| *** zenirc369 has joined #openstack-meeting-5 | 00:09 | |
| *** zenirc369 has quit IRC | 00:23 | |
| *** eanylin has joined #openstack-meeting-5 | 00:26 | |
| *** notmyname has quit IRC | 02:08 | |
| *** notmyname has joined #openstack-meeting-5 | 02:13 | |
| *** ricolin has joined #openstack-meeting-5 | 02:19 | |
| *** benj_ has quit IRC | 04:03 | |
| *** benj_ has joined #openstack-meeting-5 | 04:04 | |
| *** markvoelker has joined #openstack-meeting-5 | 04:13 | |
| *** zenirc369 has joined #openstack-meeting-5 | 04:44 | |
| *** markvoelker_ has joined #openstack-meeting-5 | 05:11 | |
| *** markvoelker has quit IRC | 05:14 | |
| *** markvoelker has joined #openstack-meeting-5 | 05:17 | |
| *** markvoel_ has joined #openstack-meeting-5 | 05:19 | |
| *** markvoelker has quit IRC | 05:19 | |
| *** markvoelker_ has quit IRC | 05:21 | |
| *** korzen has joined #openstack-meeting-5 | 05:52 | |
| *** markvoel_ has quit IRC | 05:53 | |
| *** markvoelker has joined #openstack-meeting-5 | 05:53 | |
| *** markvoelker has quit IRC | 05:54 | |
| *** markvoelker has joined #openstack-meeting-5 | 06:27 | |
| *** markvoelker has quit IRC | 06:32 | |
| *** ralonsoh has joined #openstack-meeting-5 | 06:55 | |
| *** markvoelker has joined #openstack-meeting-5 | 06:59 | |
| *** markvoelker has quit IRC | 07:03 | |
| *** markvoelker has joined #openstack-meeting-5 | 07:09 | |
| *** dixiaoli has joined #openstack-meeting-5 | 07:13 | |
| *** markvoelker has quit IRC | 07:14 | |
| *** markvoelker has joined #openstack-meeting-5 | 07:29 | |
| *** markvoelker has quit IRC | 07:37 | |
| *** markvoelker has joined #openstack-meeting-5 | 07:42 | |
| *** markvoelker has quit IRC | 07:49 | |
| *** LanceHaig has quit IRC | 08:06 | |
| *** LanceHaig has joined #openstack-meeting-5 | 08:15 | |
| *** LanceHaig has joined #openstack-meeting-5 | 08:15 | |
| *** markvoelker has joined #openstack-meeting-5 | 08:24 | |
| *** markvoelker has quit IRC | 08:30 | |
| *** markvoelker has joined #openstack-meeting-5 | 08:32 | |
| *** markvoelker has quit IRC | 08:40 | |
| *** derekh has joined #openstack-meeting-5 | 09:09 | |
| *** markvoelker has joined #openstack-meeting-5 | 09:16 | |
| *** markvoelker has quit IRC | 09:21 | |
| *** markvoelker has joined #openstack-meeting-5 | 09:54 | |
| *** markvoelker has quit IRC | 09:58 | |
| *** dixiaoli has quit IRC | 10:39 | |
| *** markvoelker has joined #openstack-meeting-5 | 10:47 | |
| *** markvoelker has quit IRC | 10:52 | |
| *** markvoelker has joined #openstack-meeting-5 | 10:55 | |
| *** markvoelker has quit IRC | 10:59 | |
| *** ricolin has quit IRC | 11:02 | |
| *** ricolin has joined #openstack-meeting-5 | 11:03 | |
| *** markvoelker has joined #openstack-meeting-5 | 11:14 | |
| *** markvoelker has quit IRC | 11:19 | |
| *** markvoelker has joined #openstack-meeting-5 | 11:29 | |
| *** markvoelker has quit IRC | 11:34 | |
| *** markvoelker has joined #openstack-meeting-5 | 12:00 | |
| *** markvoelker has quit IRC | 12:04 | |
| *** markvoelker has joined #openstack-meeting-5 | 12:11 | |
| *** markvoelker has quit IRC | 12:15 | |
| *** markvoelker has joined #openstack-meeting-5 | 12:17 | |
| *** markvoelker has quit IRC | 12:22 | |
| *** markvoelker has joined #openstack-meeting-5 | 12:56 | |
| *** korzen has quit IRC | 13:13 | |
| *** dixiaoli has joined #openstack-meeting-5 | 13:26 | |
| *** dixiaoli has quit IRC | 13:28 | |
| *** markvoelker has quit IRC | 13:33 | |
| *** eanylin has quit IRC | 13:37 | |
| *** eanylin has joined #openstack-meeting-5 | 13:51 | |
| *** markvoelker has joined #openstack-meeting-5 | 14:15 | |
| *** zenirc369 has quit IRC | 14:21 | |
| *** markvoelker has quit IRC | 14:23 | |
| *** felipemonteiro has joined #openstack-meeting-5 | 14:30 | |
| *** korzen has joined #openstack-meeting-5 | 14:52 | |
| *** markvoelker has joined #openstack-meeting-5 | 15:01 | |
| *** korzen has quit IRC | 15:02 | |
| *** marst_ has joined #openstack-meeting-5 | 15:05 | |
| *** derekh has quit IRC | 15:10 | |
| *** felipemonteiro has quit IRC | 15:14 | |
| *** hshiina has joined #openstack-meeting-5 | 15:48 | |
| *** tiendc has joined #openstack-meeting-5 | 15:52 | |
| *** phuongnh has joined #openstack-meeting-5 | 15:56 | |
| TheJulia | o/ | 16:00 |
|---|---|---|
| TheJulia | #startmeeting ironic_bfv | 16:01 |
| openstack | Meeting started Thu May 25 16:01:00 2017 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:01 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:01 |
| *** openstack changes topic to " (Meeting topic: ironic_bfv)" | 16:01 | |
| openstack | The meeting name has been set to 'ironic_bfv' | 16:01 |
| mjturek | o/ | 16:01 |
| hshiina | o/ | 16:01 |
| TheJulia | Greetings everyone! | 16:01 |
| mjturek | hey TheJulia :) | 16:01 |
| TheJulia | As always, our agenda for today is on our wiki page. | 16:01 |
| TheJulia | #link https://wiki.openstack.org/wiki/Meetings/Ironic-BFV#Agenda | 16:01 |
| TheJulia | #topic Announcements/Reminders | 16:02 |
| *** openstack changes topic to "Announcements/Reminders (Meeting topic: ironic_bfv)" | 16:02 | |
| *** dtantsur has joined #openstack-meeting-5 | 16:02 | |
| dtantsur | o/ | 16:02 |
| TheJulia | I have one announcement/ask today from myself, I'll potentially be in meetings on Thursday and wonder if someone would be willing to run this meeting next week if I'm unavailable. | 16:02 |
| TheJulia | o/ dtantsur | 16:03 |
| mjturek | I can run it! | 16:03 |
| TheJulia | mjturek: Awesome | 16:03 |
| TheJulia | #info mjturek has volunteered to run the next meeting if TheJulia is unavailable. | 16:03 |
| TheJulia | Anyone have anything else? | 16:03 |
| dtantsur | thanks mjturek! | 16:03 |
| mjturek | nppp, I'll ping you TheJulia before to confirm | 16:04 |
| TheJulia | mjturek: excellent, Thanks! | 16:04 |
| TheJulia | So if nobody has anything else for annoucements, we can move on | 16:04 |
| TheJulia | #topic Current Status | 16:05 |
| *** openstack changes topic to "Current Status (Meeting topic: ironic_bfv)" | 16:05 | |
| TheJulia | #link https://etherpad.openstack.org/p/Ironic-BFV | 16:05 |
| TheJulia | It looks like we managed to get a few patches landed this week \o/ | 16:05 |
| mjturek | so cinder driver has merged - thanks dtantsur and sambetts for the reviews | 16:05 |
| dtantsur | np :) | 16:06 |
| TheJulia | hshiina: Regarding the status ether pad, I noticed you updated a patch or two. When I went through the revisions about an hour ago, I noted their current state, so the ether pad may be slightly out of date for those revisions. | 16:07 |
| hshiina | sorry, i started rebasing a few patches. | 16:08 |
| TheJulia | mjturek: you indicated you only rebased and put a change in on the wire-in patch, I think I have some minor changes locally for that one as well, and I'll have to reconcile that and push it up | 16:08 |
| TheJulia | hshiina: no apologies necessary :) | 16:08 |
| mjturek | cool, yep I added 'cinder' to the default drivers list (and updated the commit message to indicate that). Tests were failing | 16:09 |
| TheJulia | And mjturek has a few discussion items for the ipxe patch, so I guess we should move along. :) | 16:09 |
| TheJulia | Awesome | 16:09 |
| TheJulia | Time to discuss priorities? | 16:10 |
| mjturek | oh - I uploaded a follow up patch for the cinder driver as well | 16:10 |
| TheJulia | I saw, thank you | 16:10 |
| mjturek | take a look, but I don't like that the detach_volumes signature is different than the base class :-\ | 16:10 |
| TheJulia | #topic Planning/Priorities | 16:11 |
| *** openstack changes topic to "Planning/Priorities (Meeting topic: ironic_bfv)" | 16:11 | |
| TheJulia | Priorities as I see them is the ipxe patch, the wire-in patch which I should be able to reconcile and get updated quickly, and then the deploy skip patch. The last of which may be impacted by one of the discussion items. | 16:11 |
| mjturek | sounds good to me] | 16:12 |
| hshiina | sounds good to me, too | 16:12 |
| TheJulia | Sounds like we are in agreement. | 16:13 |
| dtantsur | ++ | 16:13 |
| TheJulia | And with that, if nobody has anything else, then we should move on to discussion :) | 16:13 |
| TheJulia | #topic Discussion | 16:14 |
| *** openstack changes topic to "Discussion (Meeting topic: ironic_bfv)" | 16:14 | |
| * TheJulia hands the microphone to mjturek | 16:14 | |
| mjturek | So first item is on the wirein patch | 16:14 |
| mjturek | dtantsur mentioned that we should probably move the storage ops up from the driver level | 16:14 |
| mjturek | https://review.openstack.org/#/c/413324/20/ironic/drivers/modules/pxe.py@291 | 16:15 |
| dtantsur | yeah | 16:15 |
| mjturek | seems like the right approach as the code will need to be in each deploy module the current way | 16:15 |
| dtantsur | otherwise we're doomed to copy-paste it | 16:15 |
| mjturek | :) | 16:15 |
| dtantsur | like with networking, which is something vdrok wants to address | 16:15 |
| mjturek | I can volunteer to look into it as a followup, unless you'd like to address it in the patch TheJulia | 16:16 |
| mjturek | (or if there's a disagreement here - maybe there's a good reason for it being in the driver level?) | 16:17 |
| dtantsur | I'm fine with a follow-up, as long as it happens rather sooner than later (e.g. as the next patch in the chain) | 16:17 |
| TheJulia | So move any helpers into centralized helpers, and move the wire-in calls into the conductor? | 16:17 |
| TheJulia | Have we identified where we would call the networking in the conductor? | 16:18 |
| dtantsur | networking is a separate thing, I guess. or do we want to solve both at once? | 16:18 |
| * TheJulia feels like it would be best as a follow-up for the conductor part | 16:18 | |
| dtantsur | I'm just suggesting that we don't place volume actions in the deploy drivers (or move it away soon) | 16:18 |
| TheJulia | I feel like we are going to step on each other's feet if we don't have a basic agreement of how in place before we try and do it | 16:18 |
| mjturek | yeah I'd think we'd want to do it around the same spot as the networking | 16:19 |
| mjturek | maybe vdrok has a plan there already, I can ask him | 16:19 |
| dtantsur | he has a spec | 16:19 |
| mjturek | oh cool | 16:19 |
| dtantsur | #link https://review.openstack.org/#/c/468000/ | 16:19 |
| * dtantsur is not sure he can use #link here | 16:19 | |
| TheJulia | I guess, there're only two logic places for any of it to really go in the conductor | 16:19 |
| TheJulia | #link https://review.openstack.org/#/c/468000/ | 16:20 |
| TheJulia | I think anyone can, but we can look at the logs later to be sure :) | 16:20 |
| TheJulia | I feel like the resolution is talk to vdrok, get on the same page, and determine the happy path from there. | 16:21 |
| mjturek | sounds good | 16:21 |
| TheJulia | and well, read the spec too | 16:21 |
| mjturek | :) | 16:21 |
| TheJulia | So, your next item mjturek ? | 16:22 |
| mjturek | sure | 16:22 |
| mjturek | #link https://review.openstack.org/#/c/406290/25/ironic/drivers/modules/agent.py@396 | 16:22 |
| mjturek | whoops | 16:22 |
| mjturek | I swapped the links by accident :) | 16:23 |
| mjturek | https://review.openstack.org/#/c/413324/20/ironic/drivers/modules/pxe.py@291 | 16:23 |
| mjturek | this is the one wrt sanhook | 16:23 |
| TheJulia | I just figured that out too, but the other item does kind of fit with-in the same topic in a way | 16:23 |
| TheJulia | in part | 16:23 |
| mjturek | yep true! | 16:23 |
| TheJulia | Anyway! | 16:23 |
| mjturek | so I think Anton is correct about the separate sanhooks but I was wondering if people had more insight | 16:24 |
| mjturek | basically, for multipath the ipxe documentation shows a single call for both paths | 16:24 |
| mjturek | while we're doing one sanhook call for each | 16:24 |
| mjturek | should we be anticipating multipath here and doing the one call instead? | 16:25 |
| dtantsur | are there any reasons not to? | 16:25 |
| mjturek | honestly I'm not sure. I would think there aren't any reasons not to, but not sure if it's worth the effort or not | 16:26 |
| TheJulia | So sanhook should be explicitly executed for each "device", a single device could be defined twice which would represent another path possible, but the OS has to handle that and all | 16:26 |
| TheJulia | https://review.openstack.org/#/c/413324/20/ironic/drivers/modules/ipxe_config.template <-- does sanhooks and a sanboot | 16:27 |
| TheJulia | mjturek: last I looked it didn't reference multiple paths | 16:27 |
| TheJulia | dtantsur: no reason not to, we should pass it if we can | 16:27 |
| dtantsur | ack | 16:28 |
| mjturek | hmm okay | 16:30 |
| TheJulia | mjturek: so I'm not sure we have a good way of de-duplicating all of the possible paths. Conceptually for us, they would be unique targets | 16:30 |
| mjturek | okay so basically the OS will handle it | 16:30 |
| TheJulia | the OS is the only thing that can truly consolidate that back together by looking at data it gets via the connection and if it realizes it is a duplicate that is locked, or if it is a unique device | 16:31 |
| TheJulia | yeah | 16:31 |
| * TheJulia gets out the "There be dragons here sign" | 16:31 | |
| mjturek | fair enough :) | 16:31 |
| * TheJulia makes a nice concrete foundation for the sign, so it stays around. | 16:31 | |
| mjturek | hahaha, okay I'm cool on this then. I'll reply to Anton | 16:31 |
| TheJulia | I'm feeling like ipxe also just may have added multipath iscsi in their docs.... | 16:32 |
| mjturek | as in, they added it recently? | 16:32 |
| TheJulia | yes | 16:32 |
| mjturek | interesting | 16:33 |
| * dtantsur prepares his usual complain about ipxe versioning and releases | 16:33 | |
| * TheJulia gives dtantsur the ipxe soap box | 16:33 | |
| dtantsur | :D | 16:33 |
| mjturek | last modified 2017/03/26 so quite possible :-| | 16:33 |
| TheJulia | Yeah, I think I wrote that code late last year, so yeah | 16:34 |
| TheJulia | :( | 16:34 |
| TheJulia | Anyway, we're in a catch-22 there, but the OS should be responsible for MPIO path management | 16:34 |
| dtantsur | I hope we won't start depending on the iPXE version from this march :) | 16:34 |
| TheJulia | The device should get locked anyway along one path | 16:34 |
| mjturek | yeaaaah, sounds reasonable to me | 16:34 |
| mjturek | :) | 16:34 |
| TheJulia | dtantsur: +^one_million | 16:35 |
| TheJulia | If the device is not locked, then the OS is either doing something stupid, or the storage is doing something stupid, and chaos can then ensue regardless because it would have on its own anyway. | 16:36 |
| TheJulia | (as in, in no way shape or form mpio safe) | 16:36 |
| TheJulia | I _think_ we're good | 16:36 |
| TheJulia | Anything else? | 16:37 |
| mjturek | I'm good | 16:38 |
| dtantsur | ditto | 16:38 |
| TheJulia | #topic Open Discussion | 16:38 |
| *** openstack changes topic to "Open Discussion (Meeting topic: ironic_bfv)" | 16:38 | |
| TheJulia | Anyone have anything for open discussion? | 16:38 |
| hshiina | i noticed cinder doesn't accept metadata with dict | 16:39 |
| hshiina | when ironic sets metadata in attach/detach | 16:39 |
| hshiina | we need to change metadata not to use dict | 16:40 |
| TheJulia | to just be a string? | 16:40 |
| mjturek | hshiina: what should it be using? | 16:40 |
| hshiina | TheJulia, yes | 16:40 |
| dtantsur | hshiina: I think you mean a non-nested dict, no? /me tries to remember the patch | 16:41 |
| hshiina | dtantsur, yes, non-nested | 16:41 |
| mjturek | hshiina: https://review.openstack.org/#/c/413324/20/ironic/drivers/modules/pxe.py@291 referring to connector? | 16:41 |
| TheJulia | mjturek: I think he is referring to the metadata post to the cinder api upon attachment | 16:42 |
| TheJulia | of the extra metadata information, which I thought was listed as a dictionary :( | 16:43 |
| TheJulia | Easy enough to look it up and double check cinder's api code | 16:43 |
| mjturek | sorry wrong link | 16:43 |
| mjturek | https://review.openstack.org/#/c/366197/47/ironic/drivers/modules/storage/cinder.py@327 | 16:43 |
| hshiina | ironic creates metadata here: https://github.com/openstack/ironic/blob/767ed8002045defa0a3e982c22a0c4856796df7e/ironic/common/cinder.py#L425 | 16:45 |
| mjturek | ohhh got it | 16:45 |
| mjturek | thanks | 16:45 |
| TheJulia | http://git.openstack.org/cgit/openstack/ironic/tree/ironic/common/cinder.py#n312 | 16:46 |
| TheJulia | https://developer.openstack.org/api-ref/block-storage/v3/index.html?expanded=create-metadata-for-volume-detail#create-metadata-for-volume | 16:47 |
| TheJulia | so string of JSON I guess.... | 16:47 |
| *** skazi has quit IRC | 16:48 | |
| TheJulia | hshiina: Could you create a patch to fix it for your testing? | 16:50 |
| TheJulia | It seems super simple, since we have the utilities to do it | 16:50 |
| hshiina | TheJulia, i created an patch. https://review.openstack.org/#/c/467930/ | 16:51 |
| hshiina | This issue is not critical. warning is always logged. | 16:52 |
| hshiina | i reported it to cinder: https://bugs.launchpad.net/cinder/+bug/1690363 | 16:53 |
| openstack | Launchpad bug 1690363 in Cinder "Nested metadata to volume causes an internal server error" [Undecided,Fix released] - Assigned to luqitao (qtlu) | 16:53 |
| hshiina | only response code was fixed. cinder API doesn't accept nested metadata. | 16:53 |
| TheJulia | okay | 16:53 |
| TheJulia | That makes sense then I guess | 16:54 |
| TheJulia | Thank you! | 16:54 |
| mjturek | cool | 16:54 |
| TheJulia | Well if there is nothing else, we have six minutes left... | 16:54 |
| TheJulia | Well, I'm taking the silence as it is time to call the meeting to an end | 16:55 |
| TheJulia | Thank you everyone! | 16:55 |
| mjturek | thanks all! | 16:55 |
| mjturek | ttyl | 16:55 |
| hshiina | thank you | 16:55 |
| TheJulia | #endmeeting | 16:55 |
| *** openstack changes topic to "OpenStack Meetings || https://wiki.openstack.org/wiki/Meetings" | 16:55 | |
| openstack | Meeting ended Thu May 25 16:55:46 2017 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 16:55 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/ironic_bfv/2017/ironic_bfv.2017-05-25-16.01.html | 16:55 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/ironic_bfv/2017/ironic_bfv.2017-05-25-16.01.txt | 16:55 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/ironic_bfv/2017/ironic_bfv.2017-05-25-16.01.log.html | 16:55 |
| *** dtantsur has left #openstack-meeting-5 | 16:55 | |
| *** phuongnh has left #openstack-meeting-5 | 16:56 | |
| *** felipemonteiro has joined #openstack-meeting-5 | 16:58 | |
| *** tiendc has quit IRC | 16:59 | |
| *** ralonsoh has quit IRC | 17:01 | |
| *** hshiina has left #openstack-meeting-5 | 17:03 | |
| *** ricolin has quit IRC | 17:10 | |
| *** korzen has joined #openstack-meeting-5 | 17:22 | |
| *** korzen has quit IRC | 18:07 | |
| *** korzen has joined #openstack-meeting-5 | 18:23 | |
| *** felipemonteiro has quit IRC | 20:10 | |
| *** korzen has quit IRC | 20:13 | |
| *** markvoelker has quit IRC | 20:42 | |
| *** markvoelker has joined #openstack-meeting-5 | 20:56 | |
| *** markvoelker has quit IRC | 21:00 | |
| *** marst_ has quit IRC | 21:30 | |
| *** bzhao_ has joined #openstack-meeting-5 | 21:59 | |
| *** markvoelker has joined #openstack-meeting-5 | 22:05 | |
| *** markvoelker has quit IRC | 22:15 | |
| *** markvoelker has joined #openstack-meeting-5 | 22:16 | |
| *** markvoelker has quit IRC | 23:11 | |
| *** markvoelker has joined #openstack-meeting-5 | 23:44 | |
| *** markvoelker has quit IRC | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!