jroll | feels like neutron should handle the differences between dhcp servers, imo | 00:00 |
---|---|---|
Sukhdev | jroll: I can bring it at ptg with neutron team. In the meantime, if you can think of simplest hack I put to move forward, I would really appreciate it | 00:06 |
jroll | Sukhdev: simplest hack would be use dnsmasq imo :) | 00:06 |
* jroll knows nothing about how to do the equivalent of 'tag:!ipxe,bootfile-name' on juniper's stuff | 00:07 | |
TheJulia | Sukhdev: The patch likely won't help you, it is for... more so... static dhcp configurations, not neutron integrated | 00:07 |
*** rob0t_ has joined #openstack-ironic | 00:08 | |
Sukhdev | jroll: I do not have choice in the backend. So, a hack into the ironic side? | 00:08 |
Sukhdev | TheJulia: :-) | 00:08 |
TheJulia | Sukhdev: I think that is what we're suggesting | 00:08 |
TheJulia | At a glance, I'm not sure we can do the logic that the tags provide us.... which makes me think we would have to be able to explicitly state if it is ipxe, not ipxe. i.e. offer no upgrade capability | 00:09 |
TheJulia | :( | 00:09 |
*** rob0t_ has quit IRC | 00:12 | |
Sukhdev | H | 00:12 |
Sukhdev | Looks like I kinda stuck for now | 00:13 |
*** chedda has joined #openstack-ironic | 00:14 | |
TheJulia | Sukhdev: first step would be to disable ipxe | 00:14 |
Sukhdev | TheJulia: how to do that | 00:16 |
*** rloo has quit IRC | 00:16 | |
*** rob0t_ has joined #openstack-ironic | 00:20 | |
TheJulia | Sukhdev: IRONIC_IPXE_ENABLED=false in your local.cf | 00:20 |
Sukhdev | TheJulia: got it... Thanks | 00:21 |
*** rob0t_ has quit IRC | 00:24 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic master: WIP/DNM: Testing multinode job https://review.openstack.org/422149 | 00:28 |
*** baoli has joined #openstack-ironic | 00:35 | |
*** e0ne has quit IRC | 00:46 | |
*** jaosorior has quit IRC | 00:46 | |
*** hw_wutianwei has joined #openstack-ironic | 00:51 | |
*** rama_y has quit IRC | 01:04 | |
*** crushil has joined #openstack-ironic | 01:20 | |
*** wrouesnel has joined #openstack-ironic | 01:25 | |
*** sdake has joined #openstack-ironic | 01:28 | |
*** owalsh has joined #openstack-ironic | 01:28 | |
*** owalsh-afk has quit IRC | 01:28 | |
*** gcb has joined #openstack-ironic | 01:31 | |
*** baoli has quit IRC | 01:32 | |
*** amotoki has joined #openstack-ironic | 01:33 | |
*** rob0t_ has joined #openstack-ironic | 01:36 | |
openstackgerrit | Annie Lezil proposed openstack/ironic-python-agent master: Collect NIC name given by BIOS https://review.openstack.org/396406 | 01:43 |
*** baoli has joined #openstack-ironic | 01:48 | |
*** aNuposic has quit IRC | 01:52 | |
*** chedda has quit IRC | 01:54 | |
*** anteaya has quit IRC | 01:57 | |
*** anteaya has joined #openstack-ironic | 02:04 | |
*** rob0t_ has quit IRC | 02:07 | |
*** aNuposic has joined #openstack-ironic | 02:09 | |
*** baoli has quit IRC | 02:13 | |
*** rob0t_ has joined #openstack-ironic | 02:14 | |
*** baoli has joined #openstack-ironic | 02:17 | |
*** baoli has quit IRC | 02:18 | |
*** rajinir has quit IRC | 02:26 | |
*** mtanino has quit IRC | 02:28 | |
*** aNuposic has quit IRC | 02:30 | |
*** Syed__ has quit IRC | 02:35 | |
*** adu has joined #openstack-ironic | 02:38 | |
*** adu has quit IRC | 02:42 | |
*** amotoki_ has joined #openstack-ironic | 02:44 | |
*** amotoki has quit IRC | 02:46 | |
*** wrouesnel has quit IRC | 02:56 | |
*** wrouesnel has joined #openstack-ironic | 02:56 | |
*** amotoki_ has quit IRC | 02:57 | |
*** pmannidi has quit IRC | 02:57 | |
*** pmannidi has joined #openstack-ironic | 02:58 | |
*** baoli has joined #openstack-ironic | 02:58 | |
*** wrouesnel has quit IRC | 02:59 | |
*** wrouesnel has joined #openstack-ironic | 02:59 | |
*** wrouesnel has quit IRC | 03:01 | |
*** wrouesnel has joined #openstack-ironic | 03:01 | |
*** wrouesnel has quit IRC | 03:04 | |
*** wrouesnel has joined #openstack-ironic | 03:04 | |
*** crushil has quit IRC | 03:05 | |
*** wrouesnel has quit IRC | 03:06 | |
*** wrouesnel has joined #openstack-ironic | 03:06 | |
*** sayantan_ has quit IRC | 03:11 | |
*** ijw has quit IRC | 03:14 | |
*** wrouesnel has quit IRC | 03:15 | |
*** wrouesnel has joined #openstack-ironic | 03:15 | |
*** adu has joined #openstack-ironic | 03:18 | |
*** baoli has quit IRC | 03:19 | |
*** wrouesnel has quit IRC | 03:19 | |
*** wrouesnel has joined #openstack-ironic | 03:19 | |
*** sayantani01 has joined #openstack-ironic | 03:20 | |
*** baoli has joined #openstack-ironic | 03:23 | |
*** crushil has joined #openstack-ironic | 03:25 | |
*** owalsh has quit IRC | 03:33 | |
*** owalsh_ has joined #openstack-ironic | 03:33 | |
*** vikrant has joined #openstack-ironic | 03:34 | |
*** baoli has quit IRC | 03:37 | |
*** wrouesnel has quit IRC | 03:37 | |
*** wrouesnel has joined #openstack-ironic | 03:37 | |
*** amotoki has joined #openstack-ironic | 03:38 | |
*** wrouesnel has quit IRC | 03:40 | |
*** wrouesnel has joined #openstack-ironic | 03:40 | |
*** sdake has quit IRC | 03:41 | |
*** crushil has quit IRC | 03:41 | |
*** amotoki has quit IRC | 03:45 | |
*** wrouesnel has quit IRC | 03:53 | |
*** wrouesnel has joined #openstack-ironic | 03:53 | |
*** wrouesnel has quit IRC | 03:59 | |
*** wrouesnel has joined #openstack-ironic | 03:59 | |
*** adu has quit IRC | 04:01 | |
*** owalsh_ is now known as owalsh-afk | 04:01 | |
*** wrouesnel has quit IRC | 04:02 | |
*** wrouesnel has joined #openstack-ironic | 04:02 | |
*** crushil has joined #openstack-ironic | 04:07 | |
*** sdake has joined #openstack-ironic | 04:07 | |
*** wrouesnel has quit IRC | 04:08 | |
*** wrouesnel has joined #openstack-ironic | 04:08 | |
*** sdake_ has joined #openstack-ironic | 04:11 | |
*** nmathew has joined #openstack-ironic | 04:12 | |
*** sdake has quit IRC | 04:12 | |
*** udesale has joined #openstack-ironic | 04:13 | |
*** jaganathan has joined #openstack-ironic | 04:15 | |
*** Sukhdev has quit IRC | 04:21 | |
*** sayantani01 has quit IRC | 04:26 | |
*** crushil has quit IRC | 04:33 | |
*** crushil has joined #openstack-ironic | 04:33 | |
*** crushil has quit IRC | 04:39 | |
*** crushil has joined #openstack-ironic | 04:41 | |
*** ijw has joined #openstack-ironic | 04:43 | |
*** adu has joined #openstack-ironic | 04:47 | |
*** jvgrant has quit IRC | 04:48 | |
openstackgerrit | Aparna proposed openstack/proliantutils master: Add HPSUM firmware update support https://review.openstack.org/422004 | 04:55 |
*** jvgrant has joined #openstack-ironic | 04:57 | |
*** Nisha_Agarwal has joined #openstack-ironic | 05:06 | |
openstackgerrit | Madhuri Kumari proposed openstack/ironic master: Fix directories permission https://review.openstack.org/427077 | 05:09 |
*** adu has quit IRC | 05:14 | |
*** crushil has quit IRC | 05:16 | |
*** crushil has joined #openstack-ironic | 05:16 | |
*** ijw has quit IRC | 05:17 | |
*** aparnav has joined #openstack-ironic | 05:29 | |
*** sayantani01 has joined #openstack-ironic | 05:33 | |
*** Sukhdev has joined #openstack-ironic | 05:39 | |
*** Nisha_Agarwal has quit IRC | 05:44 | |
*** sasha2 has quit IRC | 05:50 | |
*** wrouesnel has quit IRC | 05:56 | |
*** wrouesnel has joined #openstack-ironic | 05:57 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic master: WIP/DNM: Testing multinode job https://review.openstack.org/422149 | 06:09 |
*** MattMan_1 has quit IRC | 06:12 | |
*** MattMan_1 has joined #openstack-ironic | 06:12 | |
*** crushil has quit IRC | 06:13 | |
*** rbudden has quit IRC | 06:14 | |
*** vsaienko has joined #openstack-ironic | 06:15 | |
*** wrouesnel has quit IRC | 06:16 | |
*** wrouesnel has joined #openstack-ironic | 06:16 | |
*** vsaienko has quit IRC | 06:17 | |
*** wrouesnel has quit IRC | 06:24 | |
*** wrouesnel has joined #openstack-ironic | 06:24 | |
*** mjura has joined #openstack-ironic | 06:32 | |
*** Sukhdev has quit IRC | 06:43 | |
openstackgerrit | Aparna proposed openstack/ironic master: Support `update_firmware` to hpsum based firmware update https://review.openstack.org/422572 | 06:44 |
*** sdake_ has quit IRC | 06:52 | |
*** vikrant has quit IRC | 06:53 | |
*** sayantani01 has quit IRC | 07:07 | |
*** sayantani01 has joined #openstack-ironic | 07:07 | |
*** tesseract has joined #openstack-ironic | 07:20 | |
*** rcernin has joined #openstack-ironic | 07:26 | |
*** athomas has joined #openstack-ironic | 07:28 | |
openstackgerrit | Dao Cong Tien proposed openstack/python-ironicclient master: Extends driver-list, driver-show supporting new hardware types https://review.openstack.org/419274 | 07:33 |
*** ChubYann has quit IRC | 07:44 | |
*** sayantani01 has quit IRC | 07:48 | |
*** sayantani01 has joined #openstack-ironic | 07:48 | |
*** vsaienko has joined #openstack-ironic | 07:51 | |
*** ralonsoh has joined #openstack-ironic | 08:13 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic master: TMP: Verify multinode nova fix https://review.openstack.org/414160 | 08:17 |
*** sayantani01 has quit IRC | 08:33 | |
vdrok | morning ironic! | 08:35 |
*** jaganathan has quit IRC | 08:38 | |
*** fxpester has joined #openstack-ironic | 08:55 | |
*** jaganathan has joined #openstack-ironic | 08:55 | |
*** rob0t_ has quit IRC | 08:59 | |
*** zzzeek has quit IRC | 09:00 | |
*** zzzeek has joined #openstack-ironic | 09:00 | |
*** derekh has joined #openstack-ironic | 09:13 | |
*** amarao has joined #openstack-ironic | 09:14 | |
*** dtantsur|afk is now known as dtantsur | 09:23 | |
dtantsur | morning ironic, morning vdrok | 09:23 |
dtantsur | TGIF finally | 09:23 |
openstackgerrit | Merged openstack/ironic-ui master: Improvements to enroll-node service https://review.openstack.org/426323 | 09:27 |
openstackgerrit | Merged openstack/ironic-ui master: Remove node-action deleteNodes/deletePorts functions https://review.openstack.org/426334 | 09:27 |
openstackgerrit | Merged openstack/ironic-ui master: Imported Translations from Zanata https://review.openstack.org/426667 | 09:27 |
openstackgerrit | Merged openstack/ironic-ui master: Improve error handling for the ironic-ui service https://review.openstack.org/426303 | 09:28 |
*** udesale__ has joined #openstack-ironic | 09:28 | |
*** udesale has quit IRC | 09:30 | |
*** athomas has quit IRC | 09:34 | |
*** mgoddard_ has joined #openstack-ironic | 09:37 | |
*** lucas-afk is now known as lucasagomes | 09:38 | |
lucasagomes | morning all | 09:38 |
*** chlong has quit IRC | 09:38 | |
openstackgerrit | Markos Chandras proposed openstack/bifrost master: roles: bifrost-ironic-install: Always download the latest iPXE files https://review.openstack.org/427682 | 09:43 |
*** udesale__ has quit IRC | 09:46 | |
*** udesale has joined #openstack-ironic | 09:46 | |
*** udesale has quit IRC | 09:46 | |
*** amotoki has joined #openstack-ironic | 09:47 | |
*** udesale has joined #openstack-ironic | 09:52 | |
vdrok | morning dtantsur and lucasagomes, tgif! :) | 09:57 |
lucasagomes | o/ indeed | 09:57 |
amarao | Hello everyone. I found something strange in ironic: nova shows 5 hypervisor for one compute. And this is exactly number of ironic nodes I have. Is this a coincidence? Is this a bug or a feature? | 10:03 |
*** amotoki has quit IRC | 10:05 | |
dtantsur | amarao, this is how ironic nodes are represented in nova | 10:05 |
dtantsur | so this is a feature, yes :) | 10:05 |
amarao | dtantsur: Oh, thanks. A bit unexpected at first, but reasonable after some thoughts. I think it should be noted somewhere in the docs for ironic installation. | 10:07 |
dtantsur | yeah, good point | 10:07 |
dtantsur | amarao, mind filing a bug for this missing docs bit? | 10:08 |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic master: Fix object save after refresh failure https://review.openstack.org/354744 | 10:12 |
amarao | dtanstsur, ack. | 10:13 |
*** jaganathan has quit IRC | 10:17 | |
openstackgerrit | Merged openstack/bifrost master: Remove link to modindex https://review.openstack.org/428097 | 10:22 |
*** vsaienko has quit IRC | 10:25 | |
*** sambetts|afk is now known as sambetts | 10:27 | |
sambetts | Morning all o/ | 10:27 |
lucasagomes | sambetts, o/ | 10:28 |
sambetts | hey lucasagomes | 10:30 |
amarao | I messing up with scheduler and I stumble on a question: how Ironic (nova-compute?) desides which ironic node is served by which compute? I'm trying to create availability zones. I created two computes, but I can't find a way to split nodes between those computes. | 10:30 |
*** udesale__ has joined #openstack-ironic | 10:30 | |
*** amotoki has joined #openstack-ironic | 10:31 | |
*** jaganathan has joined #openstack-ironic | 10:32 | |
*** udesale__ has quit IRC | 10:33 | |
*** udesale has quit IRC | 10:33 | |
dtantsur | morning sambetts | 10:34 |
dtantsur | amarao, I think it used a hash ring to distribute nodes. I think jroll, when he's here, can give a more detailed answer. | 10:35 |
amarao | thanks. | 10:35 |
*** owalsh-afk is now known as owalsh | 10:35 | |
amarao | But is here a way to create some kind of availability zones/aggregates? Or it is ouside of current features? | 10:35 |
sambetts | amarao: Ironic only exposes a single pool of resources, the only way that I can see aggregates working is having multiple nova-computes that point at different Ironics with different pools of resources | 10:37 |
amarao | But is this possible within a single openstack installation? I mean, to have two or more ironic in one keystone endpoints catalog? | 10:38 |
amarao | sambetts: I think it's possible to make some mechanism to associate given ironic node to a given nova-compute. After that all nova features with aggregates, availability zones, metas in flavors, etc, would start to work. | 10:39 |
*** amotoki has quit IRC | 10:40 | |
sambetts | hmm that could be interesting, I'd have to think about the HA impact of that, the way it works today means that if any nova compute died the others can take over, in a scenario where a ironic node is explictly tied to a nova-compute you lose that | 10:42 |
amarao | sambetts: It's can be easily fixed with few nova-computes with the same set of affinity. F.e. 2 compute nodes for aggregate1, 2 compute nodes for aggregate2. | 10:43 |
*** hw_wutianwei has quit IRC | 10:44 | |
amarao | So for now a single way to prefer some nodes over others during instance scheduling is to use different flavors with different meta... | 10:48 |
sambetts | amarao: yeah flavor extra specs is one way to do it currently | 10:48 |
vdrok | morning sambetts | 10:49 |
sambetts | hey vdrok | 10:49 |
amarao | Thanks. Mind if I file a bug/rfe for aggregates? | 10:49 |
vdrok | amarao: morning, in the new resource providers world that would not be needed | 10:49 |
sambetts | vdrok: oh?! will that allow us to do affinity? | 10:49 |
sambetts | vdrok: e.g. have an aggregate that pick one rack over another? | 10:50 |
vdrok | amarao: each ironic node is a resource provider in nova. in future you'll be able to associate resource providers to aggregates and scheduler will be able to schedule based on that | 10:50 |
amarao | Wow. Sounds cool. | 10:50 |
sambetts | that sounds awesome! | 10:51 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic master: Log an actual error message when failed to load new style credentials https://review.openstack.org/428690 | 10:51 |
vdrok | sambetts: if you create your aggregates that way, I think it is | 10:51 |
sambetts | we should definatly get an RFE up for it, just to track the idea | 10:51 |
sambetts | and then maybe vdrok you can leave a comment on that regarding the resource providers | 10:51 |
amarao | Ok, I'll write it now. | 10:51 |
sambetts | amarao: thanks! | 10:52 |
vdrok | but, I think it's not happening until pike, I mean, resource providers are already created in nova, but scheduler does not know yet anything about the aggregates they are associated to | 10:52 |
sambetts | +1 thats why I want to track it | 10:52 |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic master: ETAGs maintenance for all api node's requests https://review.openstack.org/392213 | 10:52 |
sambetts | because its totally a think that I want to see for affinity and stuff | 10:52 |
* dtantsur -> food and errands, back later | 10:55 | |
*** dtantsur is now known as dtantsur|bbl | 10:55 | |
*** jaganathan has quit IRC | 10:59 | |
*** yolanda has quit IRC | 11:08 | |
*** yolanda has joined #openstack-ironic | 11:10 | |
*** jaganathan has joined #openstack-ironic | 11:11 | |
openstackgerrit | Merged openstack/ironic-inspector master: Add plugin to process basic management LLDP TLVs https://review.openstack.org/406496 | 11:17 |
*** sdake has joined #openstack-ironic | 11:18 | |
*** athomas has joined #openstack-ironic | 11:21 | |
*** jaganathan has quit IRC | 11:23 | |
*** amotoki has joined #openstack-ironic | 11:26 | |
jroll | morning y'all | 11:31 |
sambetts | morning jroll | 11:31 |
jroll | vdrok: thanks for answering that, trying to think about the hash ring + aggregates made my brain hurt | 11:31 |
jroll | (I still think we may need to do some work there) | 11:31 |
vdrok | morning jroll | 11:32 |
jroll | I'm thinking we'll need to make this aggregate aware or something https://github.com/openstack/nova/blob/master/nova/virt/ironic/driver.py#L541-L549 | 11:32 |
jroll | idk | 11:32 |
jroll | tis weird | 11:32 |
* jroll needs to make a ptg list of nova conversations to have | 11:33 | |
*** cdearborn has joined #openstack-ironic | 11:33 | |
vdrok | jroll: hmm, the hash_ring was added purely to prevent overreporting of resources right? and as it seems to me that it won't be the case if resource providers are used, and we might just drop it? | 11:35 |
jroll | vdrok: well, also so that two n-cpu processes don't both try to control a single ironic node at the same time | 11:35 |
*** jaganathan has joined #openstack-ironic | 11:36 | |
jroll | nova has locks on instances but those are in-memory so can't be shared between n-cpus | 11:36 |
*** bigjools has quit IRC | 11:36 | |
openstackgerrit | Markos Chandras proposed openstack/ironic-python-agent master: imagebuild: tinyipa: Make the install-deps.sh SUSE aware https://review.openstack.org/428704 | 11:37 |
jroll | distributed locking would help, maybe completely solve things | 11:37 |
amarao | jroll, how nodes (hypervisors) are assigned to nova-computes? | 11:37 |
*** bigjools has joined #openstack-ironic | 11:37 | |
*** amotoki has quit IRC | 11:37 | |
jroll | amarao: somewhat randomly, let me find the release notes and see if that makes sense | 11:37 |
vdrok | amarao: https://specs.openstack.org/openstack/nova-specs/specs/newton/implemented/ironic-multiple-compute-hosts.html | 11:37 |
jroll | ah yeah, spec would work :P | 11:38 |
*** jkilpatr has quit IRC | 11:38 | |
amarao | vdrok: thanks. I thought to patch it for our needs to make it 'rack aware', but I looks our of my reach. | 11:40 |
*** sdake_ has joined #openstack-ironic | 11:43 | |
jroll | amarao: locality awareness is definitely something I (personally) want to do, but it might be a couple cycles before we get there | 11:45 |
*** sdake has quit IRC | 11:46 | |
vdrok | hm, yeah, locking is the problem, will have to think more about that too :) | 11:46 |
amarao | jroll, Yes, that's a difference between development and operations. I will try to do something with it today, may be at monday. If I can't fix (..em... enhance), than I will try other ways around. | 11:47 |
jroll | amarao: hm? what's the difference? | 11:48 |
amarao | jroll:Timescale: two release cycles for development, 2 days for operations. Obviously, all I can do is a crude patches, but they will be available in two days or it's no-go. | 11:50 |
lucasagomes | jroll, morning o/ | 11:50 |
*** vsaienko has joined #openstack-ironic | 11:50 | |
jroll | amarao: yeah, I hear you, though I think much of that is driven by openstack's great desire to have something mediocre that works for everyone, as opposed to doing something quickly that works for someone and iterating :) | 11:53 |
jroll | amarao: curious what you're actually trying to do here, let a nova user pick a rack? :) | 11:54 |
*** pmannidi has quit IRC | 11:54 | |
amarao | jroll ...and maintainability. 20 lines-patch for 'we need that feature yesteday' will cause some minor headache in the future, but is solvable in reasonable time. If such patch is not 20-lines, but 40k+ - good luck maintain it (I pass). | 11:55 |
jroll | heh | 11:55 |
*** crushil has joined #openstack-ironic | 11:56 | |
*** jkilpatr has joined #openstack-ironic | 11:57 | |
*** lucasagomes is now known as lucas-hungry | 11:57 | |
amarao | How crazy is idea to replace hash_ring with own implementation which will use properties of node & service to provide affinity? | 11:57 |
*** fellypefca has joined #openstack-ironic | 11:58 | |
*** fellypefca has quit IRC | 11:59 | |
jroll | amarao: yeah, this has been done, that's why I ask what your real goal is | 11:59 |
*** fellypefca has joined #openstack-ironic | 11:59 | |
amarao | jroll: sorry. I want to implement a way to use aggregates/availability-zones in nova. In my idea each nova-compute serves own aggregate, and provides (manages?) only nodes with given property. When instance is scheduled, it assigned to aggregate (through --availability-zone option). scheduler chooses nova-compute in that aggregate, and that nova-compute uses only hypervisors with proper meta. | 12:02 |
*** rob0t_ has joined #openstack-ironic | 12:02 | |
jaganathan | Hi, Please let me know steps to run test cases in ironic-python-agent | 12:02 |
jroll | amarao: right okay, I think you can do that somehow, melwitt and friends may have some tips when she's around, they've done something like that | 12:03 |
jroll | I just don't know if the hash ring messes that up, it might | 12:03 |
*** jkilpatr has quit IRC | 12:04 | |
jroll | jaganathan: tox, like most openstack projects :) | 12:04 |
jroll | jaganathan: this is for ironic, but would be similar, just different dependencies http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html | 12:05 |
sambetts | jroll: I think the problem is that all nova-computes share the same Ironic pool of resources, so you can't use an agregate to force affinity by just picking the nova-compute | 12:05 |
jroll | sambetts: yeah, there's some scheduler filters involved I think? | 12:06 |
sambetts | jroll: ah custom for Ironic? | 12:06 |
*** rob0t_ has quit IRC | 12:07 | |
jroll | sambetts: yeah something like that | 12:07 |
jroll | I know yahoo does affinity | 12:07 |
jroll | but I don't know if code is up somewhere or any sort of writeup | 12:07 |
jroll | they do rack/power affinity for sure. maybe hall/room | 12:08 |
jroll | network even? | 12:08 |
*** vsaienko has quit IRC | 12:08 | |
*** vsaienko has joined #openstack-ironic | 12:08 | |
sambetts | I think this falls into a similar catagory to conducotr affienty, e.g. having a conductor per rack/row is interesting too | 12:09 |
sambetts | then if you did that you could have nova-compute to conductor affienty | 12:09 |
jroll | yeah, +1 | 12:09 |
jroll | and -1 | 12:09 |
jroll | respectively | 12:10 |
sambetts | hehehe | 12:10 |
jroll | :) | 12:10 |
jroll | I'd love conductor + image server (or caching proxy) per rack | 12:10 |
jroll | even running inside the tor if possible | 12:11 |
jroll | would be nice for traffic to not really need to leave the rack to deploy a box | 12:11 |
sambetts | yeah that would be sweet | 12:11 |
sambetts | although we'd have to fix the IPA talking to the Ironic API thing | 12:12 |
jroll | ah yeah | 12:12 |
* sambetts really hates that layer break | 12:12 | |
sambetts | it means that I have to wire two controller boxes into my provisioning network instead of just one | 12:13 |
*** rob0t_ has joined #openstack-ironic | 12:17 | |
*** jkilpatr has joined #openstack-ironic | 12:17 | |
*** gcb has quit IRC | 12:19 | |
amarao | jroll: If images are in swift, IPA will do requests to outside of the rack. | 12:19 |
mat128 | good morning | 12:19 |
jroll | amarao: well, yeah, I'm saying I'd love to get there, it isn't trivial today | 12:20 |
jroll | morning mat128 | 12:20 |
*** rob0t_ has quit IRC | 12:21 | |
*** rob0t_ has joined #openstack-ironic | 12:23 | |
mat128 | amarao: I believe you can already attain your goal by using nova filters/weighers | 12:25 |
mat128 | you will only have a single nova-compute but instances of X zone will end up on X types of nodes, which seems to be what you want | 12:26 |
mat128 | (or multiple computes, but not in the sense of them having different config and "handling" different AZ) | 12:26 |
*** jaganathan has quit IRC | 12:27 | |
*** sdake has joined #openstack-ironic | 12:28 | |
mat128 | amarao: we built user/tenant reservation using filters for now, so the node has a tenant uuid and only that tenant can boot an instance on it | 12:28 |
*** rob0t_ has quit IRC | 12:28 | |
mat128 | amarao: you could use any custom attribute and do the same thing | 12:28 |
mat128 | based on the requested availability zone | 12:28 |
*** sdake_ has quit IRC | 12:30 | |
amarao | mat128, I didn't get it. We got different types (e.g. mem/cpu/disk) servers in two AZ. We want users have a way to say which AZ they want to use. How it will works with weights? | 12:30 |
*** rcernin has quit IRC | 12:31 | |
mat128 | amarao: you should use filters | 12:31 |
mat128 | amarao: are there servers of the same hardware in both AZ? | 12:32 |
amarao | Yep. That's the problem. | 12:32 |
mat128 | amarao: then servers could have an extra attribute in Ironic, and a custom filter would read that and the requested AZ and filter on it | 12:32 |
amarao | And how user can pass this attribute at 'nova boot' time? | 12:33 |
mat128 | amarao: they already pass an --az attribute, no? | 12:33 |
mat128 | amarao: filter_properties has a request_spec | 12:34 |
mat128 | which probably contains it | 12:34 |
amarao | Yes, they do. I don't know if it counted as 'attribute' or not. | 12:34 |
mat128 | amarao: the az is readable because the availability zone filter uses it | 12:34 |
mat128 | https://github.com/openstack/nova/blob/master/nova/scheduler/filters/availability_zone_filter.py | 12:34 |
mat128 | amarao: so it's just a matter of matching this to a "node" spec | 12:34 |
amarao | Oh, I see the idea. Throw away AggregateFilter and pass az info directly into other filters. | 12:34 |
mat128 | or keep both, but yeah | 12:35 |
mat128 | amarao: host_state.stats contains the node information | 12:35 |
mat128 | amarao: we essentially do host_state.stats.get('reserved_for_tenant_id') and set that in the node properties | 12:35 |
amarao | Wow. That's a really clever idea. | 12:36 |
mat128 | amarao: and we have this in our nova driver to expose more properties: | 12:37 |
mat128 | http://paste.openstack.org/show/597547/ | 12:37 |
openstackgerrit | Merged openstack/ironic master: Validate port info before assume we may use it https://review.openstack.org/415881 | 12:37 |
mat128 | :) | 12:37 |
amarao | mat128, thanks! | 12:39 |
openstackgerrit | Merged openstack/ironic master: Speed up irmc power unit tests https://review.openstack.org/428480 | 12:39 |
openstackgerrit | Markos Chandras proposed openstack/bifrost master: roles: bifrost-create-vm-nodes: Randomize VM XML file https://review.openstack.org/428716 | 12:44 |
*** fellypefca has quit IRC | 12:52 | |
*** rob0t_ has joined #openstack-ironic | 12:52 | |
vdrok | jroll: so it seems I can not get the problem with aggregates and hash ring, the scheduler will select a node in aggregate, both 'host' and 'nodename' are in the dictionary scheduler returns, and it sends a build request to that compute. Or is the problem that the ring can rebalance and node will move to different host? | 12:52 |
*** athomas has quit IRC | 12:52 | |
vdrok | maybe I just don't understand nova enough tho :) | 12:52 |
*** athomas has joined #openstack-ironic | 12:52 | |
*** rob0t_ has quit IRC | 12:57 | |
*** jvgrant has quit IRC | 12:58 | |
*** jvgrant_ has joined #openstack-ironic | 12:58 | |
*** lucas-hungry is now known as lucasagomes | 12:59 | |
sambetts | vdrok: I see the problem as this: because the nodes are assigned to nove-computes using the hash ring, you can't ever know before hand where a node will be assigned, so you can't create an host aggregate that makes sense just based on nova-compute | 12:59 |
openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector master: Find a node by multiple attributes https://review.openstack.org/421285 | 13:00 |
vdrok | sambetts: yup, I just mean that if you need an aggregate==room, it does not matter if the compute is the same or not? it is not the strict node affinity in nova's sense tho | 13:01 |
*** milan has joined #openstack-ironic | 13:01 | |
*** ccamacho is now known as ccamacho|lunch | 13:02 | |
milan | morning Ironic! :) | 13:02 |
vdrok | like I want a node in room1, not in room2, I do nova boot --availability-zone room1 | 13:02 |
vdrok | morning milan | 13:02 |
milan | morning vdrok! :) | 13:02 |
*** jaganathan has joined #openstack-ironic | 13:02 | |
sambetts | vdrok: right, so say AZ==room affinity, aggregates could be used for rack affinity for example, e.g. have a nova-compute per rack that manages only the nodes in that rack | 13:03 |
*** rob0t_ has joined #openstack-ironic | 13:04 | |
vdrok | sambetts: ah, gotcha | 13:05 |
vdrok | so we need to change that bit of hash ring only when we need nova to manage a specific bit of nodes | 13:05 |
sambetts | something like that yeah, you need a nova-compute to defined group of Ironic nodes mapping (maybe this could be a use for tags) | 13:08 |
*** rob0t_ has quit IRC | 13:09 | |
sambetts | vdrok: :O my brain has just clicked! | 13:09 |
vdrok | :D | 13:09 |
sambetts | we could add a configuration option that takes a list of tags to the nova ironic driver, that means that when it does the node-list it filters by list of tags! | 13:10 |
sambetts | then you can say this nova-compute only manages node with a tag of "rack11" or something | 13:10 |
vdrok | sambetts: yeah, the only problems is we don't have tags yet :) | 13:11 |
sambetts | :/ yeah... sigh... | 13:11 |
vdrok | we did something like that with chassis_uuid tho | 13:11 |
vdrok | assigning a specific chassis to a compute | 13:11 |
mat128 | sambetts: I wouldnt want to have to deploy a nova-compute per rack, there would be a ton of them | 13:11 |
*** aparnav has quit IRC | 13:12 | |
*** rob0t_ has joined #openstack-ironic | 13:12 | |
sambetts | mat128: tis but an example ;) I guess it depends on how HA you really want to get :-P | 13:13 |
*** jaganathan has quit IRC | 13:13 | |
mat128 | sambetts: yeah I figured :) | 13:14 |
sambetts | vdrok: with tags instead of chassis id though you could have nova take a list of tags, so it could be like node-tags=rack11,rack12,rack13 and then it filter nodes with those tags | 13:15 |
sambetts | or node-tags=row5,high-storage | 13:15 |
sambetts | hmmmm /me loves tags/labels | 13:16 |
*** rob0t_ has quit IRC | 13:16 | |
*** vsaienko has quit IRC | 13:19 | |
vdrok | yup, that's an advantage | 13:19 |
openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient master: Do not show chassis_uuid field when it is not specified https://review.openstack.org/427815 | 13:20 |
*** bfournie has quit IRC | 13:20 | |
*** rob0t_ has joined #openstack-ironic | 13:22 | |
*** dprince has joined #openstack-ironic | 13:23 | |
*** jheroux has joined #openstack-ironic | 13:26 | |
*** rob0t_ has quit IRC | 13:26 | |
*** g3ek has quit IRC | 13:28 | |
*** jaganathan has joined #openstack-ironic | 13:29 | |
*** awiddersheim has quit IRC | 13:29 | |
*** awiddersheim has joined #openstack-ironic | 13:30 | |
*** crushil has quit IRC | 13:32 | |
jroll | sambetts: oh, that's interesting | 13:32 |
*** trown|outtypewww is now known as trown | 13:32 | |
*** rob0t_ has joined #openstack-ironic | 13:34 | |
jroll | vdrok: the aggregate thing might actually work, idk enough about aggregates. I'm worried aggregates only care about "hosts", not "nodes", and so you might send a build request for a node to a host that doesn't know about it (which might work?). maybe it's handled in the scheduling decision, though, not sure | 13:34 |
* jroll feels like he's just throwing random words together here | 13:34 | |
*** hoonetorg has quit IRC | 13:35 | |
sambetts | haha | 13:35 |
*** g3ek has joined #openstack-ironic | 13:37 | |
*** rob0t_ has quit IRC | 13:39 | |
vdrok | jroll: yeah, aggregates will care about resource providers, not compute hosts anymore, I guess. it works now already with resource providers, and we still get (host, node) pair from the scheduler, so hopefully this will be happening in future too :) | 13:40 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc master: Fix help string about adding a VBMC using IPv4 https://review.openstack.org/428725 | 13:41 |
jroll | vdrok: yeah, I think we're fine eventually, I just wonder about today :) | 13:41 |
*** rob0t_ has joined #openstack-ironic | 13:42 | |
*** rob0t_ has quit IRC | 13:47 | |
*** hoonetorg has joined #openstack-ironic | 13:48 | |
*** amotoki has joined #openstack-ironic | 13:48 | |
*** vsaienko has joined #openstack-ironic | 13:50 | |
openstackgerrit | Galyna Zholtkevych proposed openstack/python-ironicclient master: Include driver properties to osc `driver show` https://review.openstack.org/381153 | 13:50 |
*** bfournie has joined #openstack-ironic | 13:50 | |
*** srobert has joined #openstack-ironic | 13:52 | |
*** crushil has joined #openstack-ironic | 13:53 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic master: Add Ironic standalone tests https://review.openstack.org/423556 | 13:54 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic master: [Devstack] fix waiting resources on subnode https://review.openstack.org/427675 | 13:54 |
*** srobert has quit IRC | 13:55 | |
*** srobert has joined #openstack-ironic | 13:55 | |
*** rob0t_ has joined #openstack-ironic | 13:56 | |
amarao | mat128, I dug into scheduler and I can't find any capabilities of node (ironic node) inside 'hosts' object (hosts.next()) in _schedule function in filters. I can see meta for nova-level host, but I can't see hypervisor capabilities... | 13:58 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc master: Document some usage with examples https://review.openstack.org/428737 | 13:59 |
*** crushil has quit IRC | 14:00 | |
*** dtantsur|bbl is now known as dtantsur | 14:00 | |
*** rob0t_ has quit IRC | 14:01 | |
*** rob0t_ has joined #openstack-ironic | 14:06 | |
TheJulia | Good morning | 14:09 |
dtantsur | morning TheJulia, jroll, milan | 14:10 |
TheJulia | good morning dtantsur | 14:10 |
* TheJulia needs lots of coffee this morning | 14:10 | |
*** rob0t_ has quit IRC | 14:10 | |
* dtantsur too | 14:10 | |
* jroll three | 14:11 | |
*** crushil has joined #openstack-ironic | 14:12 | |
*** rob0t_ has joined #openstack-ironic | 14:12 | |
jroll | dtantsur: I just realized neutron network_interface isn't enabled by default, is there a reason for that? vdrok ? | 14:15 |
jroll | is it because extra config or? | 14:15 |
dtantsur | jroll, because it fails without provisioning_network set? | 14:15 |
jroll | dtantsur: right, so extra config | 14:16 |
*** baoli has joined #openstack-ironic | 14:16 | |
vdrok | Yup, no specific reason apart from that | 14:16 |
jroll | I'd like to work to make multitenancy default, I think | 14:16 |
jroll | or at least enabled by default | 14:16 |
vdrok | Morning TheJulia | 14:16 |
*** rob0t_ has quit IRC | 14:17 | |
* jroll wants ironic to be sane, secure, prod-ready with defaults | 14:17 | |
*** athomas has quit IRC | 14:20 | |
*** ccamacho|lunch is now known as ccamacho | 14:20 | |
TheJulia | enabled by default makes sense, I think it is only when the driver is actually loaded that it would explode without the configuration.... and if we went to supporting node level definition of provisioning networks then the check for that would actually need to be moved | 14:20 |
dtantsur | jroll, I agree.. though as I mentioned some time ago, pre-defining a network in ironic.conf gives me a huge headache, given tripleo architecture.. | 14:21 |
*** rob0t_ has joined #openstack-ironic | 14:22 | |
*** Goneri has joined #openstack-ironic | 14:22 | |
TheJulia | dtantsur: would node level be better? | 14:23 |
jroll | dtantsur: yeah, I would think we should fix tripleo architecture in that case | 14:23 |
jroll | or we could explode late, dunno | 14:23 |
jroll | or check when we set it on a node? and check if it's set on any nodes at startup? | 14:24 |
TheJulia | not even check if it is set on nodes, check if it exists as part of one or more validation steps | 14:24 |
dtantsur | TheJulia, much better | 14:25 |
TheJulia | err, not even check if it is set on startup | 14:25 |
*** owalsh is now known as owalsh-brb | 14:25 | |
TheJulia | for some reason my macbook is lagging really badly. Time to close all of the web browsers | 14:25 |
dtantsur | jroll, "fix tripleo architecture" is not an easy win, as you may expect :) though right now it seems like Octavia has similar requirements | 14:25 |
*** jtomasek has quit IRC | 14:25 | |
jroll | yeah, we could just avoid checking at startup | 14:25 |
*** mjura has quit IRC | 14:25 | |
jroll | dtantsur: pft how hard could it be | 14:25 |
jroll | :) | 14:26 |
*** jtomasek has joined #openstack-ironic | 14:26 | |
*** mjura has joined #openstack-ironic | 14:26 | |
TheJulia | we need it to deploy, adding it into deploy validate() makes sense to me | 14:26 |
*** athomas has joined #openstack-ironic | 14:26 | |
dtantsur | TheJulia, I think I have an RFE for that, did not have time to actually code it | 14:26 |
*** rob0t_ has quit IRC | 14:27 | |
TheJulia | Yeah, there isa n RFE for it | 14:27 |
openstackgerrit | Merged openstack/bifrost master: scripts: test-bifrost.sh: Use 'unsafe' disk cache for test VMs https://review.openstack.org/427744 | 14:28 |
*** owalsh-brb has quit IRC | 14:29 | |
*** rob0t_ has joined #openstack-ironic | 14:32 | |
*** owalsh-brb has joined #openstack-ironic | 14:32 | |
*** marlinc has quit IRC | 14:32 | |
*** jcoufal has joined #openstack-ironic | 14:35 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc master: Fix help string about adding a VBMC using IPv4 https://review.openstack.org/428725 | 14:36 |
*** rob0t_ has quit IRC | 14:37 | |
jroll | ./ironic/conductor/base_manager.py:56:1: C901 'BaseConductorManager.init_host' is too complex (17) | 14:37 |
jroll | :| | 14:37 |
dtantsur | lol, I don't disagree with that | 14:37 |
jroll | heh | 14:38 |
jroll | yeah, we need to clean some of that up | 14:39 |
jroll | I just wonder how much gets cleaned up naturally if/when we kill drivers | 14:39 |
*** marlinc has joined #openstack-ironic | 14:41 | |
TheJulia | pas-ha: I reviewed 427071, biggest issue that I have is the deprecations are not indicating when we're removing the old behavior | 14:44 |
lucasagomes | jroll, yeah, well there's a lot of room for improvements there, we could start breaking things into small functions (register_conductor(), start_periodic_tasks(), start_consoles(), etc...) | 14:44 |
lucasagomes | this init_host just do it all and separated the logic with comments :-) | 14:45 |
jroll | lucasagomes: yeah totally | 14:45 |
*** sasha21 has joined #openstack-ironic | 14:46 | |
*** rob0t_ has joined #openstack-ironic | 14:48 | |
*** amotoki has quit IRC | 14:48 | |
*** amotoki has joined #openstack-ironic | 14:48 | |
*** amotoki has quit IRC | 14:49 | |
* jroll notices bcornec picked the redfish spec back up | 14:50 | |
* jroll gets excited | 14:50 | |
TheJulia | bcornec: sorry I've not replied to the email you sent me a week or two ago. Life has been rather.. hectic as of recent. | 14:51 |
TheJulia | If anyone wants to review a relatively small chunk of javascript in ironic-ui https://review.openstack.org/#/c/418520/ | 14:52 |
patchbot | patch 418520 - ironic-ui - Create base-port module to support create and upda... | 14:52 |
*** mtanino has joined #openstack-ironic | 14:52 | |
* jroll looks | 14:53 | |
TheJulia | jroll: thank you | 14:54 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic master: Improve enabled_*_interfaces config help and validation https://review.openstack.org/428219 | 14:54 |
jroll | TheJulia: I totally forgot everything in horizon is using angular... /me has a sad | 14:54 |
* jroll tries to resurrect any sort of angular knowledge from the depths | 14:55 | |
* TheJulia gives jroll a whisky | 14:55 | |
robcresswell | jroll: More of a React guy? | 14:55 |
jroll | robcresswell: yeah, I've not played with it as much as I like, but some of my friends are react people and seem to have good reasons for it | 14:56 |
*** rob0t_ has quit IRC | 14:56 | |
dtantsur | "review a relatively small chunk of javascript" http://pruebakr-1.webcindario.com/nope.jpg | 14:56 |
TheJulia | dtantsur: nice | 14:57 |
dtantsur | :-P | 14:57 |
jroll | oh I like that one | 14:57 |
TheJulia | I feel like that should be my facebook profile picture | 14:57 |
robcresswell | Yeah. Angular 1.x was a mess. I've found it reasonable to use at 1.5/1.6 as long as you strictly follow certain styles, but then you end up using about 1/3 of the framework :/ | 14:57 |
jroll | TheJulia: +A | 14:59 |
TheJulia | ppiela: ^^^ | 14:59 |
TheJulia | jroll: I'm sure ppiela thanks you :) | 15:00 |
jroll | hehe | 15:00 |
jroll | oh, that's the only one in the queue, nice | 15:00 |
* jroll sees y'all have been busy the last week or so | 15:00 | |
*** mtanino_ has joined #openstack-ironic | 15:01 | |
*** mtanino has quit IRC | 15:02 | |
openstackgerrit | Merged openstack/ironic-ui master: Create base-port module to support create and update operations https://review.openstack.org/418520 | 15:02 |
ppiela | TheJulia, jroll: I do thank you | 15:02 |
*** rob0t_ has joined #openstack-ironic | 15:03 | |
jroll | :) | 15:03 |
*** rob0t_ has quit IRC | 15:08 | |
*** mtanino_ has quit IRC | 15:09 | |
milan | morning dtantsur jroll ppiela sambetts TheJulia and all! :) | 15:09 |
jroll | morning milan :) | 15:10 |
ppiela | Hi milan | 15:10 |
*** pcaruana has quit IRC | 15:11 | |
*** mtanino has joined #openstack-ironic | 15:11 | |
robcresswell | Oh, question from the other day; are there any decent ways to test ironic in a standalone (noauth) way with a virtual driver/backend? I was looking for a way to test some UI work. I think Bifrost does what I want, but I'm an ironic noob so I'm unsure. | 15:14 |
robcresswell | I'm aware of the fake driver, but there's only so far that will get you with testing validation etc. | 15:14 |
*** pcaruana has joined #openstack-ironic | 15:17 | |
*** rob0t_ has joined #openstack-ironic | 15:19 | |
*** amotoki has joined #openstack-ironic | 15:20 | |
milan | aarefiev_afk, hey, just read your comment on https://review.openstack.org/#/c/421285/ there was yet another issue: had to specify column type for the batch_op.alter_column for MySQL | 15:21 |
patchbot | patch 421285 - ironic-inspector - Find a node by multiple attributes | 15:21 |
aarefiev_afk | milan: this issue is easy to fix :) | 15:21 |
milan | aarefiev_afk, the primary key should be OK dropping during the batch_op, now the issue seems http://logs.openstack.org/85/421285/12/check/gate-grenade-dsvm-ironic-inspector-ubuntu-xenial/e4bc510/console.html#_2017-02-03_14_28_07_738596 | 15:22 |
milan | aarefiev_afk, yeah, I've posted an update | 15:22 |
milan | I think I might have it corrected | 15:22 |
*** chlong has joined #openstack-ironic | 15:24 | |
TheJulia | good morning milan | 15:24 |
milan | o/ :) | 15:24 |
*** Nisha_Agarwal has joined #openstack-ironic | 15:25 | |
aarefiev_afk | milan: hmm, btw did you try to run tests with mysql | 15:25 |
milan | aarefiev_afk, I did in devstack | 15:26 |
*** rob0t_ has quit IRC | 15:26 | |
milan | aarefiev_afk, I tried re-running the migrations couple of times till I figured the stack trace ;) | 15:26 |
*** amotoki has quit IRC | 15:26 | |
milan | * traceback | 15:26 |
milan | as the alembic is consuming it happily ;) | 15:27 |
milan | aarefiev_afk, wdyt about this revision? | 15:27 |
* milan not sure what happened to the grenade test though | 15:27 | |
aarefiev_afk | milan: reading grenade logs | 15:27 |
*** owalsh-brb is now known as owalsh | 15:28 | |
*** jcoufal_ has joined #openstack-ironic | 15:34 | |
*** rajinir has joined #openstack-ironic | 15:36 | |
*** jcoufal has quit IRC | 15:36 | |
*** rob0t_ has joined #openstack-ironic | 15:37 | |
*** jaosorior has joined #openstack-ironic | 15:40 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic master: Add Ironic standalone tests https://review.openstack.org/423556 | 15:41 |
sambetts | TheJulia: can you help robcresswell? He wants to use bifrost with VMs but I'm not sure how that works | 15:41 |
TheJulia | o/ robcresswell | 15:41 |
robcresswell | Not a big problem if its not an option, I'm just curious :) | 15:41 |
robcresswell | \o | 15:42 |
TheJulia | Oh, it is! | 15:42 |
*** rob0t_ has quit IRC | 15:42 | |
milan | aarefiev_afk, IMO the inspector smoke test was OK (i can see the node returning back to avail in ir-cond) | 15:42 |
*** vsaienko has quit IRC | 15:42 | |
TheJulia | they would be local libvirt instances, if that works for you? | 15:42 |
milan | http://logs.openstack.org/85/421285/12/check/gate-grenade-dsvm-ironic-inspector-ubuntu-xenial/e4bc510/console.html#_2017-02-03_14_25_01_656242 has some sort of a timeout :-/ | 15:42 |
aarefiev_afk | milan: I don't see any issues, really timeout? :) | 15:42 |
milan | aarefiev_afk, 2017-02-03 14:28:07.738541 | tempest.exceptions.BuildErrorException: Server e5d55745-1eab-46ce-a834-99628cf79a29 failed to build and is in ERROR status | 15:43 |
milan | aarefiev_afk, http://logs.openstack.org/85/421285/12/check/gate-grenade-dsvm-ironic-inspector-ubuntu-xenial/e4bc510/console.html#_2017-02-03_14_28_07_738541 | 15:43 |
milan | provisioning failed, I dunno | 15:44 |
milan | recheck? | 15:44 |
aarefiev_afk | milan: yeah timeout while wait for callback and I don't see errors in bm logs | 15:44 |
aarefiev_afk | yeah lets try | 15:44 |
milan | cool | 15:44 |
robcresswell | TheJulia: I'm not picky. I'm really just looking for something to bounce API requests off without the overheads of actual baremetal, like ten minute power cycles :) | 15:45 |
openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient master: Do not show chassis_uuid field when it is not specified https://review.openstack.org/427815 | 15:45 |
robcresswell | This is all pretty new to me, so I'm kinda wandering around the dark trying to get it to work | 15:46 |
milan | sambetts, hey, do you plan to review https://review.openstack.org/#/c/375045/ ? ;) | 15:47 |
patchbot | patch 375045 - ironic-inspector-specs - Endpoint to list introspection statuses | 15:47 |
TheJulia | robcresswell: Bifrost has a script in it's repo that can be used to create a set of virtual machines. The output of which is a CSV file which you can use to deploy an image to, although I'm not really sure what of what API your wanting to hit :) | 15:47 |
milan | sambetts, and also: https://review.openstack.org/#/c/396147/ :D | 15:48 |
patchbot | patch 396147 - ironic-inspector-specs - Spec for retrieving NUMA node information | 15:48 |
robcresswell | TheJulia: Which script is that? | 15:49 |
robcresswell | TheJulia: I'm just looking to bounce a standalone UI off the ironic APIs, but setting up VMs instead of actual metal for speed of testing | 15:49 |
TheJulia | robcresswell: https://github.com/openstack/bifrost/blob/master/tools/create_vm_nodes.sh | 15:50 |
robcresswell | hey neat | 15:51 |
robcresswell | TheJulia: Thanks! | 15:51 |
openstackgerrit | Galyna Zholtkevych proposed openstack/python-ironicclient master: Include driver properties to osc `driver show` https://review.openstack.org/381153 | 15:51 |
*** rob0t_ has joined #openstack-ironic | 15:52 | |
*** Syed__ has joined #openstack-ironic | 15:56 | |
*** rob0t_ has quit IRC | 15:57 | |
sambetts | milan: I've not reviewed the NUMA one before, and I don't really have time to sit down and go through the spec right now, I want to re-read the list one though as I've been involved with that :) | 15:59 |
*** rpioso has joined #openstack-ironic | 15:59 | |
milan | sambetts, OK cool! Thanks for info! | 16:00 |
*** jcoufal_ has quit IRC | 16:03 | |
*** jcoufal has joined #openstack-ironic | 16:03 | |
*** Nisha_Agarwal has quit IRC | 16:06 | |
NobodyCam | Good Morning Ironic'ers ... and ... | 16:06 |
*** crushil has quit IRC | 16:06 | |
*** rob0t_ has joined #openstack-ironic | 16:06 | |
NobodyCam | TGIF! | 16:06 |
*** crushil has joined #openstack-ironic | 16:06 | |
*** jcoufal_ has joined #openstack-ironic | 16:07 | |
*** aparnav has joined #openstack-ironic | 16:07 | |
*** sdake_ has joined #openstack-ironic | 16:08 | |
*** fxpester has quit IRC | 16:08 | |
*** sdake has quit IRC | 16:08 | |
*** sdake has joined #openstack-ironic | 16:09 | |
*** jcoufal has quit IRC | 16:09 | |
*** ccamacho has quit IRC | 16:11 | |
*** jcoufal has joined #openstack-ironic | 16:11 | |
*** rob0t_ has quit IRC | 16:11 | |
*** sdake has quit IRC | 16:12 | |
*** sdake_ has quit IRC | 16:12 | |
*** ccamacho has joined #openstack-ironic | 16:13 | |
*** jcoufal_ has quit IRC | 16:13 | |
*** baoli has quit IRC | 16:13 | |
dtantsur | morning NobodyCam | 16:16 |
dtantsur | oh TGIF indeed | 16:16 |
*** [1]cdearborn has joined #openstack-ironic | 16:17 | |
*** jaganathan has quit IRC | 16:17 | |
lucasagomes | NobodyCam, o/ morning | 16:19 |
*** rob0t_ has joined #openstack-ironic | 16:19 | |
*** baoli has joined #openstack-ironic | 16:19 | |
milan | morning NobodyCam! TGIF ++ :D | 16:19 |
*** dansmith is now known as superdan | 16:21 | |
*** rbudden has joined #openstack-ironic | 16:21 | |
* dtantsur has a day full of puppet.. | 16:21 | |
NobodyCam | Good Morning dtantsur, lucasagomes, and milan :) | 16:21 |
* NobodyCam was getting coffee | 16:22 | |
dtantsur | milan, bfournie, also FYI we broken RDO a bit with our LLDP patch, as they don't have the required 'construct' package even in Fedora :) | 16:22 |
bfournie | dtantsur: uugh | 16:23 |
vdrok | morning NobodyCam :) | 16:24 |
*** mjura has quit IRC | 16:24 | |
NobodyCam | Morning vdrok :) | 16:24 |
bfournie | dtantsur: how can we bring that construct lib patch down, it went in a month ago | 16:24 |
dtantsur | well, we're the first one to actually use it :) | 16:25 |
bfournie | dtantsur: right | 16:25 |
mat128 | amarao: you probably need something like this in your virt driver - http://paste.openstack.org/show/597547/ | 16:25 |
*** nmathew has quit IRC | 16:25 | |
amarao | mat128, this is patch to nova? | 16:26 |
*** rob0t_ has quit IRC | 16:26 | |
mat128 | thats a function in virt driver, the paste I sent you is just us overriding the method, but you could easily patch it in | 16:26 |
mat128 | amarao: let me craft you a patch :) | 16:27 |
mat128 | which version are you running? | 16:27 |
amarao | Well, I tried to find capabilities at scheduler level, they are in hosts.next().stats .stats[] | 16:27 |
mat128 | amarao: oh, good then | 16:27 |
amarao | I will try to make a custom filter. I think it's better than patch to nova. | 16:27 |
amarao | If I fail, I will try with nova patch. Thanks for offer. | 16:28 |
mat128 | amarao: indeed, thats good | 16:28 |
mat128 | amarao: no problem, I'm always idling here so feel free to ask | 16:28 |
amarao | Right now it looks like this: ironic node-update acafa4ea-23ac-49f6-9368-3a97a3d09bf8 add properties/capabilities='availability-zone:az2' | 16:28 |
amarao | And in scheduler context it is hosts[N].stats['u'availability-zone'] == u'az2' | 16:29 |
amarao | I have no idea why it's 'stats'... | 16:29 |
mat128 | amarao: in my case capabilities are comma separated for some reason | 16:30 |
mat128 | I dont remember where this comes from, but I think nova unpacks it that way | 16:30 |
amarao | mat128, hmm... I see it upacked them to dict. | 16:30 |
*** cdearborn has quit IRC | 16:31 | |
mat128 | amarao: a random test machine dump - http://paste.openstack.org/show/597590/ | 16:31 |
amarao | Oh, you are keeping them not in capabilities, but in properties itself... I will try this too. | 16:32 |
mat128 | amarao: oh, yes | 16:33 |
*** bnemec is now known as beekneemech | 16:33 | |
mat128 | and I think this is why we have that other thing - http://paste.openstack.org/show/597547/ | 16:33 |
mat128 | we end up having to filter on many criterias, and putting them all in capabilities is a pita | 16:33 |
amarao | ... (bad suspicion) Is capablities updated by some software? | 16:33 |
mat128 | amarao: nothing in openstack, besides ironic-inspector possibly | 16:33 |
*** mjura has joined #openstack-ironic | 16:33 | |
mat128 | amarao: on our side, we have custom code to export "nodes" data from a management system into Ironic | 16:34 |
amarao | Oh, so inspector may override them. | 16:34 |
mat128 | amarao: not 100% sure, but that's where it would put hardware capabilities | 16:34 |
*** baoli has quit IRC | 16:34 | |
mat128 | amarao: it'd be nice if inspector would derive the availability zone based on some other variable | 16:34 |
amarao | mat128, ok, my first attempt will be to put AZ information somewhere else. | 16:34 |
*** rob0t_ has joined #openstack-ironic | 16:35 | |
mat128 | amarao: maybe from kernel cmdline? then you could have 1 PXE server per AZ | 16:35 |
mat128 | amarao: going into a meeting, feel free to ping and I'll check later | 16:35 |
*** mat128 is now known as mat128|mtg | 16:35 | |
amarao | mat128: my current design will not use inspector. Just to make thinks less 'software-dependent'. | 16:35 |
amarao | mat128, thank you very much! | 16:35 |
*** chlong has quit IRC | 16:36 | |
milan | dtantsur, bfournie, fun type 2 :) | 16:36 |
*** ianychoi has joined #openstack-ironic | 16:37 | |
milan | also Fedora: Freedom. Friends. Features. First. ;) | 16:37 |
*** baoli has joined #openstack-ironic | 16:38 | |
ianychoi | Hello. I am from I18n team. Can I know when "ironic-ui" will have stable/ocata branch? It is important for I18n team to create a stable version for translation sync jobs with other Horizon related projects | 16:38 |
dtantsur | sambetts, quick question re your hardware. you said you can't detect ipmi address during introspection from within the node itself, right? does it mean that IPMI access to it is not possible at all or only from within the machine? | 16:39 |
*** rob0t_ has quit IRC | 16:39 | |
jroll | ianychoi: I was thinking next week, but maybe we're ready now. TheJulia what do you think? | 16:41 |
jroll | I'm not opposed to branching now | 16:41 |
dtantsur | branch early, branch often | 16:42 |
*** rob0t_ has joined #openstack-ironic | 16:42 | |
jroll | pls don't branch often | 16:42 |
jroll | :P | 16:42 |
*** xhku is now known as fbouliane | 16:43 | |
*** aparna has joined #openstack-ironic | 16:43 | |
JayF | for i in {1..100}; do git checkout -b branch$i; done | 16:44 |
ianychoi | I18n team now deals with total 12 horizon & related repos including ironic-ui. According to https://releases.openstack.org/ocata/schedule.html , this week is for the deadline of RC1 | 16:44 |
jroll | ianychoi: ironic follows the cycle-with-intermediary model, so our deadlines are a bit different | 16:45 |
dtantsur | jroll, this is a good call, we may want to branch ironic-ui with the remaining openstack to reduce headache for the i18n team | 16:45 |
*** appu103 has joined #openstack-ironic | 16:45 | |
ianychoi | Currently 5 repos do not have stable/ocata branch, and then now I want to have more conversation with details | 16:45 |
jroll | ianychoi: I didn't realize releasing later would affect translation work, I'm happy to branch it, just want to check with TheJulia first | 16:45 |
*** aparnav has quit IRC | 16:45 | |
jroll | dtantsur: agree | 16:45 |
dtantsur | just like we do with the CLI, which is also a UI :) | 16:45 |
ianychoi | jroll, aha I see Actually I admit that I am not familiar with cycle-with-intermediary model too much | 16:46 |
*** fbouliane has quit IRC | 16:47 | |
*** rob0t_ has quit IRC | 16:47 | |
jroll | ianychoi: looking for details, but I think we should change it for ironic-ui :) | 16:47 |
ianychoi | jroll, does the model apply to all horizon plugin projects? | 16:47 |
*** aparna has quit IRC | 16:48 | |
*** fbouliane has joined #openstack-ironic | 16:48 | |
jroll | ianychoi: I don't think so, and I think we need a rule to make horizon plugins cycle-with-milestones always | 16:48 |
jroll | ianychoi: see this page for description of the models https://releases.openstack.org/reference/release_models.html | 16:48 |
dtantsur | jroll, I think changing the model is too much, just setting up a deadline like with CLI is enough | 16:48 |
dtantsur | with the model, we'll have to make M1, M2, M3, RC1.. | 16:49 |
jroll | dtantsur: ah yeah, good point | 16:49 |
robcresswell | cycle-with-intermediary should have tagged a stable/ocata by now, or should do today I believe | 16:49 |
ianychoi | jroll, aha I see thanks a lot for the link oh hello robcresswell | 16:49 |
jroll | robcresswell: we have two weeks yet, -intermediary final release must happen the same time as final RCs | 16:49 |
jroll | https://releases.openstack.org/ocata/schedule.html | 16:49 |
jroll | ianychoi: anyway, I will double check with TheJulia and probably make the stable/ocata branch today or monday. thanks for bringing this up :) | 16:50 |
robcresswell | jroll: RC1 should be when you branch stable/ocata; everything past that should be proposed as a backport. | 16:50 |
jroll | robcresswell: -intermediary doesn't do RCs | 16:50 |
robcresswell | Oh, interesting | 16:50 |
robcresswell | Well, I'll shut up then :) | 16:51 |
jroll | heh | 16:51 |
JayF | robcresswell: basically it's "just another release" for us, except we create the stable/ branch and it's capable for backporting | 16:51 |
*** chlong has joined #openstack-ironic | 16:52 | |
*** rob0t_ has joined #openstack-ironic | 16:52 | |
ianychoi | jroll and robcresswell : Thanks. Previously, I18n created stable versions for all the targeted 12 horizon & related projects just after Hard StringFreezes but maybe I18n team needs to be more aware of RC versionings with intermediary cycle model | 16:52 |
*** mtanino has quit IRC | 16:53 | |
*** mtanino has joined #openstack-ironic | 16:53 | |
*** mtanino has quit IRC | 16:53 | |
*** mtanino has joined #openstack-ironic | 16:53 | |
ianychoi | jroll, then what would be the final release date for ironic-ui during Ocata cycle (not exactly just estimation would be also fine) | 16:54 |
ianychoi | ? | 16:54 |
*** rloo has joined #openstack-ironic | 16:54 | |
*** d0ugal is now known as d0ugal|skiing | 16:54 | |
jroll | ianychoi: the plan was for next thursday, but I can do it earlier if TheJulia and ppiela are okay with it | 16:55 |
*** rloo has quit IRC | 16:55 | |
ianychoi | jroll, aha okay thanks! It would be much nice if there will be a stable/ocata branch yep please tell me :) | 16:56 |
ianychoi | Thanks :) | 16:56 |
TheJulia | jroll: I'm fine with branching it now since ppiela indicated to me a few days ago that the rest of the port work would not be ready before the end of next week | 16:56 |
robcresswell | JayF: Yeah, I just didn't realise it didnt follow the usual RC period. I just assumed intermediary releases did that anyway. | 16:56 |
TheJulia | Also, sorry for the delay, was in my xar | 16:56 |
TheJulia | Car | 16:56 |
jroll | TheJulia: no worries on delay, I'll go ahead and do that thing. thanks! | 16:56 |
ianychoi | TheJulia, oh fine. Texting in driving is so unsafe :) | 16:57 |
*** rob0t_ has quit IRC | 16:57 | |
jlvillal | vsaienk0, pas-ha vdrok: On networking- generic-switch. What do you think about just putting the config values into /etc/neutron/neutron.conf ? | 16:59 |
jlvillal | This is due to the grenade issues we are running into with multi-tenant | 17:01 |
jlvillal | vgadiraj, ^^ | 17:01 |
pas-ha | did not though about that yet, as we were just following examples of other ML2 plugins? bu we could try | 17:01 |
jlvillal | pas-ha, Yeah. Having a hard time getting the config file to be used in the 'new' phase of grenade. | 17:01 |
*** rob0t_ has joined #openstack-ironic | 17:02 | |
jlvillal | I'm going to test it to see if it helps with the grenade issue I am having | 17:02 |
pas-ha | it does not want to start q-svc with an extra config file? | 17:02 |
ppiela | TheJulia, jroll: I plan on having the patchsets for the remaining port work uploaded for review today. | 17:02 |
jlvillal | pas-ha, No. Not for the 'new' phase. The 'old' phase is fine. | 17:02 |
jroll | ppiela: oh, do you want to wait for next week to release ocata, then, or is now okay? | 17:03 |
*** yolanda has quit IRC | 17:03 | |
*** athomas has quit IRC | 17:04 | |
ppiela | jroll: I would prefer to wait, can we revisit on Monday? | 17:05 |
jroll | ppiela: sure, I'll put up a WIP patch for now | 17:05 |
dtantsur | ppiela, jroll, usually releases don't happen on Friday anyway | 17:05 |
vdrok | morning jlvillal , I don't think we have a specific reason for putting it where we do now. I see the same structure folder used in networking-cisco, but in networking-arista it's just /etc/neutron. So can be changed I guess. but not sure if just changing folders will help. worth trying :) | 17:05 |
ppiela | jroll: thanks | 17:05 |
jlvillal | vdrok, Thanks. Yeah I will give it a test | 17:06 |
*** yolanda has joined #openstack-ironic | 17:06 | |
ianychoi | One more question: where can I see "cycle-with-intermediary" tag for ironic-ui repository? I cannot see such tag on http://git.openstack.org/cgit/openstack/governance/tree/reference/projects.yaml .. | 17:06 |
dtantsur | ianychoi, I think they moved tags to releases repo | 17:07 |
*** ipukha has joined #openstack-ironic | 17:07 | |
ianychoi | dtantsur, okay I will find in release repo, thanks! | 17:07 |
*** rob0t_ has quit IRC | 17:07 | |
sambetts | dtantsur: we can use IPMI to control the machine if we want, but the IPMI server is like a bolt on API on to our own BMC, so because of that it doesn't seem to advertise its address to the kernel | 17:08 |
dtantsur | oh, I see | 17:09 |
dtantsur | thanks | 17:09 |
*** rob0t_ has joined #openstack-ironic | 17:12 | |
*** amarao has quit IRC | 17:13 | |
*** mjura has quit IRC | 17:14 | |
*** g3ek has quit IRC | 17:15 | |
openstackgerrit | Merged openstack/ironic-inspector-specs master: Spec for retrieving NUMA node information https://review.openstack.org/396147 | 17:16 |
*** chlong has quit IRC | 17:16 | |
*** rob0t_ has quit IRC | 17:17 | |
* sambetts -> heading off | 17:20 | |
sambetts | have a great weekend everyone :D | 17:20 |
sambetts | o/ | 17:20 |
*** sambetts is now known as sambetts|afk | 17:20 | |
*** rob0t_ has joined #openstack-ironic | 17:22 | |
dtantsur | going as well, have a nice weekend | 17:24 |
*** dtantsur is now known as dtantsur|afk | 17:25 | |
*** g3ek has joined #openstack-ironic | 17:26 | |
*** aparnav has joined #openstack-ironic | 17:26 | |
*** rob0t_ has quit IRC | 17:27 | |
*** appu103 has quit IRC | 17:28 | |
*** aparna has joined #openstack-ironic | 17:29 | |
*** cdearborn has joined #openstack-ironic | 17:30 | |
*** derekh has quit IRC | 17:30 | |
*** jkilpatr_ has joined #openstack-ironic | 17:31 | |
*** appu103 has joined #openstack-ironic | 17:31 | |
*** aNuposic has joined #openstack-ironic | 17:32 | |
*** aparnav has quit IRC | 17:32 | |
*** jkilpatr has quit IRC | 17:32 | |
*** rob0t_ has joined #openstack-ironic | 17:34 | |
*** aparna has quit IRC | 17:34 | |
*** trown is now known as trown|lunch | 17:35 | |
NobodyCam | have a good weekend dtantsur|afk sambetts|afk | 17:36 |
*** lucasagomes is now known as lucas-afk | 17:36 | |
NobodyCam | you too lucas-afk :) | 17:36 |
NobodyCam | :p | 17:36 |
lucas-afk | o/ | 17:36 |
lucas-afk | see ya, enjoy the weekend NobodyCam | 17:37 |
lucas-afk | and everyone else :-) | 17:37 |
NobodyCam | :) | 17:37 |
*** amarao has joined #openstack-ironic | 17:37 | |
*** rob0t_ has quit IRC | 17:39 | |
*** rama_y has joined #openstack-ironic | 17:41 | |
*** [1]cdearborn has quit IRC | 17:41 | |
*** vsaienko has joined #openstack-ironic | 17:41 | |
*** ccamacho has quit IRC | 17:43 | |
*** amarao has quit IRC | 17:48 | |
*** rob0t_ has joined #openstack-ironic | 17:49 | |
*** dprince has quit IRC | 17:50 | |
*** mgoddard_ has quit IRC | 17:52 | |
*** EmilienM has quit IRC | 17:56 | |
*** tonyb has quit IRC | 17:56 | |
*** mikal has quit IRC | 17:56 | |
*** Nakato has quit IRC | 17:56 | |
*** patchbot has quit IRC | 17:56 | |
*** hurricanerix has quit IRC | 17:56 | |
*** alex_xu has quit IRC | 17:56 | |
*** mikal has joined #openstack-ironic | 17:56 | |
*** tonyb has joined #openstack-ironic | 17:56 | |
*** devananda has quit IRC | 17:56 | |
*** Madasi has quit IRC | 17:56 | |
*** EmilienM has joined #openstack-ironic | 17:56 | |
*** rob0t_ has quit IRC | 17:56 | |
*** cinerama has quit IRC | 17:56 | |
*** sirushti has quit IRC | 17:56 | |
*** zer0c00l has quit IRC | 17:56 | |
*** hurricanerix has joined #openstack-ironic | 17:56 | |
*** zer0c00l has joined #openstack-ironic | 17:56 | |
*** alex_xu has joined #openstack-ironic | 17:56 | |
*** devananda has joined #openstack-ironic | 17:56 | |
*** cinerama has joined #openstack-ironic | 17:56 | |
*** Nakato has joined #openstack-ironic | 17:57 | |
*** sirushti has joined #openstack-ironic | 17:57 | |
*** patchbot has joined #openstack-ironic | 17:58 | |
*** Madasi has joined #openstack-ironic | 17:58 | |
jlvillal | pas-ha, vdrok vsaienk0: FYI, supposedly MultiConfigParser is deprecated and could be removed in Ocata (though I don't think it will be). https://github.com/openstack/networking-generic-switch/blob/master/networking_generic_switch/config.py#L20 | 18:00 |
*** ralonsoh has quit IRC | 18:02 | |
*** rob0t_ has joined #openstack-ironic | 18:06 | |
*** alex_xu has quit IRC | 18:06 | |
*** alex_xu has joined #openstack-ironic | 18:08 | |
mariojv | is anyone doing the ironicclient work to update the node endpoints to display new driver interfaces? | 18:09 |
mariojv | i know the driver endpoints have work underway, didn't see node work | 18:09 |
*** rob0t_ has quit IRC | 18:10 | |
*** ipukha has quit IRC | 18:11 | |
*** aparnav has joined #openstack-ironic | 18:12 | |
vdrok | mariojv: like this one? https://review.openstack.org/418359 | 18:14 |
patchbot | patch 418359 - python-ironicclient - Add args to CLI 'node-create' for selecting hardwa... (MERGED) | 18:14 |
vdrok | not sure what you mean by node endpoints | 18:15 |
*** appu103 has quit IRC | 18:15 | |
vdrok | as in case of nodes, they are just fields | 18:15 |
TheJulia | Okay, I'm back in front of a keyboard | 18:15 |
TheJulia | What a frustrating morning | 18:16 |
*** sabeen has joined #openstack-ironic | 18:16 | |
vdrok | jlvillal: ok, we'll look into that | 18:16 |
NobodyCam | :( | 18:16 |
TheJulia | ianychoi: I was parked when I was typing into irc ;) | 18:16 |
*** aparnav has quit IRC | 18:17 | |
ianychoi | TheJulia, it seems that you have wonderful driving + parking skills! | 18:17 |
TheJulia | lol! | 18:17 |
vdrok | mariojv: but still, seems like OSC set and unset for these new driver fields are still not done | 18:17 |
TheJulia | ppiela: w/r/t to the port work, awesome! :) | 18:18 |
*** pc_m has quit IRC | 18:18 | |
* vdrok leaves for the weekend | 18:19 | |
*** rob0t_ has joined #openstack-ironic | 18:19 | |
vdrok | good night everyone! | 18:19 |
TheJulia | goodnight vdrok | 18:19 |
openstackgerrit | Anup Navare proposed openstack/ironic master: Allow logical name along with UUID in port creation https://review.openstack.org/405586 | 18:19 |
jlvillal | vdrok, Good night, have a great weekend! | 18:19 |
ianychoi | BTW thanks a lot for today conversation. Now I am also starting communications with magnum-ui and murano-dashboard which are also cycle-with-intermediary but now here is with most active conversation :) | 18:19 |
*** dprince has joined #openstack-ironic | 18:20 | |
ianychoi | I think on early next week I will start to support translation sync job including ironic-ui :) | 18:20 |
TheJulia | ianychoi: I thought translations were already syncing out to ironic-ui | 18:21 |
aNuposic | goodnight vdrok | 18:21 |
ianychoi | TheJulia, yep that's right but it is just for master branch. I18n team targets total 12 horizon & plugin projects to include translated strings on the stable branch as much as possible and I18n PTL needs to create a stable version and translation sync job for stable branches at the same time. | 18:22 |
*** pc_m has joined #openstack-ironic | 18:22 | |
mariojv | vdrok: yeah that was it. have a good weekend! | 18:23 |
ianychoi | Since cycle-with-intermediary model might be dealt with, I think I need to discuss it with release team during PTG | 18:23 |
mariojv | i think that *might* make node show show the right thing, too | 18:23 |
TheJulia | ianychoi: Interesting, because I know, at least for the last release of ironic-ui, we cut the stable branch late and all of a sudden revised translations were showing up for the stable branch. | 18:26 |
*** rob0t_ has quit IRC | 18:26 | |
ianychoi | TheJulia, the main problem is that master branch strings might have differences with a stable branch for a few projects and also translators need to have reviewing period so I18n team has moved to translation on stable branches after Hard StringFreeze | 18:30 |
ianychoi | http://docs.openstack.org/developer/i18n/release_management.html => Hope that it helps. Now I am seeing number 5 but now I realize that the document does not cover cycle-with-intermediary model.. | 18:31 |
*** vsaienko has quit IRC | 18:33 | |
*** jcoufal_ has joined #openstack-ironic | 18:35 | |
*** rob0t_ has joined #openstack-ironic | 18:36 | |
*** tesseract has quit IRC | 18:38 | |
*** jcoufal has quit IRC | 18:38 | |
*** chlong has joined #openstack-ironic | 18:38 | |
*** rob0t_ has quit IRC | 18:41 | |
*** Goneri has quit IRC | 18:43 | |
*** rob0t_ has joined #openstack-ironic | 18:44 | |
*** mgoddard_ has joined #openstack-ironic | 18:45 | |
*** Sukhdev has joined #openstack-ironic | 18:45 | |
TheJulia | ianychoi: Hmm... I would be interested to be in this discussion at the PTG because naturally we don't want to overly burden anyone, and truthfully I want to understand the process/perspective from the i18n side. | 18:46 |
*** trown|lunch is now known as trown | 18:48 | |
*** rob0t_ has quit IRC | 18:49 | |
*** rob0t_ has joined #openstack-ironic | 18:52 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic master: Improve enabled_*_interfaces config help and validation https://review.openstack.org/428219 | 18:54 |
*** sdake has joined #openstack-ironic | 18:56 | |
*** cdaniel is now known as cdaniel-afk | 18:57 | |
*** rob0t_ has quit IRC | 18:57 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost master: Remove my_ip_address from use https://review.openstack.org/428975 | 18:59 |
NobodyCam | TheJulia: oh really. I wouldn't have thought that was it | 19:00 |
*** rob0t_ has joined #openstack-ironic | 19:02 | |
TheJulia | NobodyCam: defaults are calculated out upon role load, if something is still in-flight then.... I bet the exception is hiding under the covers until the variable is acessed | 19:03 |
TheJulia | at least, that is my theory | 19:03 |
NobodyCam | ahh :) | 19:03 |
*** rob0t_ has quit IRC | 19:06 | |
* NobodyCam watches the tests :p | 19:07 | |
ianychoi | TheJulia, yep I don't want your team with more burden. Thanks a lot for such consideration! | 19:09 |
*** chlong has quit IRC | 19:11 | |
*** rob0t_ has joined #openstack-ironic | 19:12 | |
*** jidar has joined #openstack-ironic | 19:12 | |
*** crushil has quit IRC | 19:15 | |
*** rob0t_ has quit IRC | 19:17 | |
TheJulia | NobodyCam: didn't work :\ | 19:17 |
*** jkilpatr_ has quit IRC | 19:18 | |
*** ChubYann has joined #openstack-ironic | 19:19 | |
*** jkilpatr has joined #openstack-ironic | 19:20 | |
*** chlong has joined #openstack-ironic | 19:21 | |
openstackgerrit | Andreas Jaeger proposed openstack/ironic master: Prepare for using standard python tests https://review.openstack.org/428994 | 19:21 |
*** rob0t_ has joined #openstack-ironic | 19:22 | |
openstackgerrit | Andreas Jaeger proposed openstack/ironic stable/newton: Prepare for using standard python tests https://review.openstack.org/428998 | 19:22 |
*** amotoki has joined #openstack-ironic | 19:23 | |
*** vsaienko has joined #openstack-ironic | 19:23 | |
*** milan has quit IRC | 19:23 | |
openstackgerrit | Merged openstack/python-ironicclient master: Do not show chassis_uuid field when it is not specified https://review.openstack.org/427815 | 19:24 |
jroll | if anyone minds a review, 19:23:32 AJaeger | dtantsur|afk, jroll: Could you review https://review.openstack.org/428994 , please? This is a nop until we merge 428995 but I'd like to have it in before ocata branch | 19:24 |
patchbot | patch 428994 - ironic - Prepare for using standard python tests | 19:24 |
jroll | would help out infra | 19:24 |
jroll | thanks TheJulia | 19:25 |
TheJulia | :) | 19:25 |
*** vsaienko has quit IRC | 19:26 | |
TheJulia | I'm a little confused why we're backporting it to stable/newton as well | 19:26 |
jroll | TheJulia: job definitions are mostly branchless | 19:27 |
TheJulia | ahh | 19:27 |
TheJulia | yeah, that they are | 19:27 |
*** rob0t_ has quit IRC | 19:27 | |
*** chlong has quit IRC | 19:27 | |
jroll | we could run -db for older ones but no reason not to backport this and keep project-config "simple" | 19:27 |
openstackgerrit | Andreas Jaeger proposed openstack/ironic stable/mitaka: Prepare for using standard python tests https://review.openstack.org/429000 | 19:27 |
*** amotoki has quit IRC | 19:28 | |
TheJulia | 2017-02-03 19:27:33.429543 | [E] E003: Indent not multiple of 4: ' SELECT 'HERE' from pg_roles where rolname='$DB_USER'")' | 19:28 |
TheJulia | 2017-02-03 19:27:33.429570 | - /home/jenkins/workspace/gate-ironic-pep8-ubuntu-xenial/tools/test-setup.sh : L42 | 19:28 |
jroll | ya | 19:28 |
jroll | I'll take care of it | 19:29 |
TheJulia | 10-4 | 19:29 |
* jroll will forget about it otherwise | 19:30 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic master: Prepare for using standard python tests https://review.openstack.org/428994 | 19:30 |
*** vsaienko has joined #openstack-ironic | 19:33 | |
jroll | huh, how did that pass on newton | 19:34 |
*** e0ne has joined #openstack-ironic | 19:35 | |
openstackgerrit | Andreas Jaeger proposed openstack/ironic-inspector master: Prepare for using standard python tests https://review.openstack.org/429004 | 19:36 |
*** vsaienko has quit IRC | 19:36 | |
openstackgerrit | Andreas Jaeger proposed openstack/ironic-inspector stable/newton: Prepare for using standard python tests https://review.openstack.org/429006 | 19:36 |
openstackgerrit | Andreas Jaeger proposed openstack/ironic-inspector stable/mitaka: Prepare for using standard python tests https://review.openstack.org/429007 | 19:37 |
*** jcoufal has joined #openstack-ironic | 19:37 | |
*** jcoufal_ has quit IRC | 19:37 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost master: Remove my_ip_address from use https://review.openstack.org/428975 | 19:38 |
*** rob0t_ has joined #openstack-ironic | 19:39 | |
openstackgerrit | Merged openstack/python-ironicclient master: Fix cleanup of resources in OSC plugin functional tests https://review.openstack.org/424528 | 19:39 |
TheJulia | jroll: bashate changes? :) | 19:39 |
openstackgerrit | Merged openstack/python-ironicclient master: Fix cleanup of resources in functional tests https://review.openstack.org/424545 | 19:39 |
*** vsaienko has joined #openstack-ironic | 19:39 | |
TheJulia | NobodyCam: ^^^ moved the setup to later, lets see if that changes the behavior in ci testing | 19:40 |
NobodyCam | :) | 19:40 |
*** chlong has joined #openstack-ironic | 19:41 | |
jroll | TheJulia: I can only assume so | 19:43 |
*** rob0t_ has quit IRC | 19:43 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic stable/newton: Ensures that OneView nodes are free for use by Ironic https://review.openstack.org/429009 | 19:44 |
openstackgerrit | Merged openstack/python-ironicclient master: Add negative test-cases for openstack node create command https://review.openstack.org/381943 | 19:45 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic stable/newton: Prepare for using standard python tests https://review.openstack.org/428998 | 19:49 |
openstackgerrit | Peter Piela proposed openstack/ironic-ui master: Fix node validation problem https://review.openstack.org/429012 | 19:50 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic stable/mitaka: Prepare for using standard python tests https://review.openstack.org/429000 | 19:51 |
jroll | JayF: if you wanna pick those stable patches up when you have a few, that'd be awesome | 19:52 |
JayF | jroll: I'm still -2 on that oneview one | 19:52 |
JayF | jroll: even with your repost | 19:52 |
*** jcoufal has quit IRC | 19:53 | |
jroll | JayF: that's a different oneview patch no? | 19:53 |
JayF | jroll: I'm 99% sure that requires a newer versino of oneviewclient, and teh requirements bump is missing from the patch | 19:53 |
JayF | jroll: and even if it was there, it's not allowed | 19:53 |
*** jcoufal has joined #openstack-ironic | 19:53 | |
jroll | JayF: what makes you think it needs a new oneview client? no code that interacts with oneview client is touched | 19:54 |
*** vsaienko has quit IRC | 19:54 | |
JayF | because apparently I'm confused | 19:54 |
JayF | going to take a secondd look at that patch | 19:54 |
jroll | heh | 19:54 |
jroll | JayF: I was mostly referring to the test-setup.sh stuff though :) | 19:54 |
*** rob0t_ has joined #openstack-ironic | 19:55 | |
JayF | jroll: both +2'd, they need to wait to merge tho obviously | 19:55 |
jroll | JayF: yeah, I thought you were in a meeting and master would land by then :) | 19:56 |
JayF | I am king of multitasking | 19:56 |
JayF | and I'm gone at noonish so, yeah | 19:56 |
*** vsaienko has joined #openstack-ironic | 19:56 | |
JayF | lol | 19:56 |
openstackgerrit | Merged openstack/ironic master: Log an actual error message when failed to load new style credentials https://review.openstack.org/428690 | 19:57 |
jroll | ah | 19:59 |
*** rob0t_ has quit IRC | 19:59 | |
* jroll steps out for a bit | 20:02 | |
*** jcoufal has quit IRC | 20:06 | |
*** jcoufal has joined #openstack-ironic | 20:06 | |
*** rob0t_ has joined #openstack-ironic | 20:07 | |
*** rob0t_ has quit IRC | 20:11 | |
vsaienko | jroll, vdrok damn scheduler picked the same node for different instances at the same time http://paste.openstack.org/show/597618/ | 20:12 |
openstackgerrit | Julia Kreger proposed openstack/bifrost master: Remove deprecated nginx_port option https://review.openstack.org/429021 | 20:12 |
openstackgerrit | Julia Kreger proposed openstack/bifrost master: Fixing typo in deprecation note https://review.openstack.org/429022 | 20:12 |
openstackgerrit | Julia Kreger proposed openstack/bifrost master: Remove deprecated test playbooks https://review.openstack.org/429023 | 20:12 |
vsaienko | jroll, vdrok latest multinode job failure https://review.openstack.org/#/c/427675/ | 20:13 |
patchbot | patch 427675 - ironic - [Devstack] fix waiting resources on subnode | 20:13 |
TheJulia | NobodyCam: cinerama some cleanup above as well | 20:13 |
*** ijw has joined #openstack-ironic | 20:16 | |
*** e0ne has quit IRC | 20:17 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic master: WIP/DNM: Testing multinode job https://review.openstack.org/422149 | 20:18 |
*** e0ne has joined #openstack-ironic | 20:19 | |
*** e0ne has quit IRC | 20:19 | |
*** e0ne has joined #openstack-ironic | 20:19 | |
*** Goneri has joined #openstack-ironic | 20:20 | |
*** e0ne has quit IRC | 20:20 | |
*** e0ne has joined #openstack-ironic | 20:20 | |
*** Sukhdev has quit IRC | 20:20 | |
*** e0ne has quit IRC | 20:21 | |
*** e0ne has joined #openstack-ironic | 20:21 | |
*** e0ne has quit IRC | 20:21 | |
*** dsneddon is now known as dsneddon_afk | 20:21 | |
*** e0ne has joined #openstack-ironic | 20:22 | |
*** rob0t_ has joined #openstack-ironic | 20:22 | |
*** jcoufal_ has joined #openstack-ironic | 20:22 | |
*** e0ne has quit IRC | 20:22 | |
*** jcoufal has quit IRC | 20:23 | |
openstackgerrit | Merged openstack/ironic master: Allow duplicate execution of update node DB api method https://review.openstack.org/427673 | 20:24 |
*** jheroux has quit IRC | 20:26 | |
*** hecliunyx has quit IRC | 20:26 | |
*** rajinir has quit IRC | 20:26 | |
*** rob0t_ has quit IRC | 20:26 | |
TheJulia | NobodyCam: so all those other changes, they passed the two jobs that have been misbehaving sporatically, I guess that is a good sign :) | 20:29 |
NobodyCam | :) i'll take a look in a bit | 20:30 |
TheJulia | jroll: shouldn't 428219 have a regenerated sample config? | 20:31 |
openstackgerrit | Mario Villaplana proposed openstack/ironic master: Update api-ref for dynamic drivers https://review.openstack.org/428460 | 20:31 |
mariojv | cc: jroll dtantsur|afk that's ready ^ | 20:31 |
*** rob0t_ has joined #openstack-ironic | 20:32 | |
openstackgerrit | Merged openstack/ironic-inspector master: Trivial: improve logging in FSM transition https://review.openstack.org/427853 | 20:33 |
*** rob0t_ has quit IRC | 20:36 | |
*** sasha21 has quit IRC | 20:40 | |
*** awiddersheim has quit IRC | 20:42 | |
*** rob0t_ has joined #openstack-ironic | 20:42 | |
*** awiddersheim has joined #openstack-ironic | 20:43 | |
*** rob0t_ has quit IRC | 20:47 | |
openstackgerrit | Merged openstack/ironic master: Prepare for using standard python tests https://review.openstack.org/428994 | 20:48 |
*** rob0t_ has joined #openstack-ironic | 20:52 | |
*** dsneddon has joined #openstack-ironic | 20:55 | |
*** rob0t_ has quit IRC | 21:01 | |
*** sdake has quit IRC | 21:06 | |
TheJulia | wow the api-ref update is a bit big :( | 21:22 |
*** mgoddard_ has quit IRC | 21:22 | |
jroll | TheJulia: ah damn, I keep forgetting to do that, lol | 21:23 |
jroll | uno momento | 21:23 |
*** amotoki has joined #openstack-ironic | 21:24 | |
jroll | vsaienko: I'd rather wait for CI to be green on that, if it's okay with you | 21:25 |
jroll | mariojv: holy cow that's big. you rock, thanks for doing that | 21:25 |
*** crushil has joined #openstack-ironic | 21:26 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic master: Improve enabled_*_interfaces config help and validation https://review.openstack.org/428219 | 21:26 |
jroll | TheJulia: bam | 21:26 |
TheJulia | jroll: merci beaucoup | 21:27 |
jroll | de nada :P | 21:27 |
*** amotoki has quit IRC | 21:28 | |
*** amarao has joined #openstack-ironic | 21:32 | |
mariojv | yw jroll | 21:33 |
*** r-daneel has joined #openstack-ironic | 21:33 | |
*** mgoddard_ has joined #openstack-ironic | 21:35 | |
*** crushil has quit IRC | 21:35 | |
*** bfournie has quit IRC | 21:41 | |
*** chlong has quit IRC | 21:44 | |
mariojv | i +1ed that also - looks like JayF had +2 before so someone can probably land it (428219) | 21:49 |
mariojv | since there is also another +2 there | 21:49 |
vsaienko | jroll: CI is green https://review.openstack.org/#/c/427675/ | 21:52 |
patchbot | patch 427675 - ironic - [Devstack] fix waiting resources on subnode | 21:52 |
*** jcoufal has joined #openstack-ironic | 21:53 | |
jroll | vsaienko: +A | 21:53 |
jroll | thanks for that | 21:54 |
*** jcoufal_ has quit IRC | 21:54 | |
vsaienko | jroll thanks! | 21:56 |
jroll | np :) | 21:58 |
*** pester has joined #openstack-ironic | 21:58 | |
*** srobert_ has joined #openstack-ironic | 21:59 | |
*** srobert has quit IRC | 22:02 | |
*** srobert_ has quit IRC | 22:03 | |
*** pester is now known as fxpester | 22:04 | |
*** jcoufal has quit IRC | 22:06 | |
jroll | mariojv: reviewed, lookin good, just a few things | 22:09 |
jroll | mariojv: side note, did you have any issues playing with the new stuff? how (non-)obvious was it? | 22:10 |
mariojv | jroll: no issues except with the client really | 22:10 |
mariojv | it seemed quite intuitive to me | 22:10 |
*** jcoufal has joined #openstack-ironic | 22:10 | |
mariojv | i think the more confusing thing will be understanding classic vs. dynamic drivers | 22:11 |
mariojv | i.e. it'd be better if it was just dynamic | 22:11 |
jroll | what was up with the client, just the default version? | 22:11 |
jroll | yeah | 22:11 |
jroll | I think in pike I'd like to transition everyone to dynamic drivers | 22:11 |
mariojv | no, even when you specify --ironic-api-version=1.31 | 22:12 |
mariojv | node-show doesn't show the interfaces | 22:12 |
jroll | it does for me O_o | 22:12 |
jroll | though, only on master | 22:12 |
jroll | not the latest release | 22:12 |
mariojv | oh, i was pulling down a patched version | 22:13 |
mariojv | might have been done on the wrong base | 22:13 |
jroll | ah probably | 22:13 |
jroll | I think those patches were in separate trees | 22:13 |
mariojv | (patched with the updates to driver show/list/etc) | 22:13 |
jroll | ya :) | 22:13 |
jroll | cool, I'm glad it wasn't a horrible experience | 22:14 |
mariojv | that's this one https://review.openstack.org/#/c/419274/ | 22:14 |
patchbot | patch 419274 - python-ironicclient - Extends driver-list, driver-show supporting new ha... | 22:14 |
aNuposic | mariojv: rama_y: thnx for reviewing, will fix the issues. | 22:14 |
jroll | I just did a last sanity check and it appears to work fine in devstack, with nova and everything | 22:14 |
aNuposic | rama_y: wrt https://review.openstack.org/#/c/405586/6/ironic/api/controllers/v1/port.py@89 name would be None | 22:14 |
patchbot | patch 405586 - ironic - Allow logical name along with UUID in port creation | 22:14 |
mariojv | it made a heck of a lot more sense than reading spec/code (though those weren't too difficult either) | 22:14 |
mariojv | great | 22:14 |
jroll | lol | 22:14 |
jroll | the internals are the weird part, I guess | 22:14 |
mariojv | ya | 22:14 |
mariojv | i'm heading out, have a great weekend ironic o/ | 22:15 |
mariojv | (go patriots) | 22:15 |
jroll | \o | 22:15 |
* jroll walks his dog, may or may not return | 22:15 | |
jroll | if not, see y'all monday :) | 22:15 |
*** Sukhdev has joined #openstack-ironic | 22:16 | |
openstackgerrit | Varun Gadiraju proposed openstack/ironic master: WIP/DNM: Testing grenade multi-tenant experimental job for Tempest failures https://review.openstack.org/403895 | 22:21 |
*** dprince has quit IRC | 22:24 | |
*** trown is now known as trown|outtypewww | 22:25 | |
openstackgerrit | Merged openstack/bifrost master: Allow enroll-dynamic on remote host https://review.openstack.org/411299 | 22:31 |
openstackgerrit | Merged openstack/bifrost master: roles: bifrost-create-vm-nodes: Randomize VM XML file https://review.openstack.org/428716 | 22:32 |
*** sdake has joined #openstack-ironic | 22:32 | |
openstackgerrit | Peter Piela proposed openstack/ironic-ui master: Add support for additional port attributes https://review.openstack.org/429066 | 22:37 |
*** baoli has quit IRC | 22:37 | |
*** Sukhdev has quit IRC | 22:38 | |
*** jcoufal has quit IRC | 22:39 | |
*** owalsh is now known as owalsh-afk | 22:40 | |
*** jcoufal has joined #openstack-ironic | 22:40 | |
*** chlong has joined #openstack-ironic | 22:42 | |
*** amarao has quit IRC | 22:45 | |
*** cdearborn has quit IRC | 22:48 | |
*** jcoufal has quit IRC | 22:50 | |
*** jcoufal has joined #openstack-ironic | 22:51 | |
*** jkilpatr has quit IRC | 22:51 | |
*** jaosorior has quit IRC | 22:52 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic master: WIP/DNM: Testing multinode job https://review.openstack.org/422149 | 22:52 |
*** vsaienko has quit IRC | 23:00 | |
*** mgoddard_ has quit IRC | 23:00 | |
*** jcoufal has quit IRC | 23:06 | |
*** dsneddon has quit IRC | 23:07 | |
*** jcoufal has joined #openstack-ironic | 23:09 | |
*** mgoddard_ has joined #openstack-ironic | 23:10 | |
*** sdake has quit IRC | 23:11 | |
*** dsneddon has joined #openstack-ironic | 23:14 | |
joanna | vdrok: Thank you for the review :) I replied to the comments :) | 23:18 |
*** pmannidi has joined #openstack-ironic | 23:20 | |
*** jcoufal has quit IRC | 23:21 | |
*** jcoufal has joined #openstack-ironic | 23:21 | |
*** amotoki has joined #openstack-ironic | 23:24 | |
*** Sukhdev has joined #openstack-ironic | 23:25 | |
*** jkilpatr has joined #openstack-ironic | 23:25 | |
*** pmannidi has quit IRC | 23:26 | |
*** pmannidi has joined #openstack-ironic | 23:29 | |
*** amotoki has quit IRC | 23:29 | |
openstackgerrit | Merged openstack/ironic master: [Devstack] fix waiting resources on subnode https://review.openstack.org/427675 | 23:29 |
*** mgoddard_ has quit IRC | 23:36 | |
*** adu has joined #openstack-ironic | 23:46 | |
openstackgerrit | Peter Piela proposed openstack/ironic-ui master: Add support for editing Ironic network ports https://review.openstack.org/429111 | 23:49 |
*** adu has quit IRC | 23:55 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!