has joined #openstack-ironic | 02:24 | |
| *** SpamapS has quit IRC | 02:24 | |
| *** SpamapS has joined #openstack-ironic | 02:24 | |
| *** BLZbubba has quit IRC | 02:26 | |
| *** davidlenwell has joined #openstack-ironic | 02:28 | |
| *** max_lobur has quit IRC | 02:28 | |
| *** SpamapS has quit IRC | 02:29 | |
| *** SpamapS has joined #openstack-ironic | 02:30 | |
| *** coolsvap has quit IRC | 02:30 | |
| *** max_lobur has joined #openstack-ironic | 02:30 | |
| *** nosnos has joined #openstack-ironic | 02:32 | |
| *** dkehn__ has quit IRC | 02:33 | |
| *** davidlenwell_ has quit IRC | 02:33 | |
| *** nosnos_ has quit IRC | 02:33 | |
| *** aignatov has quit IRC | 02:34 | |
| *** russell_h has joined #openstack-ironic | 02:34 | |
| *** dkehn has quit IRC | 02:35 | |
| *** SpamapS has quit IRC | 02:36 | |
| *** SpamapS has joined #openstack-ironic | 02:36 | |
| *** SpamapS has quit IRC | 02:36 | |
| *** SpamapS has joined #openstack-ironic | 02:36 | |
| *** BLZbubba has joined #openstack-ironic | 02:38 | |
| *** pquerna_ has quit IRC | 02:38 | |
| *** pquerna has joined #openstack-ironic | 02:38 | |
| *** aignatov_ has joined #openstack-ironic | 02:39 | |
| *** aignatov_ is now known as aignatov | 02:40 | |
| *** thedodd has joined #openstack-ironic | 02:41 | |
| *** dkehn_ has joined #openstack-ironic | 02:42 | |
| *** mdurnosvistov has joined #openstack-ironic | 02:43 | |
| *** russellb_ has joined #openstack-ironic | 02:45 | |
| *** BLZbubba_ has joined #openstack-ironic | 02:46 | |
| *** pquerna has quit IRC | 02:47 | |
| *** BLZbubba has quit IRC | 02:47 | |
| *** marty has joined #openstack-ironic | 02:47 | |
| *** creiht_ has joined #openstack-ironic | 02:48 | |
| *** mgoddard_ has joined #openstack-ironic | 02:49 | |
| *** mikal has joined #openstack-ironic | 02:49 | |
| *** nosnos has quit IRC | 02:49 | |
| *** mikal_ has quit IRC | 02:49 | |
| *** mdurnosvistov_ has quit IRC | 02:49 | |
| *** rwsu has quit IRC | 02:50 | |
| *** hstimer has quit IRC | 02:50 | |
| *** mgoddard has quit IRC | 02:50 | |
| *** lifeless has quit IRC | 02:50 | |
| *** jrist has quit IRC | 02:50 | |
| *** anteaya has quit IRC | 02:50 | |
| *** russellb has quit IRC | 02:50 | |
| *** pquerna has joined #openstack-ironic | 02:50 | |
| *** russellb_ is now known as russellb | 02:50 | |
| *** nosnos has joined #openstack-ironic | 02:56 | |
| *** anteaya has joined #openstack-ironic | 02:56 | |
| *** jrist has joined #openstack-ironic | 02:57 | |
| *** lifeless has joined #openstack-ironic | 02:58 | |
| *** pquerna_ has joined #openstack-ironic | 02:59 | |
| *** rwsu has joined #openstack-ironic | 03:01 | |
| *** pleia2 has quit IRC | 03:03 | |
| *** pleia2 has joined #openstack-ironic | 03:04 | |
| *** pquerna has quit IRC | 03:04 | |
| *** creiht_ is now known as creiht | 03:16 | |
| *** dkehn_ has quit IRC | 03:21 | |
| *** dkehn has joined #openstack-ironic | 03:22 | |
| *** hemna has joined #openstack-ironic | 03:29 | |
| *** harlowja_ has quit IRC | 03:30 | |
| *** harlowja_ has joined #openstack-ironic | 03:30 | |
| *** mikal has quit IRC | 04:26 | |
| *** mikal has joined #openstack-ironic | 04:27 | |
| *** thedodd has quit IRC | 05:12 | |
| *** john-n-seattle2 has quit IRC | 05:40 | |
| *** pradipta has joined #openstack-ironic | 05:41 | |
| *** dkehn has quit IRC | 05:52 | |
| *** loki__ has joined #openstack-ironic | 05:54 | |
| *** dkehn has joined #openstack-ironic | 05:54 | |
| *** marty has quit IRC | 05:54 | |
| *** dkehn_ has joined #openstack-ironic | 06:06 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/71192 | 06:06 |
| openstackgerrit | Yongli He proposed a change to openstack/ironic: Remove extraneous vim configuration comments for horizon and tools https://review.openstack.org/73160 | 06:12 |
| Haomeng | devananda: ping | 06:12 |
| openstackgerrit | A change was merged to openstack/ironic: Updated from global requirements https://review.openstack.org/71449 | 06:14 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add testing and doc sections to docs/dev-quickstart https://review.openstack.org/71638 | 06:17 |
| *** pquerna_ is now known as pquerna | 06:24 | |
| loki__ | hi planning to add documentation to ironic let me know the requirements | 06:32 |
| *** mgoddard has joined #openstack-ironic | 06:35 | |
| *** john-n-seattle has joined #openstack-ironic | 06:37 | |
| *** mgoddard_ has quit IRC | 06:37 | |
| openstackgerrit | Yongli He proposed a change to openstack/python-ironicclient: Remove vim header https://review.openstack.org/73166 | 06:40 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Fix Iterface misspelling from node-validate cmd https://review.openstack.org/73082 | 06:43 |
| *** harlowja_ is now known as harlowja_away | 06:44 | |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Remove tox locale overrides https://review.openstack.org/72237 | 06:45 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Add node.states() to the client library https://review.openstack.org/70979 | 06:50 |
| openstackgerrit | A change was merged to openstack/ironic: Remove unnecessary json dumps/loads from tests https://review.openstack.org/70413 | 06:52 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/python-ironicclient: Node power state is not printed after set https://review.openstack.org/73170 | 06:52 |
| openstackgerrit | A change was merged to openstack/ironic: Fix API inconsistence when changing node's states https://review.openstack.org/67506 | 06:54 |
| *** john-n-seattle has quit IRC | 07:03 | |
| openstackgerrit | Yongli He proposed a change to openstack/python-ironicclient: Remove vim header https://review.openstack.org/73166 | 07:08 |
| *** rongze has joined #openstack-ironic | 07:17 | |
| *** dkehn__ has joined #openstack-ironic | 07:19 | |
| *** rsacharya has joined #openstack-ironic | 07:21 | |
| *** dkehn__ has quit IRC | 07:21 | |
| *** dkehn__ has joined #openstack-ironic | 07:21 | |
| *** dkehn_ has quit IRC | 07:22 | |
| *** dkehn has quit IRC | 07:22 | |
| *** dkehn has joined #openstack-ironic | 07:24 | |
| *** lazy_prince has joined #openstack-ironic | 07:40 | |
| GheRivero | morning all | 07:43 |
| *** john-n-seattle has joined #openstack-ironic | 07:44 | |
| *** ifarkas has joined #openstack-ironic | 07:47 | |
| *** john-n-seattle has quit IRC | 07:57 | |
| *** hemna has quit IRC | 08:05 | |
| *** SpamapS has quit IRC | 08:14 | |
| *** greghaynes has quit IRC | 08:14 | |
| *** bigjools_ has joined #openstack-ironic | 08:15 | |
| *** SpamapS has joined #openstack-ironic | 08:15 | |
| *** SpamapS has quit IRC | 08:15 | |
| *** SpamapS has joined #openstack-ironic | 08:15 | |
| *** bigjools has quit IRC | 08:15 | |
| *** creiht has quit IRC | 08:15 | |
| *** aignatov has quit IRC | 08:15 | |
| *** creiht has joined #openstack-ironic | 08:16 | |
| *** openstackgerrit has quit IRC | 08:16 | |
| *** aignatov_ has joined #openstack-ironic | 08:17 | |
| *** aignatov_ is now known as aignatov | 08:17 | |
| *** lifeless has quit IRC | 08:17 | |
| *** lifeless has joined #openstack-ironic | 08:19 | |
| *** comstud has quit IRC | 08:21 | |
| *** comstud has joined #openstack-ironic | 08:22 | |
| *** romcheg1 has joined #openstack-ironic | 08:46 | |
| *** yuriyz has joined #openstack-ironic | 08:47 | |
| *** saju_m has joined #openstack-ironic | 08:50 | |
| Haomeng | GheRivero: morning:) | 08:52 |
| *** dshulyak has joined #openstack-ironic | 09:06 | |
| *** derekh has joined #openstack-ironic | 09:09 | |
| romcheg1 | Morning Haomeng, GheRivero! How are you? | 09:09 |
| romcheg1 | nick /romcheg_ltp | 09:10 |
| romcheg1 | whoops | 09:10 |
| *** romcheg1 is now known as romcheg_ltp | 09:10 | |
| yuriyz | morning all | 09:10 |
| romcheg_ltp | Morning yuriyz | 09:10 |
| romcheg_ltp | yuriyz: We had a discussion about you patch that exposes lock breaking features to the API last night | 09:11 |
| romcheg_ltp | So if you're interested you can read the logs | 09:11 |
| yuriyz | ok, thanks | 09:11 |
| yuriyz | romcheg, what is ltp? | 09:12 |
| *** john-n-seattle has joined #openstack-ironic | 09:12 | |
| romcheg_ltp | LapToP :) | 09:12 |
| romcheg_ltp | I have IRC client running on my PC at the office so I cannot use romcheg | 09:13 |
| *** john-n-seattle has quit IRC | 09:21 | |
| *** lucasagomes has joined #openstack-ironic | 09:30 | |
| *** john-n-seattle has joined #openstack-ironic | 09:30 | |
| *** athomas has joined #openstack-ironic | 09:37 | |
| mdurnosvistov | Morning all! :) | 09:52 |
| Haomeng | romcheg_ltp: morning:) | 09:52 |
| Haomeng | romcheg_ltp: find and you? | 09:52 |
| romcheg_ltp | Haomeng: And I find :D | 09:53 |
| Haomeng | romcheg_ltp: :) | 09:53 |
| *** tatyana has joined #openstack-ironic | 09:55 | |
| *** bigjools_ is now known as bigjools | 10:04 | |
| *** bigjools has quit IRC | 10:05 | |
| *** bigjools has joined #openstack-ironic | 10:05 | |
| lucasagomes | morning all! | 10:06 |
| *** greghaynes has joined #openstack-ironic | 10:14 | |
| *** martyntaylor has joined #openstack-ironic | 10:26 | |
| max_lobur | morning Ironic | 10:28 |
| *** killer_prince has quit IRC | 10:30 | |
| Haomeng | max_lobur: morning:) | 10:47 |
| Haomeng | lucasagomes: morning:) | 10:47 |
| *** rsacharya has quit IRC | 11:06 | |
| *** pbrooko has joined #openstack-ironic | 11:10 | |
| romcheg_ltp | Morning lucasagomes | 11:16 |
| romcheg_ltp | and max_lobur | 11:16 |
| max_lobur | morning | 11:16 |
| Haomeng | max_lobur: :) | 11:18 |
| *** pleia2 has quit IRC | 11:31 | |
| *** pleia2 has joined #openstack-ironic | 11:32 | |
| *** dkehn_ has joined #openstack-ironic | 11:32 | |
| *** dkehn_ has quit IRC | 11:35 | |
| *** dkehn__ has quit IRC | 11:35 | |
| *** dkehn has quit IRC | 11:35 | |
| lucasagomes | Haomeng, romcheg_ltp morning | 11:35 |
| *** dkehn_ has joined #openstack-ironic | 11:35 | |
| lucasagomes | max_lobur, morning | 11:35 |
| *** dkehn has joined #openstack-ironic | 11:37 | |
| max_lobur | hey lucasagomes | 11:40 |
| lucasagomes | anyone can have a q look at https://review.openstack.org/#/c/72698/ ? | 11:40 |
| max_lobur | needs +2? | 11:41 |
| Haomeng | lucasagomes: Lucas, I remember nova compute driver has such same interface methods:) | 11:41 |
| lucasagomes | max_lobur, heh yea | 11:41 |
| lucasagomes | cause I want to add another patch which will be a dependency of 72998 | 11:42 |
| max_lobur | lucasagomes, done | 11:42 |
| lucasagomes | but in gerrit I can't have 2 patches as dependency | 11:42 |
| lucasagomes | only linear dependency :( | 11:42 |
| lucasagomes | max_lobur, cheers | 11:42 |
| Haomeng | lucasagomes: how do you think what is differece between our nova computer console interface and our ironic console interface? | 11:42 |
| lucasagomes | Haomeng, yea :D they do for diff types of protocols | 11:42 |
| lucasagomes | vnc, spice etc | 11:42 |
| Haomeng | lucasagomes: yes | 11:42 |
| max_lobur | btw, how you create dependent patches in gerrit? | 11:42 |
| Haomeng | lucasagomes: but how nova launch our ironic console? | 11:42 |
| lucasagomes | Haomeng, their driver don't have a dedicated console interface | 11:42 |
| max_lobur | do I need to create branch from parent patch branch? | 11:43 |
| lucasagomes | so they need to add one method for each protocol | 11:43 |
| lucasagomes | where us have an abstraction | 11:43 |
| Haomeng | lucasagomes: yes | 11:43 |
| lucasagomes | max_lobur, exactly | 11:43 |
| max_lobur | and gerrit handles that automatically? | 11:43 |
| max_lobur | or needs some additional comments magic> | 11:43 |
| max_lobur | ? | 11:43 |
| lucasagomes | Haomeng, so the api trigger is being added by 72998 | 11:43 |
| Haomeng | lucasagomes: but i still has concern, how to launch our console from nova? | 11:43 |
| lucasagomes | max_lobur, ah so | 11:43 |
| yuriyz | max_lobur, http://www.mediawiki.org/wiki/Gerrit/Advanced_usage | 11:43 |
| lucasagomes | max_lobur, you create a new branch from the parent one | 11:44 |
| lucasagomes | add the changes | 11:44 |
| Haomeng | lucasagomes: yes we have api already, great work:) | 11:44 |
| lucasagomes | and then do a "git review -R" | 11:44 |
| lucasagomes | Haomeng, yea not yet :) being added | 11:44 |
| lucasagomes | but not merged yet heh | 11:44 |
| lucasagomes | hope soon | 11:44 |
| Haomeng | lucasagomes: great | 11:44 |
| lucasagomes | max_lobur, yuriyz links describe it better than I did :D thanks yuriyz | 11:45 |
| Haomeng | lucasagomes: for this bp - https://blueprints.launchpad.net/ironic/+spec/serial-console-access , we support serial console for ironic bm nodes | 11:46 |
| Haomeng | lucasagomes: so how to launch this serial console from horizon like the existing console function | 11:46 |
| max_lobur | lucasagomes, yuriyz thanks | 11:47 |
| Haomeng | lucasagomes: should we add nova compute interface and add horizon code to support our serial console, I think that is Unified entrance to launch console | 11:47 |
| lucasagomes | Haomeng, yea, first we need a way to trigger it in the ironic api | 11:47 |
| Haomeng | lucasagomes: yes | 11:47 |
| lucasagomes | Haomeng, then we could use the ironic driver on nova | 11:47 |
| Haomeng | lucasagomes: ok | 11:48 |
| lucasagomes | to get the request from nova | 11:48 |
| Haomeng | lucasagomes: great:) | 11:48 |
| lucasagomes | and route to ironic | 11:48 |
| lucasagomes | using our client libraries | 11:48 |
| Haomeng | lucasagomes: yes | 11:48 |
| Haomeng | lucasagomes: The final effect is we can launch our ironic ipmi node serial console from horizon:) | 11:49 |
| lucasagomes | Haomeng, yea | 11:49 |
| lucasagomes | that would be great | 11:50 |
| Haomeng | lucasagomes: maybe next release:) | 11:50 |
| lucasagomes | yea, in the moment the requirement is having us to support it | 11:50 |
| Haomeng | lucasagomes: yes | 11:51 |
| lucasagomes | integration with other projects will probably be part of J release | 11:51 |
| lucasagomes | cause of time | 11:51 |
| Haomeng | lucasagomes: yea | 11:51 |
| Haomeng | lucasagomes: we have to focus our ironic part first:) | 11:51 |
| lucasagomes | +1 | 11:52 |
| lucasagomes | :D | 11:52 |
| Haomeng | lucasagomes: :) | 11:53 |
| *** vkozhukalov has joined #openstack-ironic | 11:59 | |
| max_lobur | Anybody familiar with oslo.config? | 12:00 |
| max_lobur | at what point it considered fully loaded | 12:00 |
| max_lobur | e.g. the values from ironic.conf already available | 12:01 |
| agordeev | max_lobur: let me help you | 12:04 |
| max_lobur | agordeev, would be great | 12:06 |
| agordeev | max_lobur: i found the calling of parser. it's cfg.CONF() in prepare_service at ironic/common/service.py | 12:11 |
| max_lobur | https://github.com/openstack/ironic/blob/master/ironic/common/service.py#L65 | 12:12 |
| Haomeng | agordeev: cfg.CONF(argv[1:], project='ironic')? | 12:12 |
| max_lobur | this one? | 12:12 |
| agordeev | max_lobur: right! | 12:12 |
| max_lobur | thanks | 12:12 |
| max_lobur | this helps | 12:12 |
| Haomeng | max_lobur: Max, what is your question? | 12:12 |
| max_lobur | I need to check first | 12:13 |
| max_lobur | seems I found a bug | 12:13 |
| Haomeng | max_lobur: ok:) | 12:13 |
| *** loki__ has quit IRC | 12:14 | |
| Haomeng | max_lobur: I have another clue, cfg.CONF will try to search the <project>.conf file from different folders | 12:14 |
| Haomeng | max_lobur: the first folder is current folder, 2nd one is the current user's home, then more folders including /etc/<project>/<project>.conf | 12:15 |
| Haomeng | max_lobur: hope these helps you:) | 12:15 |
| max_lobur | ah, thanks | 12:16 |
| Haomeng | max_lobur: :) | 12:16 |
| max_lobur | mine is slightly different | 12:16 |
| max_lobur | filling it | 12:16 |
| Haomeng | max_lobur: ok | 12:16 |
| Haomeng | max_lobur: about "./tools/config/generate_sample.sh" ? | 12:18 |
| max_lobur | Haomeng, haha, nope :D | 12:21 |
| max_lobur | cold | 12:21 |
| Haomeng | max_lobur: :) | 12:22 |
| agordeev | max_lobur: so your question is at what stage do the option get into cfg.CONF ? | 12:24 |
| max_lobur | agordeev, yea, we have a few places broken | 12:24 |
| max_lobur | because they read CONF prior to initialization | 12:25 |
| max_lobur | almost completed the bug | 12:25 |
| agordeev | afair it gets loaded on python module importing stage | 12:25 |
| Haomeng | before prepare_service() method call | 12:26 |
| max_lobur | Haomeng, + | 12:26 |
| Haomeng | :) | 12:26 |
| Haomeng | max_lobur: because our "CONF = cfg.CONF" is global , outside of any class, will be run on py initialization time | 12:28 |
| max_lobur | yea, the object will be ok, but the options are default | 12:29 |
| Haomeng | max_lobur: yea | 12:29 |
| Haomeng | let me check other project's behavior | 12:29 |
| *** dshulyak has quit IRC | 12:30 | |
| *** dshulyak has joined #openstack-ironic | 12:31 | |
| Haomeng | can we move "cfg.CONF(argv[1:], project='ironic')" code to the position before prepare_service() call? | 12:33 |
| max_lobur | https://bugs.launchpad.net/ironic/+bug/1279774 | 12:35 |
| Haomeng | cool:) | 12:36 |
| max_lobur | well, not very cool :) | 12:37 |
| Haomeng | :) | 12:37 |
| max_lobur | Idk how to fix https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L350 without obfuscating the code | 12:37 |
| max_lobur | periodic services would need to configured after start | 12:38 |
| max_lobur | and so on | 12:38 |
| max_lobur | yuriyz, lucasagomes , what do you think about 2nd part of deva's coment ? https://review.openstack.org/#/c/69135/5/ironic/conductor/worker_pool.py | 12:40 |
| max_lobur | do you feel like a worker pool should better be built in into conductor? | 12:40 |
| max_lobur | or it's better to have it as a separate util | 12:41 |
| max_lobur | I thought it might be used outside conductor as well.. but not sure | 12:42 |
| max_lobur | devananda, once you're here, pls take a look at my last comment at https://review.openstack.org/#/c/69135/4/ironic/conductor/worker_pool.py (I noticed that with worker pool we actually doubling possible IO workload) | 12:47 |
| *** bigjools has quit IRC | 12:48 | |
| *** bigjools has joined #openstack-ironic | 12:48 | |
| *** bigjools has joined #openstack-ironic | 12:48 | |
| lucasagomes | max_lobur, I will take a look, yea interesting comment about oslo instantiation I didn't know that as well | 12:48 |
| lucasagomes | about being baked inside the conductor hmm | 12:49 |
| lucasagomes | only conductors are going to use that pool right? So I think it makes sense to have a scope on it | 12:49 |
| lucasagomes | (the worker_pool.py is placed /conductor/ as well) | 12:50 |
| max_lobur | lucasagomes, that's the question :) Do we want to issue background workers from some deploy utils, or from conductor only | 12:50 |
| max_lobur | yea, probably better to issue from conductor only, otherwise it will be mess | 12:51 |
| lucasagomes | max_lobur, heh I'm not sure | 12:51 |
| lucasagomes | well if it's our use case for now we could just have it in the conductor | 12:51 |
| lucasagomes | if we need it on the utils after we could move it | 12:51 |
| lucasagomes | and even create a singleton class that would manage instantiate and manage the pool | 12:52 |
| lucasagomes | would instantiate and manage* | 12:52 |
| max_lobur | lucasagomes, agree | 12:52 |
| lucasagomes | so other parts of the code can use that | 12:52 |
| lucasagomes | but for now just adding to the conductor is grand | 12:52 |
| lucasagomes | and does the job we need | 12:52 |
| * lucasagomes going to grab some lunch | 12:53 | |
| *** lucasagomes is now known as lucas-hungry | 12:53 | |
| *** nosnos has quit IRC | 13:10 | |
| *** jdob has joined #openstack-ironic | 13:14 | |
| *** jbjohnso has joined #openstack-ironic | 13:17 | |
| *** rloo has joined #openstack-ironic | 13:38 | |
| *** lucas-hungry is now known as lucasagomes | 13:48 | |
| *** matty_dubs|gone is now known as matty_dubs | 13:58 | |
| *** tatyana_ has joined #openstack-ironic | 14:07 | |
| *** tatyana has quit IRC | 14:09 | |
| *** tatyana_ is now known as tatyana | 14:09 | |
| yuriyz | lucasagomes, next bug in wsme https://bugs.launchpad.net/wsme/+bug/1279757 | 14:13 |
| lucasagomes | yuriyz, ouch lemme take a look | 14:14 |
| lucasagomes | yuriyz, I've two patches in wsme that I submitted long time ago | 14:14 |
| lucasagomes | :( no reviews | 14:14 |
| lucasagomes | well there's reviews but it's hanging in the queue for looooong time | 14:15 |
| lucasagomes | https://review.openstack.org/#/q/status:open+project:stackforge/wsme,n,z | 14:15 |
| *** rongze has quit IRC | 14:21 | |
| *** saju_m has quit IRC | 14:28 | |
| *** rongze has joined #openstack-ironic | 14:29 | |
| max_lobur | bbiaw | 14:35 |
| *** max_lobur is now known as max_lobur_afk | 14:36 | |
| NobodyCam | good morning Ironic says the man mak'n coffee | 15:08 |
| matty_dubs | NobodyCam: for all of us? | 15:09 |
| matty_dubs | NobodyCam: ;) Good morning! | 15:09 |
| NobodyCam | matty_dubs: hehhee I only have a single cup coffee maker in this trailer :-p | 15:09 |
| NobodyCam | oh and good morning :) | 15:11 |
| *** rloo has quit IRC | 15:15 | |
| *** rloo has joined #openstack-ironic | 15:16 | |
| *** rloo has quit IRC | 15:16 | |
| *** rloo has joined #openstack-ironic | 15:17 | |
| GheRivero | morning all | 15:19 |
| matty_dubs | Howdy GheRivero | 15:19 |
| lucasagomes | morning folks NobodyCam matty_dubs GheRivero | 15:19 |
| NobodyCam | morning GheRivero & lucasagomes | 15:20 |
| devananda | morning, all | 15:20 |
| lucasagomes | morning devananda | 15:21 |
| NobodyCam | morning devananda :) | 15:21 |
| devananda | lucasagomes: try poking jd__ on irc? | 15:21 |
| NobodyCam | max_lobur_afk: any reason you didn't +a https://review.openstack.org/#/c/72698/ ?? | 15:22 |
| lucasagomes | devananda, not directly but I've asked on the #wsme channel for reviews | 15:23 |
| matty_dubs | Anyone heard from Sun Jing? I just noticed new comments from her on https://review.openstack.org/#/c/64100/ | 15:24 |
| matty_dubs | Wonder if it's getting revived | 15:24 |
| devananda | lucasagomes: i'd try pinging him or ryanpetrello in openstack-dev, particularly if if's fixing a bug and taking this long and you're not getting a response elsewhere | 15:25 |
| *** loki__ has joined #openstack-ironic | 15:26 | |
| NobodyCam | matty_dubs: Sun got reassigned by IBM | 15:26 |
| NobodyCam | I have been looking for Linggao | 15:26 |
| NobodyCam | who was going to pick up Sun's patches | 15:27 |
| lucasagomes | devananda, ack lemme ping them | 15:27 |
| matty_dubs | NobodyCam: She just commented on the console patch last night, though. Hence my confusion. | 15:27 |
| devananda | so folks, just got an answer from ttx re: graduation deadline -- "some time in march, before the 28th" | 15:27 |
| devananda | somewhere between i3 and release | 15:28 |
| NobodyCam | devananda: ok :) | 15:28 |
| lucasagomes | devananda, oh good | 15:29 |
| devananda | so the code sprint in Santa CLara will reeallly be a sprint :) | 15:29 |
| NobodyCam | :) | 15:29 |
| devananda | quick check - do the folks here want to try our first code jam today? | 15:30 |
| lucasagomes | btw, devananda max_lobur_afk, take a look at #64336 please. markmc just commented on that | 15:30 |
| lucasagomes | I think we should consider his points | 15:31 |
| lucasagomes | tl;dr, sync the last object code from nova, get the dick-like behaviour removed in nova and submit the code to oslo | 15:31 |
| NobodyCam | lucasagomes: you have any time for a quick look at https://review.openstack.org/#/c/70468 | 15:32 |
| devananda | lucasagomes: yep - he's spot on. I tried to get it into oslo when we started, but there was push-back because ironic was new | 15:33 |
| devananda | lucasagomes: i also think there's absolutely no benefit to removing the dict-like behavoir in the first place | 15:33 |
| lucasagomes | devananda, consistence | 15:33 |
| lucasagomes | but yea, it's not urgent | 15:33 |
| devananda | i mean,yes, consistency - but if we only used dict-like access, that's fine too | 15:34 |
| lucasagomes | NobodyCam, sure thing! /me spawns a vm | 15:34 |
| lucasagomes | devananda, yes I agree with that | 15:34 |
| devananda | anyhow, i would rather see someone (mikhail?) move this into oslo than keep diverging it | 15:36 |
| lucasagomes | yeah, maybe we should accept the patches from that series but the one modifying the base.py | 15:37 |
| devananda | ++ | 15:37 |
| *** loki__ has quit IRC | 15:38 | |
| devananda | hm, so i need to take a call in ~20 minutes... | 15:39 |
| *** rloo has quit IRC | 15:39 | |
| devananda | i'll be unavailable at the start of our code jam, assuming folks want to do that today | 15:39 |
| *** rloo has joined #openstack-ironic | 15:39 | |
| lucasagomes | ack, we can start with the more cosmetic patches in the queue | 15:40 |
| lucasagomes | that won't raise much doubts etc | 15:40 |
| NobodyCam | i was going to ask max_lobur_afk about: https://review.openstack.org/#/c/68980 & https://review.openstack.org/#/c/68985 | 15:41 |
| NobodyCam | I was thinking start on page #3 of the outstanding reviews and work oldest to newest | 15:41 |
| *** dshulyak has quit IRC | 15:42 | |
| *** vkozhukalov has quit IRC | 15:42 | |
| devananda | NobodyCam: ++ | 15:43 |
| devananda | ya'll should also be aware of http://status.openstack.org/reviews/ | 15:44 |
| devananda | it may not be the best choice for ordering today -- the goal of that page is to put the most important / hottest activity at the top | 15:44 |
| devananda | but it's a good resource to check when reviewing | 15:44 |
| *** rloo has quit IRC | 15:45 | |
| NobodyCam | devananda: ++ /me bookmarks | 15:45 |
| *** rloo has joined #openstack-ironic | 15:46 | |
| lucasagomes | NobodyCam, ok I gave it a try, it didn't work but my env is not reliable | 15:46 |
| lucasagomes | NobodyCam, I was checking whether neutron would start dnsmasq at least | 15:46 |
| lucasagomes | it should right? | 15:46 |
| lucasagomes | I'm fine with that patch tho, code looks good | 15:46 |
| lucasagomes | thinking about +a that and then we can dig more into the problem | 15:47 |
| *** openstackgerrit has joined #openstack-ironic | 15:47 | |
| NobodyCam | lucasagomes: ++ thats the plan for today.. land and fix! | 15:47 |
| lucasagomes | NobodyCam, ack | 15:47 |
| lucasagomes | will do | 15:47 |
| lucasagomes | NobodyCam, +a | 15:49 |
| devananda | lucasagomes: on my doc patch, mind land+fixing it? | 15:49 |
| lucasagomes | devananda, grand for me | 15:49 |
| devananda | lucasagomes: there are, i believe, folks who are coming on board in HP now() who are looking at docs.openstack.org for that info | 15:49 |
| lucasagomes | devananda, ack, will +a and fix the fedora/rhel related steps | 15:50 |
| devananda | awesome, ty | 15:50 |
| lucasagomes | devananda, done | 15:52 |
| *** rloo has quit IRC | 15:52 | |
| *** rloo has joined #openstack-ironic | 15:53 | |
| rloo | devananda, lucasagomes, ahhh. wait. i have some other comments. | 15:53 |
| lucasagomes | rloo, ouch! | 15:53 |
| lucasagomes | rloo, idk if I can cancel that :/ | 15:53 |
| NobodyCam | lucasagomes: up for a rebase on https://review.openstack.org/#/c/68980 | 15:53 |
| lucasagomes | mind adding the comments anyway and I can address it at the review fixing the rhel/fedora stuff? | 15:53 |
| lucasagomes | rloo, ^ | 15:53 |
| *** rloo has quit IRC | 15:54 | |
| lucasagomes | NobodyCam, sure | 15:54 |
| *** max_lobur has joined #openstack-ironic | 15:54 | |
| max_lobur | back | 15:54 |
| NobodyCam | Wb max_lobur | 15:54 |
| *** aignatov is now known as aignatov_ | 15:54 | |
| NobodyCam | max_lobur: any reason you didn't +a https://review.openstack.org/#/c/72698/ ?? | 15:54 |
| *** rloo has joined #openstack-ironic | 15:54 | |
| lucasagomes | NobodyCam, btw, looking at max_lobur review at #68980 | 15:55 |
| NobodyCam | yes | 15:55 |
| lucasagomes | max_lobur, yea I'd like to remove the node parameters | 15:55 |
| lucasagomes | but hmm idk if it would be better to do it in a separated patch | 15:55 |
| lucasagomes | removing the node from all methods | 15:55 |
| lucasagomes | since it's node needed by any method not only validate | 15:55 |
| lucasagomes | thoughts? | 15:55 |
| max_lobur | NobodyCam: nope | 15:56 |
| max_lobur | +A 'ing | 15:57 |
| lucasagomes | NobodyCam, max_lobur if you guys can +a that I would be very happy, cause I want that patch to depend on another patch (the one adding the wsme custom boolean type) | 15:57 |
| lucasagomes | max_lobur, magic! thanks | 15:57 |
| NobodyCam | lucasagomes: 68980? | 15:58 |
| lucasagomes | NobodyCam, nop | 15:58 |
| lucasagomes | NobodyCam, 72698 | 15:58 |
| NobodyCam | Approved | 15:58 |
| NobodyCam | :-p | 15:58 |
| *** rloo has quit IRC | 15:59 | |
| max_lobur | lucasagomes: devananda was going to fix other method in subsequent patch | 15:59 |
| * devananda hops on a phone call .... bbiab | 15:59 | |
| lucasagomes | max_lobur, removing the node from validate u mean? | 16:00 |
| lucasagomes | max_lobur, right | 16:00 |
| *** rloo has joined #openstack-ironic | 16:00 | |
| max_lobur | yep | 16:00 |
| max_lobur | seems he commented that | 16:00 |
| NobodyCam | ok how are We tracking the fixes to patches we land that need them | 16:00 |
| NobodyCam | max_lobur: you mean his comment on https://review.openstack.org/#/c/70468 | 16:01 |
| *** rloo has quit IRC | 16:01 | |
| *** rloo has joined #openstack-ironic | 16:01 | |
| max_lobur | ah, right | 16:01 |
| max_lobur | sorry | 16:02 |
| max_lobur | looking what is 68980 | 16:02 |
| max_lobur | https://review.openstack.org/#/c/68980/2/ironic/conductor/manager.py | 16:03 |
| max_lobur | lucasagomes: I agree, separate patch would be ok | 16:03 |
| lucasagomes | max_lobur, ack | 16:03 |
| max_lobur | lemme change to +1 then | 16:03 |
| lucasagomes | max_lobur, I will rebase i | 16:03 |
| lucasagomes | it* | 16:03 |
| max_lobur | ah, ok | 16:03 |
| max_lobur | NobodyCam: seems we don't have any doc to track that | 16:04 |
| max_lobur | so we may forget | 16:04 |
| NobodyCam | with the remove node praram should we tg the reviews with : https://bugs.launchpad.net/ironic/+bug/1259346 | 16:04 |
| NobodyCam | do we want to file bugs.. that seems a tad overkill | 16:05 |
| max_lobur | NobodyCam: + | 16:05 |
| max_lobur | just need to publish fix right after landing | 16:06 |
| max_lobur | NobodyCam: what is tg? :) | 16:06 |
| NobodyCam | llol a type-o for tag | 16:07 |
| *** mtaylor is now known as mordred | 16:07 | |
| *** mordred has quit IRC | 16:07 | |
| *** mordred has joined #openstack-ironic | 16:07 | |
| max_lobur | :) | 16:07 |
| *** saju_m has joined #openstack-ironic | 16:08 | |
| max_lobur | NO | 16:08 |
| max_lobur | sorry) | 16:08 |
| NobodyCam | hehee | 16:08 |
| max_lobur | ignore this) | 16:08 |
| NobodyCam | don't need to yell | 16:08 |
| NobodyCam | lol | 16:08 |
| NobodyCam | :-p | 16:08 |
| max_lobur | NobodyCam: looks like that's it | 16:08 |
| max_lobur | IMO we should tag | 16:08 |
| NobodyCam | :) | 16:09 |
| *** ifarkas has quit IRC | 16:10 | |
| max_lobur | brb | 16:11 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Task object as paramater to validate() methods https://review.openstack.org/68980 | 16:13 |
| max_lobur | back | 16:15 |
| NobodyCam | we talked about switching away from sql Alchemy to alembic, but I don't recall what we decided? this is in ref to https://review.openstack.org/#/c/67415/ | 16:15 |
| NobodyCam | Wb max_lobur | 16:15 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: PXE and SSH validate() method to check for a port https://review.openstack.org/68985 | 16:16 |
| max_lobur | As I remember we decided to switch, because alchemy_migrate doesn't supported anymore | 16:16 |
| max_lobur | other projects migrating as well | 16:16 |
| lucasagomes | yea, devananda also sent an email about it | 16:18 |
| lucasagomes | the strategy is to move to alambic yes | 16:18 |
| lucasagomes | alembic* | 16:19 |
| max_lobur | yep, remembered email too | 16:20 |
| lucasagomes | uhul! https://review.openstack.org/#/c/64715/ got approved | 16:20 |
| NobodyCam | :) | 16:21 |
| openstackgerrit | A change was merged to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 16:23 |
| openstackgerrit | A change was merged to openstack/ironic: Add testing and doc sections to docs/dev-quickstart https://review.openstack.org/71638 | 16:23 |
| * devananda is back, sort of | 16:28 | |
| *** rsacharya has joined #openstack-ironic | 16:29 | |
| NobodyCam | Wb (kinda) devananda :-p | 16:29 |
| yuriyz | lucasagomes, https://review.openstack.org/#/c/71449/ | 16:30 |
| lucasagomes | yuriyz, ohhhh | 16:30 |
| lucasagomes | yuriyz, sorry I haven't seem that | 16:31 |
| lucasagomes | yuriyz, will re-vote | 16:31 |
| yuriyz | thanks | 16:31 |
| openstackgerrit | A change was merged to openstack/ironic: Add get_console() method https://review.openstack.org/72698 | 16:31 |
| lucasagomes | yuriyz, I'm sorry, I pulled my master yesterday so it didn't have the change when I looked at it | 16:32 |
| devananda | so, hi guys! looks like ya'll are getting it done! awesome! :-D | 16:33 |
| NobodyCam | i;m thinking the alambic would be best to wait on as it may collide with https://review.openstack.org/#/c/68413 | 16:33 |
| NobodyCam | hi devananda :) | 16:33 |
| devananda | I saw a question about how are we tracking follow-in patchsets - maybe an etherpad? | 16:34 |
| max_lobur | NobodyCam: nope, I think they different | 16:34 |
| max_lobur | devananda: I think we can just post them right after the original . If we track them - they become not follow-in :) | 16:37 |
| max_lobur | and it will be one more point to forget to update etc | 16:37 |
| devananda | i mean - who's posting them :) | 16:37 |
| devananda | max_lobur: but yes, they should just be posted | 16:37 |
| *** hemna has joined #openstack-ironic | 16:38 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Fix autodoc formatting in pxe.py https://review.openstack.org/66997 | 16:38 |
| NobodyCam | I'm thinking for the next review day we actually come up with some kinda of review agenda | 16:39 |
| max_lobur | so, are we starting our first review hangout hangout? :) | 16:39 |
| NobodyCam | max_lobur: we kinda have already | 16:39 |
| NobodyCam | :-p | 16:39 |
| max_lobur | :) | 16:39 |
| NobodyCam | but sure | 16:39 |
| devananda | NobodyCam: care to invite me? | 16:40 |
| max_lobur | how about to start from Get rid object model `dict` | 16:41 |
| max_lobur | 5 patches | 16:41 |
| max_lobur | 6th will be abandoned as per latest updates | 16:41 |
| *** pbrooko has quit IRC | 16:41 | |
| NobodyCam | max I dont have your gmail | 16:41 |
| devananda | romcheg_ltp: g-hang invite sent. max_lobur - i don't have your google contact either | 16:41 |
| max_lobur | mlobur@mirantis.com | 16:41 |
| lucasagomes | max_lobur, I think #1 also modifies base.py | 16:42 |
| max_lobur | that's google | 16:42 |
| *** athomas has quit IRC | 16:42 | |
| * max_lobur looking | 16:43 | |
| devananda | max_lobur: giving your latest patchset for the race condition a quick exercise | 16:43 |
| devananda | if nothing breaks, i think we should land it | 16:43 |
| max_lobur | devananda: k | 16:44 |
| max_lobur | I plan to post vendor patch today | 16:44 |
| max_lobur | lucasagomes: youre right | 16:44 |
| max_lobur | 2 lines in base | 16:44 |
| max_lobur | mdurnosvistov: are u around? | 16:44 |
| lucasagomes | yea | 16:44 |
| NobodyCam | max get the invite? | 16:45 |
| *** yuriyz has quit IRC | 16:45 | |
| max_lobur | yup | 16:45 |
| NobodyCam | :) | 16:45 |
| max_lobur | ah that's real hangout :) | 16:45 |
| max_lobur | lemme check if it works on my laptop | 16:46 |
| NobodyCam | :) | 16:46 |
| NobodyCam | https://etherpad.openstack.org/p/IronicReviewDay | 16:47 |
| *** athomas has joined #openstack-ironic | 16:49 | |
| NobodyCam | https://review.openstack.org/#/c/67415/ | 16:50 |
| devananda | http://paste.openstack.org/show/65096/ | 16:59 |
| *** matty_dubs is now known as matty_dubs|lunch | 17:01 | |
| NobodyCam | https://review.openstack.org/#/c/71974/ | 17:04 |
| NobodyCam | https://review.openstack.org/#/c/69135/6/ironic/api/controllers/v1/node.py | 17:07 |
| *** anniec has joined #openstack-ironic | 17:14 | |
| devananda | https://review.openstack.org/#/c/72632/ | 17:16 |
| *** rongze has quit IRC | 17:19 | |
| romcheg_ltp | mdurnosvistov: Ping | 17:19 |
| *** rongze has joined #openstack-ironic | 17:20 | |
| *** rongze has quit IRC | 17:24 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove None as default value for dict get() https://review.openstack.org/72632 | 17:25 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Log IPMI power on/off timeouts https://review.openstack.org/58266 | 17:28 |
| NobodyCam | https://review.openstack.org/#/c/72851 | 17:30 |
| *** martyntaylor has quit IRC | 17:31 | |
| NobodyCam | https://review.openstack.org/#/c/64674 | 17:33 |
| *** lazy_prince has quit IRC | 17:35 | |
| openstackgerrit | Roman Prykhodchenko proposed a change to openstack/ironic: Removes use of timeutils.set_time_override https://review.openstack.org/67432 | 17:36 |
| devananda | https://review.openstack.org/#/c/73005/ | 17:39 |
| NobodyCam | 73005 | 17:39 |
| devananda | rloo: for your patch ^, we'll need a change to the python lib and the CLI -- are you working on those? | 17:41 |
| devananda | rloo: also, thanks for ^ :) | 17:42 |
| rloo | devananda will work on it but didn't want to if this patch wasn't ok ;) | 17:43 |
| devananda | rloo: ah, so, actually we shouldn't load drivers in the AP Iservice | 17:43 |
| *** vkozhukalov has joined #openstack-ironic | 17:43 | |
| devananda | rloo: list of available drivers is specific to each conductor instance and can vary by config | 17:43 |
| devananda | they're loaded via entrypoints, so it's possible for different systems to have different drivers installed | 17:44 |
| rloo | devananda in meeting now. | 17:45 |
| devananda | k, i'll leave comments | 17:45 |
| openstackgerrit | A change was merged to openstack/ironic: Fix race condition when changing node states https://review.openstack.org/69135 | 17:45 |
| *** tatyana has quit IRC | 17:48 | |
| *** anniec has quit IRC | 17:48 | |
| rloo | devananda: back now. | 17:49 |
| openstackgerrit | Roman Prykhodchenko proposed a change to openstack/ironic: Process public API list as regular expressions https://review.openstack.org/66621 | 17:53 |
| openstackgerrit | A change was merged to openstack/ironic: Add topic to the change_node_maintenance_mode() RPC method https://review.openstack.org/72724 | 17:55 |
| *** matty_dubs|lunch is now known as matty_dubs | 17:55 | |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Return node_uuid from a port-show cmd https://review.openstack.org/73091 | 17:55 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Node power state is not printed after set https://review.openstack.org/73170 | 17:57 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Remove vim header https://review.openstack.org/73166 | 17:58 |
| NobodyCam | https://review.openstack.org/#/c/73006 | 17:59 |
| *** killer_prince has joined #openstack-ironic | 18:02 | |
| romcheg_ltp | https://review.openstack.org/#/c/72889 | 18:04 |
| *** rsacharya has quit IRC | 18:04 | |
| *** anniec has joined #openstack-ironic | 18:08 | |
| *** hstimer has joined #openstack-ironic | 18:12 | |
| *** hstimer has quit IRC | 18:13 | |
| *** marty_ has joined #openstack-ironic | 18:13 | |
| *** marty_ is now known as Guest19634 | 18:13 | |
| *** derekh has quit IRC | 18:13 | |
| *** harlowja_away is now known as harlowja_ | 18:15 | |
| NobodyCam | https://etherpad.openstack.org/p/IronicReviewDay | 18:16 |
| NobodyCam | Awesome meeting Guys! | 18:17 |
| NobodyCam | Thank You all | 18:17 |
| max_lobur | :) | 18:17 |
| max_lobur | did anybody use webcam from mac book air for hangouts? | 18:18 |
| max_lobur | mine does not recognized | 18:18 |
| NobodyCam | I did | 18:18 |
| max_lobur | what's the trick? | 18:18 |
| max_lobur | haswell air? | 18:18 |
| NobodyCam | ?? I just click video | 18:18 |
| max_lobur | nope | 18:18 |
| max_lobur | no device found | 18:18 |
| lucasagomes | gnight guys | 18:18 |
| * lucasagomes runs to catch the train | 18:19 | |
| max_lobur | night lucasagomes! | 18:19 |
| NobodyCam | night lucasagomes | 18:19 |
| *** lucasagomes has quit IRC | 18:19 | |
| matty_dubs | I have a 2010-vintage Air; Hangouts work for me but really tax the CPU | 18:19 |
| rloo | oops, missed lucasagomes. NobodyCam: what about https://review.openstack.org/#/c/71638/? Is lucas going to fix? | 18:19 |
| NobodyCam | I have two vbox vm running along with several other windows open | 18:19 |
| max_lobur | I'll try to google | 18:20 |
| max_lobur | would like to speak from beach as well :-З | 18:20 |
| NobodyCam | :) | 18:21 |
| NobodyCam | rloo: that is approved | 18:21 |
| rloo | NobodyCam. Yes, but lucas had two? comments about it and I think he said he would fix later. and I also had some comments. ie, some of the instructions don't work. | 18:22 |
| rloo | so should i just fix/address the comments i put in? | 18:22 |
| max_lobur | Also I'd like to discuss https://review.openstack.org/#/c/68413/4 once everyone here | 18:22 |
| max_lobur | brb | 18:22 |
| rloo | NobodyCam: just ignore me. I'll just add a fix to it, seems easiest. | 18:23 |
| NobodyCam | rloo: yea | 18:23 |
| NobodyCam | ok :-p | 18:23 |
| NobodyCam | rloo: also please add to https://etherpad.openstack.org/p/IronicReviewDay just so we know its being worked on | 18:24 |
| rloo | NobodyCam: okey dokey | 18:24 |
| NobodyCam | :) | 18:26 |
| *** anniec has quit IRC | 18:33 | |
| max_lobur | gate-ironic-python26 LOST what the hell | 18:34 |
| max_lobur | NobodyCam: what's the phone are you waiting for? :) | 18:36 |
| NobodyCam | lol a Lg f3q | 18:38 |
| NobodyCam | :-p only phone i could find with a real keyboard | 18:38 |
| NobodyCam | :-p i'm so old school | 18:38 |
| max_lobur | heh | 18:38 |
| romcheg_ltp | I' have to go for a while and clean up the apartment. Otherwise I'm going to be killed | 18:39 |
| romcheg_ltp | Will be later | 18:39 |
| NobodyCam | lol ok romcheg_ltp :) | 18:40 |
| NobodyCam | enjoy | 18:40 |
| romcheg_ltp | No :) | 18:40 |
| *** wusWWW has joined #openstack-ironic | 18:41 | |
| max_lobur | I recently bought http://en.meizu.com/products/mx3-product.html | 18:41 |
| max_lobur | and it's surprisingly cool | 18:42 |
| NobodyCam | nice | 18:42 |
| max_lobur | was about 450 $ here | 18:42 |
| *** athomas has quit IRC | 18:42 | |
| *** anniec has joined #openstack-ironic | 18:43 | |
| NobodyCam | max_lobur: http://www.youtube.com/watch?v=ct1_r_61sk8 | 18:45 |
| devananda | eep, jenkins failed all the remaining patch sets | 18:46 |
| * devananda goes and gets food anyawy | 18:46 | |
| max_lobur | NobodyCam: lol, something like that :D | 18:47 |
| NobodyCam | uggg | 18:47 |
| NobodyCam | :) | 18:47 |
| openstackgerrit | A change was merged to openstack/ironic: Add samples to serve API through Apache mod_wsgi https://review.openstack.org/72069 | 18:52 |
| openstackgerrit | A change was merged to openstack/ironic: Remove POST calls in tests for resource creation https://review.openstack.org/64674 | 18:53 |
| NobodyCam | oh I forgot to mention on the call I will be away most of tomorrow. Will be here in the morning but away after that | 18:55 |
| * NobodyCam needs to spend time with family | 18:56 | |
| openstackgerrit | A change was merged to openstack/ironic: Fix autodoc formatting in pxe.py https://review.openstack.org/66997 | 18:57 |
| openstackgerrit | A change was merged to openstack/ironic: Remove _load_one_plugin fallback https://review.openstack.org/73006 | 18:58 |
| max_lobur | NobodyCam: ack | 18:58 |
| *** anniec has quit IRC | 19:00 | |
| *** anniec has joined #openstack-ironic | 19:02 | |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic: Update docs/dev-quickstart https://review.openstack.org/73380 | 19:15 |
| * devananda has food now | 19:23 | |
| devananda | anyone working on the merge fails? NobodyCam ? max_lobur ? | 19:23 |
| devananda | if not, i'll start | 19:24 |
| NobodyCam | not I | 19:28 |
| *** rloo has quit IRC | 19:28 | |
| devananda | ok, taking stab at https://review.openstack.org/#/c/67415/ | 19:28 |
| *** rloo has joined #openstack-ironic | 19:28 | |
| max_lobur | not I yet | 19:35 |
| NobodyCam | started recheck on https://review.openstack.org/#/c/72851 | 19:35 |
| max_lobur | I'll fix my follow-ups firsh | 19:36 |
| max_lobur | *first | 19:36 |
| openstackgerrit | Victor Sergeyev proposed a change to openstack/ironic: WIP: Use oslo.db module (DRAFT) https://review.openstack.org/42159 | 19:38 |
| devananda | looks like the py26 failures may be coming from -infra | 19:38 |
| devananda | not a change we landed | 19:38 |
| devananda | they moved the py26 jobs to different nodes | 19:38 |
| NobodyCam | ya | 19:41 |
| NobodyCam | I just started a recheck on https://review.openstack.org/#/c/73248 | 19:41 |
| *** rloo has quit IRC | 19:42 | |
| *** rloo has joined #openstack-ironic | 19:42 | |
| devananda | if you're seeig py26 failures, hold off | 19:43 |
| devananda | infra is working on that | 19:43 |
| devananda | if it's something else -- carry on :) | 19:43 |
| NobodyCam | ack | 19:46 |
| *** rloo has quit IRC | 19:46 | |
| *** rloo has joined #openstack-ironic | 19:47 | |
| max_lobur | https://review.openstack.org/#/c/68413/ folks, what do you think about last comment? | 19:49 |
| NobodyCam | max_lobur: to show your encoding something a little different | 19:53 |
| NobodyCam | i like that | 19:53 |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace nonexistent mock assert methods with real ones https://review.openstack.org/70349 | 19:57 |
| devananda | max_lobur: https://review.openstack.org/#/c/73380/ could use another +2 | 20:04 |
| max_lobur | done | 20:06 |
| devananda | NobodyCam: https://review.openstack.org/#/c/70382/ as well | 20:06 |
| NobodyCam | looking | 20:07 |
| max_lobur | https://review.openstack.org/#/c/70382/ this one may not merge | 20:07 |
| max_lobur | I'm currently checking it | 20:07 |
| max_lobur | because of conflicts with race patch | 20:07 |
| devananda | ah | 20:09 |
| max_lobur | yep, merging it | 20:10 |
| max_lobur | there some conflicts | 20:11 |
| *** saju_m has quit IRC | 20:19 | |
| *** jrist is now known as jrist-afk | 20:24 | |
| *** jrist-afk is now known as jrist | 20:24 | |
| *** epim has joined #openstack-ironic | 20:36 | |
| openstackgerrit | A change was merged to openstack/ironic: Remove None as default value for dict get() https://review.openstack.org/72632 | 20:39 |
| *** anniec has quit IRC | 20:51 | |
| *** anniec has joined #openstack-ironic | 20:51 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/70382 | 20:57 |
| NobodyCam | so we're in holding pattern for the py26 fixes from infra? | 20:57 |
| *** anniec has quit IRC | 20:57 | |
| * NobodyCam looks for food! | 20:58 | |
| *** anniec has joined #openstack-ironic | 21:02 | |
| devananda | https://review.openstack.org/#/c/58266/ could use another +2/+A | 21:02 |
| devananda | NobodyCam: looks like they were transient? not sure yet | 21:02 |
| *** mdurnosvistov_ has joined #openstack-ironic | 21:05 | |
| NobodyCam | devananda: +a'd | 21:08 |
| *** thedodd has joined #openstack-ironic | 21:09 | |
| NobodyCam | your rechecking 67415 | 21:09 |
| NobodyCam | :0 | 21:09 |
| NobodyCam | :) | 21:09 |
| devananda | also needs a +A https://review.openstack.org/#/c/70349/ | 21:10 |
| openstackgerrit | Mikhail Durnosvistov proposed a change to openstack/ironic: Removes use of timeutils.set_time_override https://review.openstack.org/67432 | 21:14 |
| NobodyCam | ty for rebase mdurnosvistov | 21:17 |
| *** romcheg_ltp has quit IRC | 21:18 | |
| *** rloo has quit IRC | 21:21 | |
| *** rloo has joined #openstack-ironic | 21:21 | |
| NobodyCam | devananda: we're still getting 26 failures | 21:22 |
| devananda | :( | 21:22 |
| NobodyCam | :( seems intermitent | 21:23 |
| *** derekh has joined #openstack-ironic | 21:23 | |
| *** rloo has quit IRC | 21:24 | |
| *** rloo has joined #openstack-ironic | 21:25 | |
| devananda | https://review.openstack.org/#/c/70382/ passed | 21:29 |
| NobodyCam | ya | 21:30 |
| NobodyCam | and 67432 | 21:30 |
| NobodyCam | recheck? | 21:31 |
| devananda | hang on | 21:31 |
| devananda | fungi and i are looking at it. something is off.... | 21:31 |
| NobodyCam | :) | 21:31 |
| * NobodyCam jumps in to infra to lurk | 21:32 | |
| openstackgerrit | A change was merged to openstack/ironic: Log IPMI power on/off timeouts https://review.openstack.org/58266 | 21:38 |
| *** derekh is now known as derekh_afk | 21:39 | |
| *** vkozhukalov has quit IRC | 21:41 | |
| devananda | NobodyCam: looks like non-deterministic test fail in py26 ... | 21:44 |
| NobodyCam | oh joy | 21:44 |
| devananda | max_lobur: i see you working on https://review.openstack.org/#/c/70382/ -- is it ready to go? | 21:46 |
| openstackgerrit | A change was merged to openstack/ironic: Replace nonexistent mock assert methods with real ones https://review.openstack.org/70349 | 21:47 |
| *** anniec has quit IRC | 21:48 | |
| max_lobur | devananda: yea it's ready | 21:53 |
| max_lobur | there's also a note in etherpad - Fix order of in test_conductor_utils | 21:54 |
| devananda | NobodyCam: trivial +A https://review.openstack.org/#/c/70382/ | 21:54 |
| devananda | so i am going to need to head out fairly soon, and will probably be gone for the rest of the day | 21:55 |
| max_lobur | i thought we came up that those patch which landed recently fixed only API tests, I don't see why we should fix test_conductor_utils as follow-up. Maybe to fix all the remaining assertEqual stuff? though there will be a lo | 21:55 |
| devananda | if ya'll keep hitting the py26 issue, i've opened a bug -- https://bugs.launchpad.net/ironic/+bug/1279992 | 21:55 |
| max_lobur | * a lot | 21:55 |
| devananda | max_lobur: right. it's not a required follow-up. but if folks want to do clean up work, it's there | 21:56 |
| devananda | just a wishlist item | 21:56 |
| max_lobur | ah, ok | 21:56 |
| max_lobur | makes sense | 21:56 |
| max_lobur | Is something still needs to be investigated/fixed? | 22:03 |
| max_lobur | from our today's patches | 22:03 |
| devananda | we have 3 patches that were +A'd but didnt land | 22:04 |
| devananda | because of bug 1279992 | 22:04 |
| devananda | ** 2 patches, | 22:05 |
| devananda | https://review.openstack.org/#/c/73248/ | 22:05 |
| devananda | https://review.openstack.org/#/c/67415/ | 22:05 |
| devananda | this one also failing, but wasn't approved: https://review.openstack.org/#/c/66621/ | 22:05 |
| *** jdob has quit IRC | 22:05 | |
| max_lobur | ack | 22:07 |
| max_lobur | any Ideas about DriverNotFound: Failed to load driver fake_pxe. | 22:07 |
| max_lobur | ? | 22:07 |
| devananda | nope | 22:08 |
| max_lobur | strange that it's only 2.6 | 22:08 |
| devananda | well, one | 22:08 |
| devananda | race condition between greenthreads | 22:09 |
| devananda | some other test not cleaning up | 22:09 |
| devananda | just an idea -- and probably wrong | 22:09 |
| max_lobur | hmm | 22:09 |
| devananda | oddly, all the failures so far are in the same test | 22:09 |
| devananda | test__update_neutron_fails | 22:09 |
| max_lobur | related to recent merges? | 22:09 |
| NobodyCam | well we just landed the update neutron [atch | 22:10 |
| max_lobur | so we probably introduced change that made the test unstable | 22:10 |
| max_lobur | it's passed its gate | 22:10 |
| max_lobur | but failing now | 22:10 |
| devananda | failing only some of the time | 22:11 |
| NobodyCam | https://review.openstack.org/#/c/70468/ | 22:11 |
| NobodyCam | line 338 of test_pxe | 22:11 |
| NobodyCam | added that test | 22:11 |
| max_lobur | but | 22:12 |
| max_lobur | Failed to load driver fake_pxe. | 22:12 |
| max_lobur | why do you think it's race.. | 22:12 |
| max_lobur | I'd expect NodeAlreadyLocked from race | 22:13 |
| *** matty_dubs is now known as matty_dubs|gone | 22:15 | |
| devananda | PXEPrivateMethodsTestCase does not start a conductorManager | 22:15 |
| devananda | it's loading drivers directly via TaskManager | 22:16 |
| devananda | there's a singleton for the driver_factory | 22:16 |
| devananda | so my *guess* is a race somehow in driver_factory | 22:16 |
| devananda | but like i said, i haven't dug into it -- jsut guessing | 22:16 |
| *** anniec has joined #openstack-ironic | 22:17 | |
| devananda | max_lobur: oh -- we also landed your greenthread patch, which refactored task_manager | 22:25 |
| max_lobur | yes | 22:25 |
| devananda | max_lobur: i have a nagging suspicion __exit__ isn't working the way you think | 22:25 |
| devananda | i recall (but couldn't find on google) that contextlib was necessary to avoid a race with __exit__ not cleaning up | 22:26 |
| max_lobur | not called on exception? | 22:26 |
| max_lobur | I remember I tested that before removing context lib, though i'm not sure i covered everything | 22:26 |
| max_lobur | yep, basically there was a try/finally block, that might change situation | 22:27 |
| max_lobur | ah, of course, i tried that on 2.7. probably __exit__ behaving different on 2.6 | 22:28 |
| *** jbjohnso has quit IRC | 22:28 | |
| max_lobur | can you reproduce that 2.6 fail on local env? (mine is not complete, i can try only tomorrow on office desktop) | 22:30 |
| devananda | i need to spin up a 2.6 as well :( | 22:30 |
| devananda | max_lobur: ahhh | 22:31 |
| devananda | so here is a problem | 22:31 |
| devananda | ugh, i should have seen this | 22:31 |
| devananda | old task_manager.acquire | 22:32 |
| devananda | when passed multiple nodes | 22:32 |
| devananda | will acquire resources for al lof them | 22:32 |
| devananda | 124 for id in node_ids: | 22:32 |
| devananda | 125 t.resources.append(resource_manager.NodeManager.acquire( | 22:32 |
| devananda | if it fails part way through, finally: clause kicks in | 22:32 |
| devananda | and it releases them | 22:32 |
| devananda | !!! | 22:32 |
| openstack | devananda: Error: "!!" is not a valid command. | 22:32 |
| max_lobur | true | 22:33 |
| max_lobur | that's bad | 22:33 |
| devananda | max_lobur: are we sure that, on exception inside of acquire, it's going to call __exit__ | 22:33 |
| max_lobur | I completely missed that | 22:33 |
| devananda | it never actually enterrred | 22:33 |
| devananda | I remembers looking at an issue like this and decided to use contextlib with a generator when coding this almost a year ago :) | 22:34 |
| *** anniec has left #openstack-ironic | 22:35 | |
| devananda | ok. unfortunately i need to run to an appt | 22:35 |
| max_lobur | :) | 22:36 |
| max_lobur | thought it may be fixed without context lib | 22:36 |
| devananda | indeed | 22:36 |
| devananda | i would expect there's more than one way to solve it | 22:36 |
| max_lobur | by wrapping in try/except in acquire shortcut | 22:36 |
| max_lobur | here https://github.com/openstack/ironic/blob/master/ironic/conductor/task_manager.py#L131-L133 | 22:37 |
| max_lobur | makes sense? | 22:37 |
| max_lobur | I'll make a patch | 22:37 |
| devananda | max_lobur: what about moving functionality that is inside TaskManager.acquire_resources into TaskManager.__init__ | 22:37 |
| devananda | and adding try: except: | 22:38 |
| devananda | i think your approach would work but does nto address TaskManager.acquire_resources | 22:38 |
| devananda | which is a public method and may be called directly | 22:38 |
| devananda | ok - i'm really late now .... | 22:39 |
| max_lobur | right | 22:39 |
| devananda | :) ttyl! | 22:39 |
| NobodyCam | night max_lobur | 22:39 |
| max_lobur | :D | 22:40 |
| max_lobur | I said right? :D | 22:40 |
| NobodyCam | oh see I thought you said you were late :-p | 22:41 |
| max_lobur | :) | 22:41 |
| NobodyCam | i have told you I need glasses havent I? | 22:41 |
| max_lobur | hehe | 22:42 |
| max_lobur | I liked you monocle | 22:42 |
| NobodyCam | lol | 22:42 |
| max_lobur | wear it :D | 22:42 |
| NobodyCam | sad part I REALLY do need them.. just haven't gone to get them | 22:43 |
| max_lobur | last time I'm feeling like I need glasses as well.. | 22:44 |
| max_lobur | need to spend more time outside | 22:44 |
| NobodyCam | ya :) | 22:44 |
| *** med_ has quit IRC | 22:52 | |
| *** med_ has joined #openstack-ironic | 22:52 | |
| *** thedodd has quit IRC | 22:56 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Fix task_manager acquire post landing for c4f2f26ed https://review.openstack.org/73440 | 22:58 |
| *** mdurnosvistov_ has quit IRC | 22:59 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Fix task_manager acquire post landing for c4f2f26ed https://review.openstack.org/73440 | 23:01 |
| NobodyCam | lol I just opened the first one | 23:01 |
| max_lobur | Hope this will help | 23:02 |
| max_lobur | that was the con of quick review :) | 23:03 |
| max_lobur | I hope the last one for today | 23:03 |
| NobodyCam | yep, but we knew that going in | 23:03 |
| NobodyCam | need to do a quick walkies.. brb | 23:04 |
| NobodyCam | max_lobur: py27 failing on ^^^ | 23:10 |
| max_lobur | NobodyCam: looking | 23:12 |
| max_lobur | I don't have full-func local env now so have to check code through jenkins | 23:12 |
| max_lobur | ahh | 23:13 |
| max_lobur | tests | 23:13 |
| max_lobur | right | 23:13 |
| max_lobur | thanks mr jenkins :) | 23:13 |
| NobodyCam | __init__() takes at least 3 arguments (2 given) | 23:13 |
| NobodyCam | :-p | 23:13 |
| max_lobur | yup | 23:13 |
| *** ndipanov has joined #openstack-ironic | 23:16 | |
| openstackgerrit | A change was merged to openstack/ironic: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/70382 | 23:24 |
| *** ndipanov has quit IRC | 23:29 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Fix task_manager acquire post landing for c4f2f26ed https://review.openstack.org/73440 | 23:29 |
| *** anniec has joined #openstack-ironic | 23:31 | |
| NobodyCam | success with py27 | 23:33 |
| NobodyCam | :) | 23:33 |
| max_lobur | :) | 23:37 |
| max_lobur | 26 more interesting | 23:37 |
| NobodyCam | 26 running now | 23:39 |
| max_lobur | I'm watching as well :) | 23:39 |
| max_lobur | yey! | 23:41 |
| max_lobur | seems that was the problem | 23:42 |
| max_lobur | k, I'm going sleep | 23:46 |
| max_lobur | night NobodyCam | 23:46 |
| max_lobur | ah, tempest started | 23:47 |
| max_lobur | can't sleep without results | 23:47 |
| NobodyCam | hehehe | 23:53 |
| NobodyCam | another 20 minutes | 23:54 |
| *** derekh_afk is now known as derekh | 23:56 | |
| max_lobur | :) | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!