| *** shakamunyi has joined #openstack-ironic | 00:41 | |
| *** zackf has quit IRC | 00:49 | |
| *** shakamunyi has quit IRC | 00:50 | |
| *** matsuhashi has joined #openstack-ironic | 00:55 | |
| *** nosnos has joined #openstack-ironic | 00:59 | |
| *** zackf has joined #openstack-ironic | 01:09 | |
| *** shakamunyi has joined #openstack-ironic | 01:16 | |
| *** killer_prince has quit IRC | 01:17 | |
| *** shakamunyi has quit IRC | 01:26 | |
| *** shakamunyi has joined #openstack-ironic | 01:53 | |
| *** shakamunyi has quit IRC | 01:57 | |
| *** shakamunyi has joined #openstack-ironic | 01:58 | |
| *** shakamunyi has quit IRC | 02:05 | |
| *** shakamunyi has joined #openstack-ironic | 02:30 | |
| *** shakamunyi has quit IRC | 02:40 | |
| *** shakamunyi has joined #openstack-ironic | 03:07 | |
| *** shakamunyi has quit IRC | 03:17 | |
| *** matsuhashi has quit IRC | 03:19 | |
| *** coolsvap|afk is now known as coolsvap | 03:29 | |
| *** nosnos has quit IRC | 03:33 | |
| *** coolsvap is now known as coolsvap|afk | 03:41 | |
| *** shakamunyi has joined #openstack-ironic | 03:43 | |
| *** coolsvap|afk is now known as coolsvap | 03:48 | |
| *** shakamunyi has quit IRC | 03:53 | |
| *** shakamunyi has joined #openstack-ironic | 04:19 | |
| *** rameshg87 has joined #openstack-ironic | 04:20 | |
| *** matsuhashi has joined #openstack-ironic | 04:21 | |
| *** nosnos has joined #openstack-ironic | 04:21 | |
| lifeless | Hey Ironic folk - https://bugs.launchpad.net/ironic/+bug/1323152 - I've emailed the list, but its currently causing grief in tripleo :(. | 04:27 |
|---|---|---|
| openstackgerrit | Sirushti Murugesan proposed a change to openstack/ironic: Adds support for deploying whole disk images https://review.openstack.org/73054 | 04:27 |
| *** shakamunyi has quit IRC | 04:29 | |
| *** lazy_prince has joined #openstack-ironic | 04:42 | |
| *** shakamunyi has joined #openstack-ironic | 04:55 | |
| *** shakamunyi has quit IRC | 05:05 | |
| *** sabah has joined #openstack-ironic | 05:13 | |
| *** lazy_prince has quit IRC | 05:15 | |
| *** sabah has quit IRC | 05:15 | |
| *** subah has joined #openstack-ironic | 05:15 | |
| *** radsy has quit IRC | 05:17 | |
| *** lazy_prince has joined #openstack-ironic | 05:21 | |
| *** shakamunyi has joined #openstack-ironic | 05:32 | |
| *** shakamunyi_ has joined #openstack-ironic | 05:37 | |
| *** coolsvap is now known as coolsvap|afk | 05:37 | |
| *** shakamunyi has quit IRC | 05:39 | |
| *** shakamunyi_ has quit IRC | 05:39 | |
| *** alexizz has joined #openstack-ironic | 05:40 | |
| *** coolsvap|afk is now known as coolsvap | 05:45 | |
| *** k4n0 has joined #openstack-ironic | 05:45 | |
| *** k4n0 has quit IRC | 06:03 | |
| *** coolsvap is now known as coolsvap|afk | 06:09 | |
| *** coolsvap|afk is now known as coolsvap | 06:20 | |
| mrda | hey ironic: Just trying to understand how our tests work... it appears our ironic tests only use test pecan servers, rather than wsme. Does that mean if I want to do any testing that might be wsme specific, I might need to do it in tempest where we would instantiate a full ironic server (and hence use wsme)? I think devananda might know the answer, but anyone else feel free to jump in :) | 06:28 |
| *** k4n0 has joined #openstack-ironic | 06:32 | |
| *** coolsvap is now known as coolsvap|afk | 06:38 | |
| dtantsur | Morning Ironic! | 06:40 |
| mrda | dtantsur: \o | 06:42 |
| dtantsur | mrda, o/ | 06:43 |
| mrda | dtantsur: where are you based? | 06:44 |
| dtantsur | mrda, Brno, Czech | 06:44 |
| mrda | wow, ok, thanks | 06:45 |
| mrda | I'm Adelaide, Australia | 06:45 |
| dtantsur | cool! | 06:48 |
| dtantsur | so it's quite late at your place now, right? | 06:49 |
| *** coolsvap|afk is now known as coolsvap | 06:49 | |
| mrda | well, 4:20pm | 06:49 |
| mrda | almost time to be afk | 06:50 |
| mrda | :) | 06:50 |
| dtantsur | nearly 9am here :) | 06:51 |
| dtantsur | timezones don't stop to surprise me | 06:52 |
| mrda | dtantsur: and do you stay online for the ironic meeting? Or is that too late? | 06:52 |
| dtantsur | mrda, I do. It's 9pm, so it's a bit troublesome, but possible | 06:52 |
| dtantsur | btw are we having a meeting today? | 06:52 |
| mrda | well, that's better than the 4:30am that it is for me! :) | 06:53 |
| mrda | I think so (hope so, since I'm planning on getting up early) | 06:53 |
| dtantsur | ouch | 06:53 |
| dtantsur | 4:30am sounds hard :( | 06:53 |
| mrda | I should have joined another project ;P | 06:53 |
| * dtantsur is a night owl | 06:54 | |
| *** vinbs has joined #openstack-ironic | 06:57 | |
| *** viktors|afk has joined #openstack-ironic | 07:06 | |
| *** zul has quit IRC | 07:26 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 07:40 | |
| *** zul has joined #openstack-ironic | 07:40 | |
| *** foexle has joined #openstack-ironic | 07:53 | |
| mrda | see y'all tomorrow | 07:56 |
| *** mrda is now known as mrda-away | 07:57 | |
| vinbs | Hello everyone | 08:07 |
| vinbs | I'm new to ironic | 08:08 |
| vinbs | I have an all-in-one node openstack setup | 08:08 |
| vinbs | and also have ironic service setup | 08:08 |
| vinbs | when I add a node using the ironic node-create command, the baremetal node gets added | 08:10 |
| vinbs | and I'm able to power on power off the node | 08:10 |
| vinbs | I'm using the pxe_ipmitool driver | 08:10 |
| vinbs | but the baremetal node is not getting listed in the list of hypervisors | 08:11 |
| vinbs | could anyone point me to what I might be doing wrong? | 08:12 |
| *** jistr has joined #openstack-ironic | 08:13 | |
| *** derekh_ has joined #openstack-ironic | 08:23 | |
| *** lucasagomes has joined #openstack-ironic | 08:33 | |
| *** lsmola has joined #openstack-ironic | 08:39 | |
| *** rameshg87 has quit IRC | 08:41 | |
| dtantsur | vinbs, hello, what exactly do you mean by "list of hypervisors"? | 08:44 |
| *** viktors|afk is now known as viktors | 08:45 | |
| *** rameshg87 has joined #openstack-ironic | 08:48 | |
| viktors | hello All. Can someone look at bug https://bugs.launchpad.net/ironic/+bug/1321787 ? | 08:48 |
| *** vkdrao has joined #openstack-ironic | 08:51 | |
| *** coolsvap is now known as coolsvap|afk | 08:54 | |
| *** max_lobur has joined #openstack-ironic | 08:54 | |
| *** max_lobur has quit IRC | 08:56 | |
| *** max_lobur has joined #openstack-ironic | 08:56 | |
| *** coolsvap|afk is now known as coolsvap | 09:01 | |
| lucasagomes | viktors, whtat's the output of ironic node-validate $NODE? | 09:02 |
| vinbs | dtantsur: nova hypervisor-list | 09:02 |
| lucasagomes | it looks like permission or some network problem | 09:02 |
| viktors | lucasagomes: I've already removed this environment :( .Please inform me, what data should I collect and I'll try to reproduce this some later | 09:04 |
| lucasagomes | viktors, check if the username/[password, keyfile, keydata] is correct at the Node's driver_info attribute | 09:05 |
| *** alexizz has quit IRC | 09:05 | |
| lucasagomes | try to login to the machine using the conventional ssh command to see if it works as well | 09:05 |
| lucasagomes | I don't think it's an ironic problem, it might be some wrong credentials or network problem that is preventing u to access the machine | 09:06 |
| lucasagomes | might worth checking the permissions that the ssh keys have in the filesystem (if you're using keys) | 09:06 |
| dtantsur | vinbs, I'm not sure, why are you expecting to see node in hypervisor list. It's not a hypervisor | 09:07 |
| dtantsur | lucasagomes, hi, I'm a bit stuck, could you help me? | 09:08 |
| lucasagomes | dtantsur, hey sure, if I know how to I will | 09:08 |
| dtantsur | lucasagomes, have a look at commit message here: https://review.openstack.org/#/c/73054 . I can't decide, whether the new parameter should go to driver_info or instance_info | 09:08 |
| dtantsur | maybe it's lack of coffee and getting up at 6am :( | 09:09 |
| viktors | lucasagomes: it looks like some races in Ironic code, IMO. I've got a similar (but not totally same) error with this code http://paste.openstack.org/show/81487/ | 09:10 |
| viktors | lucasagomes: but there are was some cases, when I got totally same error | 09:11 |
| lucasagomes | dtantsur, hmm thinking as well... I would say instance_info because it's related to the image that is being deployed... but this parameter looks like something that nova should set and idk if nova knows about it | 09:12 |
| lucasagomes | dtantsur, if we have a metadata in the image | 09:12 |
| * lucasagomes thinking heh | 09:12 | |
| *** matsuhashi has quit IRC | 09:13 | |
| dtantsur | yeah, that's interesting :) I'll go ahead a leave a review without this point, but we need to decide, I think | 09:13 |
| lucasagomes | dtantsur, +1, so there's a metadata in the image that says it's a whole disk image right? | 09:13 |
| dtantsur | lucasagomes, yep | 09:14 |
| lucasagomes | dtantsur, if so, and nova doesn't look at it, we could make the prepare() method of the pxe set it in the instance_info | 09:14 |
| lucasagomes | when it's downloading the image and converting to raw | 09:14 |
| lucasagomes | dtantsur, but as the instance_info doesn't work right now I think this patch might be safe until we get it merged | 09:14 |
| dtantsur | lucasagomes, I'd better ask the author to rebase on top of your patch. I don't feel like introducing more merge conflicts | 09:15 |
| lucasagomes | dtantsur, cool yeah that works also | 09:15 |
| lucasagomes | viktors, hmm I see, well if you hit it again I think we should try to debug it deeply then | 09:16 |
| dtantsur | but right, I realize I don't understand how that patch works with Nova | 09:16 |
| dtantsur | we need to chat with the author, I guess | 09:16 |
| dtantsur | sirushti, hi, are you around? | 09:16 |
| dtantsur | we're discussing your "whole disk deploy" patch | 09:17 |
| viktors | lucasagomes: ok, I'll ping you in this case | 09:17 |
| lucasagomes | viktors, ack | 09:17 |
| dtantsur | btw, lucasagomes, do we have any kind of "Starting guide" for folks, who install Ironic themselves? I remember reading a lot of devstack source code :( | 09:18 |
| viktors | lucasagomes: no, some later :) | 09:18 |
| sirushti | dtantsur, here | 09:19 |
| dtantsur | sirushti, please have a look at our discussion with Lucas above | 09:19 |
| lucasagomes | dtantsur, you mean just running it to test the services? there's a quick guide in our dev page | 09:20 |
| sirushti | dtantsur, reading now :-) | 09:20 |
| *** matsuhashi has joined #openstack-ironic | 09:20 | |
| lucasagomes | dtantsur, http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#exercising-the-services-locally | 09:20 |
| dtantsur | lucasagomes, no, I mean proper setup, with booting real nodes. I did it once, but it's not a peace-of-cake | 09:20 |
| dtantsur | it's for vinbs | 09:20 |
| lucasagomes | dtantsur, ahh I don't have any tutorial on how to do it completely manually no :( | 09:21 |
| dtantsur | vinbs, as you see ^^^, we don't have anything ready-to-use. My suggestion is: I create/use existing etherpad, where I copy-paste required commands | 09:22 |
| dtantsur | you may need to fix them a bit, updating an etherpad | 09:22 |
| dtantsur | so that we have kind of guide for your case | 09:22 |
| dtantsur | what do you think? | 09:22 |
| *** ams0 has joined #openstack-ironic | 09:23 | |
| *** ams0 has quit IRC | 09:23 | |
| vinbs | dtantsur. sure.. I will need to try out things then | 09:24 |
| vinbs | I will create an etherpad | 09:24 |
| sirushti | dtantsur, If I understand correctly, you're asking how the deploy_disk field is set by nova? It checks the image_properties of the image for kernel/ramdisk | 09:24 |
| dtantsur | vinbs, yeah, give it a meaningful name and send me the link | 09:25 |
| vinbs | dtantsur, sure | 09:26 |
| dtantsur | sirushti, so you're getting this information from image properties? Like setting deploy_disk=True there? | 09:26 |
| sirushti | dtantsur, well, if the image has kernel/ramdisk in its properties dict, then based on that, I set a boolean deploy_disk value. | 09:27 |
| dtantsur | sirushti, first of all, I'm somewhat against doing it if the image has no kernel/ramdisk. That can be just an operator's mistake, and you'll try to do whole-disk deployment | 09:29 |
| dtantsur | sirushti, I'd vote for a special flag 'deploy_disk' or smth in image properties dict | 09:29 |
| sirushti | dtantsur, That was how it was initially during the early stages of the patch. It was later advised to be changed to a 'find out if the image has kernel/ramdisk and figure out what type of image it is'. | 09:31 |
| dtantsur | sirushti, can we really "figure out what type of image it is"? | 09:32 |
| *** matsuhashi has quit IRC | 09:33 | |
| dtantsur | sirushti, also, could you point me to this particularly comment? I can't find it | 09:34 |
| sirushti | dtantsur, type of image in the sense - Disk or Partition Image. Well, I think it's an assumption that has to be made. If the operator fails to register a kernel/ramdisk along with the image, it's considered a whole disk image. | 09:34 |
| sirushti | dtantsur, Let me try and find it | 09:34 |
| dtantsur | yes please | 09:34 |
| *** matsuhas_ has joined #openstack-ironic | 09:38 | |
| sirushti | dtantsur, https://etherpad.openstack.org/p/icehouse-ironic-windows-support - here | 09:38 |
| *** rameshg87 has quit IRC | 09:40 | |
| sirushti | dtantsur, do you think there is a hard necessity to set a glance tag to determine if its a disk/partition image? | 09:43 |
| *** igor_ has joined #openstack-ironic | 09:50 | |
| dtantsur | aha, what I see here: | 09:51 |
| dtantsur | 1. full disk doesn't imply local boot <-- what do you think on this? | 09:51 |
| dtantsur | 2. Full disk controlled by a tag in glance <-- nothing about having kernel/ramdisk | 09:52 |
| dtantsur | 3. full disk image cannot be deployed to a flavor that has ephemeral or swap sizes <-- do you check it? | 09:52 |
| dtantsur | sirushti, ^^^ | 09:53 |
| *** shakamunyi has joined #openstack-ironic | 10:01 | |
| *** coolsvap is now known as coolsvap|afk | 10:09 | |
| *** rameshg87 has joined #openstack-ironic | 10:13 | |
| *** shakamunyi has quit IRC | 10:15 | |
| dtantsur | lucasagomes, can we have this particular change: https://review.openstack.org/#/c/86092/15/ironic/conductor/utils.py split and be merged asap? I think it should be more than useful for say agent driver | 10:20 |
| dtantsur | it will also allow you make all management patches independent of each other | 10:21 |
| *** rameshg87 has left #openstack-ironic | 10:21 | |
| *** rameshg87 has joined #openstack-ironic | 10:22 | |
| lucasagomes | dtantsur, I bet we could, but this patch is the first of the series wouldn't be easier to get it merged? (I need to rebase it) | 10:22 |
| lucasagomes | dtantsur, I will rebase it and try to poke the guys to take a look at it | 10:22 |
| dtantsur | lucasagomes, well, to me the whole series is relatively easy to merge :) thanks | 10:22 |
| *** Sirushti_afk has joined #openstack-ironic | 10:25 | |
| *** SpamapS has quit IRC | 10:25 | |
| *** SpamapS has joined #openstack-ironic | 10:26 | |
| *** SpamapS has quit IRC | 10:26 | |
| *** SpamapS has joined #openstack-ironic | 10:26 | |
| Sirushti_afk | dtantsur, sorry, brb, in a meeting | 10:27 |
| dtantsur | ok | 10:27 |
| *** Sirushti_afk has quit IRC | 10:31 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove 'node' parameter from the Console and Rescue interfaces https://review.openstack.org/91097 | 10:32 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove 'node' parameter from the Power interface https://review.openstack.org/91093 | 10:32 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove 'node' parameter from the Deploy interface https://review.openstack.org/91062 | 10:32 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove 'node' parameter from the validate() methods https://review.openstack.org/91104 | 10:32 |
| *** shakamunyi has joined #openstack-ironic | 10:33 | |
| *** vinbs has quit IRC | 10:42 | |
| *** vinbs has joined #openstack-ironic | 10:43 | |
| *** martyntaylor has joined #openstack-ironic | 10:55 | |
| *** martyntaylor has left #openstack-ironic | 10:57 | |
| sirushti | dtantsur, back, free now? | 10:58 |
| dtantsur | sirushti, will be in 10 minutes :) | 10:58 |
| sirushti | dtantsur, cool :-) | 10:59 |
| *** max_lobur1 has joined #openstack-ironic | 11:09 | |
| *** max_lobur has quit IRC | 11:10 | |
| *** rameshg87 has quit IRC | 11:10 | |
| dtantsur | sirushti, here | 11:13 |
| *** boris-42 has quit IRC | 11:13 | |
| *** rameshg87 has joined #openstack-ironic | 11:14 | |
| sirushti | dtantsur, okay, so after re-reading the spec, I probably misinterpreted the glance tag bit | 11:14 |
| dtantsur | sirushti, left a comment, repeating what I fetched from the etherpad | 11:17 |
| *** vkdrao has quit IRC | 11:18 | |
| dtantsur | sirushti, also we discussed here, whether deploy_disk field should be in driver properties or in instance properties | 11:19 |
| sirushti | dtantsur, just looking at the instance_info patch and scrolled back | 11:21 |
| *** boris-42 has joined #openstack-ironic | 11:22 | |
| dtantsur | sirushti, I suspect it should be in instance_info | 11:23 |
| sirushti | dtantsur, yes, same here, because different deploys can have different images being deployed to them? | 11:24 |
| sirushti | dtantsur, so anything that changes per deploy should probably go to instance_info | 11:24 |
| sirushti | no? | 11:24 |
| dtantsur | sirushti, probably. E.g. image_source will be in instance_info | 11:24 |
| dtantsur | I think so, yes | 11:24 |
| sirushti | dtantsur, so IIUC, deploy_disk based on glance tag, and that's put into instance_info | 11:29 |
| dtantsur | yes, seems correct | 11:29 |
| dtantsur | to get proper instance_info you may want to rebase on top of lucasagomes' patch | 11:29 |
| sirushti | cool, I'll do that then :-) thanks | 11:31 |
| lucasagomes | :) | 11:31 |
| *** lucasagomes is now known as lucas-hungry | 11:31 | |
| sirushti | dtantsur, any comments on localboot? It seems like certain images require certain a certain boot order. There's a problem of the user himself setting the boot order. Do we restrict the user from changing the boot order? | 11:32 |
| dtantsur | re localboot, I think it should be optional. But! I'm ok with not making it optional in this patch | 11:33 |
| dtantsur | Of course you should get opinion of other folks here on this | 11:34 |
| dtantsur | but to me you may start with localboot = on, but keep in mind, it should be eventually optional and configurable | 11:34 |
| dtantsur | (that't my interpretation of the etherpad. I personally don't know use cases for whole-disk-image w/o localboot) | 11:34 |
| dtantsur | lucas-hungry, ^^^ ? | 11:35 |
| sirushti | dtantsur, hmm, it seems weird however to have a 'pxe driver' set localboot for any image. I'd assume that should be managed elsewhere but not too aware about this. | 11:36 |
| sirushti | dtantsur, will dig in more | 11:36 |
| sirushti | dtantsur, thanks again :-) | 11:39 |
| dtantsur | ur welcome :) | 11:40 |
| *** jcoufal has joined #openstack-ironic | 11:43 | |
| *** ifarkas has joined #openstack-ironic | 11:43 | |
| *** jcoufal has quit IRC | 11:43 | |
| *** lazy_prince has quit IRC | 11:46 | |
| *** nosnos has quit IRC | 12:00 | |
| openstackgerrit | Sandhya Balakrishnan proposed a change to openstack/ironic: Update Ironic Guide post landing for 569c0f395a9 https://review.openstack.org/94604 | 12:04 |
| *** dtantsur is now known as dtantsur|lunch | 12:07 | |
| *** matsuhas_ has quit IRC | 12:08 | |
| *** k4n0 has quit IRC | 12:10 | |
| *** subah has quit IRC | 12:14 | |
| *** vinbs_ has joined #openstack-ironic | 12:15 | |
| *** lazy_prince has joined #openstack-ironic | 12:17 | |
| *** vinbs has quit IRC | 12:18 | |
| *** vinbs has joined #openstack-ironic | 12:20 | |
| *** vinbs_ has quit IRC | 12:21 | |
| *** igordcard has joined #openstack-ironic | 12:22 | |
| *** matsuhashi has joined #openstack-ironic | 12:28 | |
| *** igordcard has quit IRC | 12:29 | |
| *** rloo has joined #openstack-ironic | 12:33 | |
| *** lucas-hungry is now known as lucasagomes | 12:35 | |
| *** rloo has quit IRC | 12:37 | |
| *** rloo has joined #openstack-ironic | 12:37 | |
| *** lazy_prince has quit IRC | 12:37 | |
| lucasagomes | dtantsur|lunch, sirushti yeah local boot should be configurable | 12:37 |
| openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: Add IloDriver and its utils https://review.openstack.org/89500 | 12:38 |
| *** shausy has joined #openstack-ironic | 12:45 | |
| viktors | lucasagomes: ping | 12:48 |
| lucasagomes | viktors, pong | 12:48 |
| lucasagomes | viktors, did you reproduce that problem? | 12:49 |
| viktors | lucasagomes: yes, I've got same bug just now. | 12:49 |
| viktors | lucasagomes: waht should I do for debug? | 12:49 |
| lucasagomes | viktors, right... what's the output of "ironic node-validate $NODE" ? | 12:49 |
| lucasagomes | what does the power interface says? | 12:50 |
| viktors | lucasagomes: deploy | False | Can not validate PXE bootloader. The following parameters were not passed to ironic: ['pxe_root_gb', 'pxe_image_source', 'pxe_deploy_kernel', 'pxe_deploy_ramdisk'] | 12:50 |
| lucasagomes | the power interface | 12:50 |
| lucasagomes | it says True?! | 12:51 |
| viktors | yes | 12:51 |
| lucasagomes | the deploy is ok, because this parameters will be passed via the nova driver | 12:51 |
| lucasagomes | right so the ssh seems to be fine | 12:51 |
| viktors | lucasagomes: what's next? | 12:51 |
| lucasagomes | viktors, hmmm checking the error in the bug ticket 1 sec | 12:55 |
| viktors | lucasagomes: sure | 12:55 |
| lucasagomes | viktors, what version of paramiko ur using? | 12:56 |
| lucasagomes | viktors, seems that something similar was fixed on the version 1.11, https://github.com/paramiko/paramiko/commit/e1851788768b5132181690e5ab03d4d65c466e42 | 12:57 |
| lucasagomes | viktors, another way to work it around, seems greenthreads doesn't seems to work nicely with paramiko | 12:57 |
| lucasagomes | maybe would be to have it in a subprocess? | 12:58 |
| *** ifarkas has quit IRC | 12:58 | |
| viktors | lucasagomes: I'm a little confused, that nobody got this error before | 12:59 |
| lucasagomes | viktors, yeah, well ssh driver is only used for tests so | 12:59 |
| viktors | >>> paramiko.__version__ | 13:00 |
| viktors | '1.14.0' | 13:00 |
| lucasagomes | https://answers.launchpad.net/nova/+question/190171 | 13:00 |
| lucasagomes | but no answers :( | 13:00 |
| viktors | yes, it looks similar | 13:00 |
| lucasagomes | viktors, ah https://bugs.launchpad.net/sahara/+bug/1212341 | 13:01 |
| lucasagomes | viktors, https://review.openstack.org/#/c/45716/ | 13:01 |
| *** zackf has quit IRC | 13:01 | |
| *** dtantsur|lunch is now known as dtantsur | 13:01 | |
| viktors | lucasagomes: oh :) | 13:02 |
| * viktors looking at the patch | 13:03 | |
| *** matsuhashi has quit IRC | 13:06 | |
| *** matsuhashi has joined #openstack-ironic | 13:06 | |
| *** shakamunyi has quit IRC | 13:07 | |
| viktors | lucasagomes: thanks for your help! | 13:08 |
| lucasagomes | viktors, ah no problem, thank you for digging into the problem :) | 13:08 |
| *** matsuhas_ has joined #openstack-ironic | 13:10 | |
| *** matsuhashi has quit IRC | 13:10 | |
| dtantsur | rameshg87, ping re https://review.openstack.org/#/c/89500 | 13:13 |
| *** igordcard has joined #openstack-ironic | 13:13 | |
| rameshg87 | hi dtantsur | 13:14 |
| dtantsur | rameshg87, hi! I just wondering, whether you should now merge two iLo patches together | 13:15 |
| dtantsur | https://review.openstack.org/#/c/89500 no longer makes much sense on it's own | 13:15 |
| dtantsur | I don't think you'll have a separate Deploy driver for your hardware, right? | 13:15 |
| rameshg87 | dtantsur, yes we have a separate deploy driver | 13:16 |
| dtantsur | wow | 13:16 |
| dtantsur | I see. And still I suggest merging https://review.openstack.org/#/c/89500 https://review.openstack.org/#/c/90294/ | 13:16 |
| dtantsur | it's hard to reason e.g. about common code without anything actually using it | 13:17 |
| rameshg87 | dtantsur: it will use the virtual media to boot the proliant server for both deploy and boot - aimed at people who don't want to use pxe | 13:17 |
| rameshg87 | dtantsur: we will submit the design spec once it is ready, most probably this week | 13:17 |
| rameshg87 | dtantsur, i agree, there isn't anything in there in the common code.. | 13:18 |
| dtantsur | ok, let Deploy be a separate spec and separate patch | 13:18 |
| rameshg87 | dtantsur, so we will need to abort the ongoing 2 reviews and merge it into a separate review ? | 13:18 |
| dtantsur | rameshg87, ok, abandon https://review.openstack.org/#/c/90294/ and update https://review.openstack.org/#/c/89500 with all changes | 13:19 |
| rameshg87 | dtantsur: i mean should we raise a new review ? | 13:19 |
| dtantsur | I don't want to lose history in the first patch | 13:19 |
| rameshg87 | dtantsur, yeah seems fine to me .. | 13:20 |
| dtantsur | thanks | 13:20 |
| rameshg87 | dtantsur, will do that :-) | 13:20 |
| rameshg87 | thanks | 13:20 |
| NobodyCam | good morning Ironic | 13:21 |
| NobodyCam | woo hoo coffee is ready | 13:22 |
| lucasagomes | NobodyCam, morning! | 13:22 |
| lucasagomes | NobodyCam, how was ur Towel Day? | 13:23 |
| NobodyCam | towel day? humm need to look at one up | 13:24 |
| NobodyCam | good morning lucasagomes | 13:24 |
| lucasagomes | NobodyCam, heh yeah may 25 | 13:25 |
| NobodyCam | heavy reains here in NE .. the A/c over the bed started to leak last night so I need to get up on hte roof today and seal it :-p | 13:26 |
| lucasagomes | ouch | 13:26 |
| NobodyCam | :-p | 13:26 |
| *** kaizer has joined #openstack-ironic | 13:26 | |
| lucasagomes | here it's also raining, it rained almost the whole week last week | 13:27 |
| lucasagomes | today's raining as well :( | 13:27 |
| lucasagomes | but light rain | 13:27 |
| kaizer | Hi guys! | 13:27 |
| lucasagomes | so not a problem | 13:27 |
| dtantsur | morning NobodyCam :) | 13:27 |
| dtantsur | kaizer, hi | 13:28 |
| NobodyCam | morning dtantsur, and kaizer | 13:28 |
| dtantsur | NobodyCam, how is your problem with devstack? | 13:28 |
| NobodyCam | huh, devstack? | 13:29 |
| *** igordcard has quit IRC | 13:29 | |
| NobodyCam | I'm still poking folks for 93133 | 13:29 |
| NobodyCam | poke poke | 13:30 |
| NobodyCam | lol | 13:30 |
| dtantsur | NobodyCam, I mean, you were unable to validate https://review.openstack.org/#/c/92625 :) | 13:30 |
| kaizer | My team will soon start looking at Ironic, i would like to know if its yet in a kind of prototype version or its really used in production? | 13:31 |
| dtantsur | 93133? Oh, it still was not merged... | 13:31 |
| NobodyCam | :-p | 13:33 |
| *** shakamunyi has joined #openstack-ironic | 13:34 | |
| lucasagomes | kaizer, we still incubated and we are planning to get integrated at the end of this release and then i would say we are production ready | 13:36 |
| lucasagomes | kaizer, but, you already can use ironic to deploy machines and all, so it might be good to start experimenting it since now | 13:36 |
| dtantsur | NobodyCam, left 1 non-critical suggestion at 93133, otherwise lgtm | 13:39 |
| dtantsur | NobodyCam, could you retry https://review.openstack.org/#/c/92625 ? | 13:40 |
| dtantsur | (or just believe check-tempest-dsvm-virtual-ironic ;) | 13:40 |
| dtantsur | also lucasagomes, NobodyCam, could you also have a look at https://review.openstack.org/#/c/94371/ ? It haven't had any reviews for ~week :( | 13:42 |
| lucasagomes | dtantsur, :( sure will do, I'm just trying to finish the spec for instance_info here quickly and then I will review it | 13:43 |
| dtantsur | ack, thank you | 13:43 |
| dtantsur | lucasagomes, btw, will you mention fields for whole-disk deployment in your spec? | 13:44 |
| lucasagomes | dtantsur, idk if I can, cause it's not merged yet | 13:44 |
| *** Alexei_9871 has joined #openstack-ironic | 13:44 | |
| dtantsur | hmm, I thought spec has to go before code | 13:44 |
| lucasagomes | dtantsur, but he can rebase his patch on mine and add a "Related to blueprint" rag | 13:44 |
| lucasagomes | tag* | 13:44 |
| kaizer | lucasagomes, thanks! I am also looking at provisioning Linux and Windows with Ironic. I know there was some issues with Windows, do you have a status on what is working and what is not?! | 13:45 |
| *** rameshg87 has left #openstack-ironic | 13:45 | |
| lucasagomes | dtantsur, hmm yeah that's a good point... | 13:45 |
| dtantsur | kaizer, https://etherpad.openstack.org/p/icehouse-ironic-windows-support | 13:45 |
| lucasagomes | maybe I should include it ? hmm | 13:45 |
| dtantsur | tl;dr: it's been worked on, bot ready yet | 13:45 |
| dtantsur | s/bot/not/ | 13:45 |
| lucasagomes | kaizer, yup, for windows we have a patch waiting for review | 13:45 |
| *** shakamunyi has quit IRC | 13:47 | |
| kaizer | dtantsur, lucasagomes : Thanks for fast answers, we might eventually contribute also on our side since we need this working ASAP! | 13:48 |
| dtantsur | kaizer, this is patch in question: https://review.openstack.org/#/c/73054/ | 13:49 |
| dtantsur | it should be the 1st step to supporting Windows | 13:49 |
| lucasagomes | kaizer, np, awesome! Thank you for that, we are really looking into more people to contribute with code and also reviews | 13:49 |
| rloo | NobodyCam, others: is there an ironic meeting today? (Memorial Day in the US) | 13:50 |
| *** vinbs_ has joined #openstack-ironic | 13:50 | |
| *** vinbs has quit IRC | 13:51 | |
| *** vinbs_ is now known as vinbs | 13:52 | |
| NobodyCam | rloo: I believe there is(at least I was planning on it) | 13:52 |
| dtantsur | rloo, hi. I was also planning to attend. | 13:53 |
| rloo | NobodyCam: ok thx. I am thinking of taking the day off but will try to attend :-) | 13:53 |
| NobodyCam | I never said I wasn't taking the day off..lol | 13:53 |
| NobodyCam | hehehhe | 13:53 |
| NobodyCam | :-p | 13:53 |
| rloo | Well, it isn't a holiday for me, but thought I'd take a vacation day ;) | 13:54 |
| NobodyCam | :) | 13:54 |
| NobodyCam | and good morning rloo :) | 13:54 |
| rloo | heh, morning NobodyCam! | 13:55 |
| lucasagomes | rloo, morning :D | 13:58 |
| NobodyCam | dtantsur: That logging was suggested by dev a, see his comment on patch 29 (I think, lost count of the patches) :-p | 13:59 |
| NobodyCam | lucasagomes: speaking of, can you take a look at how I worked the passing of node.uuid in? | 14:01 |
| *** vinbs has quit IRC | 14:02 | |
| max_lobur1 | morning Ironic! | 14:02 |
| *** urulama has joined #openstack-ironic | 14:02 | |
| NobodyCam | morning max_lobur1 :) | 14:02 |
| lucasagomes | NobodyCam, ? | 14:02 |
| max_lobur1 | we're back from the vacation US trip | 14:02 |
| *** urulama has left #openstack-ironic | 14:02 | |
| lucasagomes | I'm a bit lost sorry | 14:02 |
| lucasagomes | max_lobur1, morning | 14:02 |
| NobodyCam | lucasagomes: on 93133, | 14:02 |
| lucasagomes | NobodyCam, right, one sec I will finish wrting the spec for the instance_info before looking at other patches otherwise I get a bit confused | 14:03 |
| NobodyCam | :) | 14:03 |
| *** shakamunyi has joined #openstack-ironic | 14:04 | |
| rloo | morning lucasagomes! | 14:07 |
| NobodyCam | wow glad we made it through Roswell, they flooded last night :-p | 14:07 |
| rloo | NobodyCam: just re-reviewed 93133. The comments aren't quite correct. | 14:08 |
| rloo | NobodyCam: are you sure the aliens didn't get you? | 14:09 |
| rloo | max_lobur1: how was NY? | 14:10 |
| dtantsur | NobodyCam, what he suggested, as I understand, was separating catching exceptions from blockdev and dd | 14:10 |
| dtantsur | NobodyCam, what I suggest is to make sure we have real exception message in logs | 14:10 |
| NobodyCam | rloo: ok I'll update, after lucas has a look. | 14:10 |
| rloo | dtantsur, NobodyCam: I added a comment related to that try/except, just to confuse the issue. | 14:11 |
| dtantsur | NobodyCam, the easiest way is to move try..except.. there (while still keeping 2 try blocks separate) | 14:11 |
| NobodyCam | dtantsur: :) ack will push up another rev in bit... want to get everyone to look at it first :) | 14:12 |
| dtantsur | rloo, oh, that's true! | 14:12 |
| dtantsur | max_lobur1, morning! | 14:12 |
| max_lobur1 | rloo: awesome | 14:12 |
| * max_lobur1 preparing some photos to share | 14:13 | |
| *** max_lobur1 is now known as max_lobur | 14:13 | |
| NobodyCam | rloo: just fyi: I did not want any of this to raise a execption as that would block the deploy and as Adam pointed out on rev #1 | 14:21 |
| NobodyCam | the wipe may fail | 14:21 |
| NobodyCam | But I can redo to catch a error differently | 14:22 |
| rloo | NobodyCam: yeah, no exception 'at the end', but I was wondering whether having the get_dev_block_size raise an exception (and caught by destroy_mbr_gpt) might make things easier to understand. | 14:22 |
| NobodyCam | that I can do :) | 14:23 |
| rloo | NobodyCam: my comment was wrt deva and dtantsur's comments. I was fine with the previous stuff ;) | 14:24 |
| NobodyCam | :) | 14:24 |
| *** rloo is now known as rloo_away | 14:29 | |
| NobodyCam | rloo_away: will you be back | 14:29 |
| NobodyCam | I had a question on it only clears 512 bytes of the GPT table, not the entire table. comment | 14:30 |
| rloo_away | NobodyCam: yes, I'll be back later. Haven't left yet though. What's your question? | 14:30 |
| rloo_away | the entire gpt table is bigger than 512 bytes. i think you're erasing 1024 bytes, right? | 14:30 |
| NobodyCam | :-p i am wipping 1024 bytes so I not quite sure the working your looking for here | 14:31 |
| NobodyCam | s/working/wording/ | 14:31 |
| rloo_away | NobodyCam: https://wiki.archlinux.org/index.php/GUID_Partition_Table#GUID_Partition_Table | 14:31 |
| max_lobur | photos (from my cell only). There are a few from Georgia Aquarium, Manhattan, NY Intrepid NAVY museum, Museum of glass in Corning, Niagara falls, Watkins Glen State Park, Seneca Lake | 14:31 |
| rloo_away | NobodyCam: I'd be fine with 'part of the GPT table'. | 14:31 |
| max_lobur | https://www.dropbox.com/sh/72rk601vps6erju/AABua0BaLYiWHCBP15iS46-4a | 14:32 |
| NobodyCam | :) | 14:32 |
| NobodyCam | ok the 512 threw me off | 14:32 |
| NobodyCam | TY | 14:32 |
| rloo_away | NobodyCam: otherwise, someone might think you're erasing the entire GPT table which isn't true. | 14:32 |
| NobodyCam | :) | 14:32 |
| NobodyCam | ack | 14:32 |
| rloo_away | max_lobur: thx for photos. I see they let you into Canada for a few hours ;) | 14:35 |
| max_lobur | hehe | 14:35 |
| max_lobur | actually they don't. Instead they changed the rules and now you may get into that boat without Canadian visa :) | 14:37 |
| NobodyCam | brb | 14:37 |
| *** matsuhas_ has quit IRC | 14:37 | |
| rloo_away | max_lobur: cool. Guess you're OK as long as you don't try to swim to land or something! | 14:38 |
| max_lobur | yea :D I think no one will want to swim there once he see that flow | 14:43 |
| NobodyCam | brb | 14:54 |
| *** shausy has quit IRC | 14:58 | |
| *** viktors is now known as viktors|afk | 15:16 | |
| *** shakamunyi has quit IRC | 15:16 | |
| NobodyCam | leak sealed (I hope) :-p | 15:19 |
| *** zackf has joined #openstack-ironic | 15:28 | |
| *** igordcard has joined #openstack-ironic | 15:31 | |
| *** zackf has quit IRC | 15:33 | |
| *** foexle has quit IRC | 15:38 | |
| *** igordcard has quit IRC | 15:38 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Nova Ironic Driver to not get deploy r&k from flavor https://review.openstack.org/95551 | 15:40 |
| devananda | g'morning, all | 15:45 |
| lucasagomes | devananda, morning | 15:45 |
| lucasagomes | devananda, the ironic-specs repo is broken :( can't submit a new review and the patches approved are not going through | 15:45 |
| NobodyCam | happy Memorial Day devananda | 15:45 |
| lucasagomes | I pinged the guys on infra they r looking at it now | 15:46 |
| devananda | lucasagomes: ack | 15:46 |
| NobodyCam | lucasagomes: I think that is from the name change | 15:46 |
| NobodyCam | but not sure | 15:46 |
| lucasagomes | NobodyCam, yeah it's | 15:46 |
| *** rloo_away has quit IRC | 15:47 | |
| lucasagomes | apparently zuul is not picking up the jobs | 15:47 |
| lucasagomes | fungi and clarkb are looking at it right now | 15:47 |
| NobodyCam | :) | 15:47 |
| devananda | did you guys see lifeless' email re: the virtual-ironic job being broken recently? | 15:48 |
| lucasagomes | devananda, nop | 15:49 |
| lucasagomes | lemme check | 15:50 |
| lucasagomes | devananda, https://review.openstack.org/#/c/91062/ seems grand | 15:50 |
| NobodyCam | hum Unable to enable dhcp for uuid | 15:50 |
| lucasagomes | check-tempest-dsvm-virtual-ironic SUCCESS in 28m 54s | 15:50 |
| * lucasagomes check the logs | 15:51 | |
| devananda | indeed | 15:51 |
| devananda | so the job isn't broken all the time | 15:51 |
| *** BadCub has joined #openstack-ironic | 15:51 | |
| lucasagomes | devananda, new release of the python-ironicclient | 15:59 |
| lucasagomes | ops | 15:59 |
| lucasagomes | we need a new release of the* | 15:59 |
| lucasagomes | after https://review.openstack.org/#/c/90040/ << this shouldn't have be merged without a release really | 15:59 |
| lucasagomes | we also will need to pin the ironicclient version in the requirements to the new version | 16:00 |
| lucasagomes | NobodyCam, also if u get a time can you look at #91062, pretty much rebase and commit message updated | 16:01 |
| NobodyCam | lucasagomes: ack | 16:02 |
| lucasagomes | devananda, another thing (sorry for the spam heh), you might want to unblock this: https://review.openstack.org/#/c/89301/ after what we decided in the summit | 16:03 |
| *** igor_ has quit IRC | 16:05 | |
| *** igor_ has joined #openstack-ironic | 16:06 | |
| openstackgerrit | Jeremy Stanley proposed a change to openstack/ironic-specs: Updated .gitreview file for repo rename https://review.openstack.org/95295 | 16:06 |
| *** romcheg has joined #openstack-ironic | 16:07 | |
| *** romcheg has left #openstack-ironic | 16:08 | |
| *** romcheg has joined #openstack-ironic | 16:08 | |
| *** romcheg has left #openstack-ironic | 16:08 | |
| *** romcheg has joined #openstack-ironic | 16:09 | |
| *** romcheg has left #openstack-ironic | 16:09 | |
| *** romcheg has joined #openstack-ironic | 16:09 | |
| *** romcheg has left #openstack-ironic | 16:09 | |
| devananda | lucasagomes: it looks like 90040 didn't break any tests -- why's that? | 16:09 |
| devananda | lucasagomes: re: 89301 - yes, thanks for the reminder | 16:09 |
| lucasagomes | devananda, I'm not sure, it broke my local tests | 16:09 |
| lucasagomes | AttributeError: 'module' object has no attribute 'ConnectionRefused' | 16:10 |
| *** romcheg has joined #openstack-ironic | 16:10 | |
| lucasagomes | but yeah... odd enough it didn't break the other tests, lemme investigate | 16:10 |
| *** igor_ has quit IRC | 16:10 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/python-ironicclient: Added argcomplete lib to enable cli completition https://review.openstack.org/65865 | 16:11 |
| devananda | lucasagomes: also on 89301, shouldn't there be a unit test for the new CLI method? | 16:11 |
| *** romcheg has left #openstack-ironic | 16:11 | |
| *** romcheg has joined #openstack-ironic | 16:11 | |
| *** romcheg has left #openstack-ironic | 16:11 | |
| lucasagomes | devananda, so (shame on us really) we don't have any tests to the _shell stuff in the client :( | 16:13 |
| *** romcheg has joined #openstack-ironic | 16:14 | |
| *** romcheg has left #openstack-ironic | 16:14 | |
| lucasagomes | there's a bug open about it somewhere | 16:14 |
| openstackgerrit | Christian Berendt proposed a change to openstack/python-ironicclient: WIP: replace dict.iteritems() with six.iteritems(dict) https://review.openstack.org/95562 | 16:14 |
| lucasagomes | devananda, https://bugs.launchpad.net/python-ironicclient/+bug/1235431 | 16:14 |
| lucasagomes | it was assigned to me urgh | 16:15 |
| lucasagomes | well I might take a look at it when I find some time, removing the assign for now | 16:15 |
| devananda | :) | 16:15 |
| devananda | lucasagomes: the priority on that seems misaligned | 16:16 |
| lucasagomes | devananda, yeah, true, medium would be enough | 16:16 |
| devananda | lucasagomes: https://wiki.openstack.org/wiki/BugTriage#Task_2:_Prioritize_confirmed_bugs_.28bug_supervisors.29 | 16:16 |
| lucasagomes | it's not a big deal because the commands in the shell are pretty straight forward | 16:17 |
| *** Alexei_9871 has left #openstack-ironic | 16:17 | |
| lucasagomes | but needs tests anyway | 16:17 |
| devananda | it's really Low | 16:17 |
| devananda | as it's not preventing any feature at all | 16:17 |
| lucasagomes | oh that's a diff description of the one I read before | 16:17 |
| devananda | oh? | 16:17 |
| devananda | ah: https://wiki.openstack.org/wiki/Bugs | 16:17 |
| lucasagomes | yeah | 16:18 |
| devananda | well, that guideline results in the same priority IMO | 16:18 |
| NobodyCam | brb | 16:18 |
| lucasagomes | devananda, wishlist seems adequate | 16:18 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/python-ironicclient: Added argcomplete lib to enable cli completition https://review.openstack.org/65865 | 16:19 |
| devananda | lucasagomes: oh - we should be running the virtual-ironic job on the client - but we're not | 16:20 |
| lucasagomes | devananda, a-ha +1 | 16:21 |
| devananda | er, wait, we ware | 16:21 |
| NobodyCam | lucasagomes: 91062 +a'd | 16:21 |
| * devananda looks closer | 16:21 | |
| lucasagomes | NobodyCam, oh wow, thanks | 16:21 |
| devananda | yea, i need more coffee | 16:21 |
| *** shausy has joined #openstack-ironic | 16:21 | |
| lucasagomes | doesns't seem we are running it no | 16:22 |
| lucasagomes | would be good to enable, since changes in the client libs can break the deploy | 16:22 |
| *** vinbs has joined #openstack-ironic | 16:23 | |
| NobodyCam | friendly poke for 93133 :-p .. just looking to get as many comments as I can before pushing up the next rev | 16:24 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix chassis bookmark link url https://review.openstack.org/94597 | 16:30 |
| openstackgerrit | Christian Berendt proposed a change to openstack/python-ironicclient: replace dict.iteritems() with six.iteritems(dict) https://review.openstack.org/95562 | 16:32 |
| *** igor_ has joined #openstack-ironic | 16:36 | |
| * lucasagomes will be afk until the meeting hour | 16:36 | |
| *** lucasagomes is now known as lucas-afk | 16:36 | |
| dtantsur | Folks, can someone review/give final +2 to https://review.openstack.org/#/c/92625/ ? I'm starting to forget what I did there :) | 16:37 |
| dtantsur | Also follow-up needs some review: https://review.openstack.org/#/c/94371/ | 16:37 |
| dtantsur | thanks :) | 16:37 |
| * dtantsur afk until the meeting | 16:39 | |
| *** igor_ has quit IRC | 16:41 | |
| *** romcheg has joined #openstack-ironic | 16:44 | |
| romcheg | Good morning/evening guy! | 16:44 |
| romcheg | guys even | 16:44 |
| NobodyCam | morning romcheg :) | 16:44 |
| romcheg | Finally I've returned from the vacation but today I'm adopting after a jetlag. | 16:45 |
| NobodyCam | hey hey romcheg want to jump in on https://review.openstack.org/#/c/93133 | 16:45 |
| romcheg | I will try to be present on the meeting but cannot guarantee that | 16:45 |
| NobodyCam | hehehhe :) | 16:45 |
| *** derekh_ has quit IRC | 16:48 | |
| *** rloo_away has joined #openstack-ironic | 16:51 | |
| *** rloo_away has quit IRC | 16:51 | |
| *** rloo_away has joined #openstack-ironic | 16:52 | |
| *** BadCub has quit IRC | 17:00 | |
| *** jistr has quit IRC | 17:13 | |
| *** Mikhail_D_ltp has quit IRC | 17:15 | |
| *** shausy has quit IRC | 17:17 | |
| * devananda edits the meeting agenda | 17:21 | |
| *** romcheg has quit IRC | 17:21 | |
| NobodyCam | ya its a little out of date ... may 5th :-p | 17:25 |
| * devananda saves update | 17:30 | |
| *** rloo has joined #openstack-ironic | 17:33 | |
| *** rloo has quit IRC | 17:33 | |
| *** rloo has joined #openstack-ironic | 17:33 | |
| *** rloo_away has quit IRC | 17:34 | |
| devananda | lucas-afk: on https://review.openstack.org/#/c/91062/ i see you used "recheck no bug" twice -- we should really be looking into why these failed, and rechecking the appropriate bug ###, so that it can be tracked and fixes can be prioritized | 17:34 |
| NobodyCam | qick walkies brb | 17:35 |
| devananda | even if they're all due to https://bugs.launchpad.net/ironic/+bug/1321494, we should be recording that | 17:36 |
| *** igor_ has joined #openstack-ironic | 17:37 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/python-ironicclient: Added argcomplete lib to enable cli completition https://review.openstack.org/65865 | 17:38 |
| *** rloo has quit IRC | 17:39 | |
| *** rloo has joined #openstack-ironic | 17:40 | |
| openstackgerrit | A change was merged to openstack/ironic: Remove 'node' parameter from the Deploy interface https://review.openstack.org/91062 | 17:41 |
| *** igor_ has quit IRC | 17:42 | |
| NobodyCam | devananda: so I should not have anded that ^^^??? | 17:43 |
| devananda | NobodyCam: no - it's fine | 17:44 |
| devananda | but in general, when we see test failures, folks should only use "recheck no bug" as a last resort | 17:44 |
| devananda | we should look into the failure and follow the guidelines on rechecks here: https://wiki.openstack.org/wiki/GerritJenkinsGit#Test_Failures | 17:45 |
| devananda | some of the time, we'll get hit by other project's issues, and it helps them when we "recheck bug ###" | 17:45 |
| devananda | and some of the time, it's a bug in ironic, and then it helps US to "recheck bug ###" | 17:45 |
| devananda | recheck no bug doesn't help anyone fix bugs -- it just hides the non-deterministic failures | 17:46 |
| NobodyCam | ack :) | 17:46 |
| *** praefect has joined #openstack-ironic | 18:05 | |
| NobodyCam | wanders afk for a few minutes | 18:07 |
| Shrews | devananda, NobodyCam: wow, such dedication, working on a holiday | 18:11 |
| NobodyCam | :) its what we do | 18:22 |
| *** igor_ has joined #openstack-ironic | 18:38 | |
| *** igor_ has quit IRC | 18:43 | |
| lifeless | morning deva | 18:46 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Workaround for conflicting volume labels https://review.openstack.org/93133 | 18:48 |
| NobodyCam | morning lifeless :) | 18:48 |
| devananda | morning, lifeless | 18:49 |
| *** lucas-afk is now known as lucasagomes | 18:55 | |
| NobodyCam | welcome back lucasagomes :) | 18:55 |
| lucasagomes | NobodyCam, cheers ;) | 18:56 |
| lucasagomes | devananda, yeah :( I will put a bug last time, the problem was that it failed on 2 different places, the first recheck was because of the check-tempest-dsvm-virtual-ironic (the rest ended up with SUCCESS) | 18:56 |
| lucasagomes | then check-tempest-dsvm-ironic failed (and the rest ended up with SUCCESS) | 18:56 |
| devananda | yay for non-deterministic failures | 18:57 |
| lucasagomes | both errors didn't look related to the change that's why I just issued recheck twice | 18:57 |
| lucasagomes | yeah | 18:57 |
| lucasagomes | :( | 18:57 |
| devananda | exactly why we need elastic rechecks | 18:57 |
| devananda | created for those sorts of things | 18:57 |
| devananda | to help us track them and fix them | 18:57 |
| lucasagomes | s/last/next* | 18:57 |
| lucasagomes | yeah I will pay more attention to it next time | 18:57 |
| lucasagomes | and put a bug there | 18:58 |
| devananda | thanks | 18:59 |
| devananda | meeting starts in a minute, for those that are around | 18:59 |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: NodeLocked causing random test failures https://review.openstack.org/95573 | 19:04 |
| max_lobur | http://www.convergedigest.com/2014/05/amds-seamicro-server-sets-openstack.html?utm_campaign=OpenStack%20Now&utm_content=5585027&utm_medium=social&utm_source=twitter | 19:06 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/python-ironicclient: Added argcomplete lib to enable cli completition https://review.openstack.org/65865 | 19:19 |
| *** romcheg has joined #openstack-ironic | 19:20 | |
| *** igor_ has joined #openstack-ironic | 19:39 | |
| *** igor_ has quit IRC | 19:43 | |
| *** sysexit has joined #openstack-ironic | 19:45 | |
| NobodyCam | lifeless: side note on check-tripleo-ironic-undercloud-precise check found this in neutron-dhcp log http://paste.openstack.org/show/VwcWMHsMKE27BTWJf6XN/ | 19:51 |
| NobodyCam | brb | 20:01 |
| lucasagomes | have a great night everybody! | 20:02 |
| dtantsur | g'night, Ironic :) | 20:02 |
| *** lucasagomes is now known as lucas-dinner | 20:02 | |
| lucas-dinner | dtantsur, night :) | 20:02 |
| dtantsur | lucas-dinner, night | 20:02 |
| devananda | specs-core updated! current list: devananda, lucas-dinner, lifeless, dtantsur, russell_h | 20:02 |
| devananda | g'night lucas-dinner | 20:02 |
| * devananda steps afk for a few minutes | 20:03 | |
| *** dtantsur is now known as dtantsur|afk | 20:03 | |
| *** ccit has joined #openstack-ironic | 20:08 | |
| *** praefect has quit IRC | 20:10 | |
| NobodyCam | night lucas-dinner dtantsur|afk | 20:11 |
| *** sysexit has quit IRC | 20:17 | |
| *** sysexit has joined #openstack-ironic | 20:18 | |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Workaround for conflicting volume labels https://review.openstack.org/93133 | 20:23 |
| rloo | NobodyCam: oops, I just gave you some comments for 93133 | 20:23 |
| NobodyCam | :) its ok :) ty rloo | 20:24 |
| lifeless | NobodyCam: interesting | 20:24 |
| rloo | NobodyCam: maybe I'm being too picky wrt the comment. It is just that it took me awhile to try to understand what was going on. | 20:24 |
| NobodyCam | :) I'll push a fix in a couple if minutes | 20:26 |
| NobodyCam | lifeless: thats on the neutron dhcp log find.. ya, I'm not sure how important it is just found it in a resent run | 20:27 |
| NobodyCam | rloo: I can /will change the doc string there but its kinda starting to feel like we're traching folks about how mbr and gpt work, which i'm not sure is our job. | 20:30 |
| rloo | NobodyCam: training isn't our job. BUT... I'm not sure what we're doing is totally correct. eg, should the GPT table be erased too? So I think since we are making some assumptions, the assumptions need to be spelled out. | 20:31 |
| NobodyCam | got to give it a few min for others to comment on it too...if they wish | 20:31 |
| rloo | NobodyCam: otherwise, people can ask similar questions like 'why 1024 bytes'? For MBR, is it correct to wipe out the end of the disk? etc. | 20:32 |
| NobodyCam | rloo: ack :) | 20:32 |
| rloo | NobodyCam: I'm not even sure why we're wiping out 1024 bytes from the end for GPT, except that it works ;) | 20:32 |
| *** sysexit has quit IRC | 20:32 | |
| rloo | NobodyCam: at least if it is wrong or doesn't work for some platform, the assumptions are there so someone can understand why we/you did what we did ;) | 20:33 |
| NobodyCam | :) yes! | 20:33 |
| *** sysexit has joined #openstack-ironic | 20:34 | |
| *** Hefeweiz1n has quit IRC | 20:35 | |
| *** romcheg has quit IRC | 20:35 | |
| *** romcheg has joined #openstack-ironic | 20:35 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 20:36 | |
| *** igor_ has joined #openstack-ironic | 20:40 | |
| *** romcheg has quit IRC | 20:40 | |
| *** ccit has quit IRC | 20:41 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic-specs: Standardize logging and improve log coverage https://review.openstack.org/94924 | 20:41 |
| *** vinbs has quit IRC | 20:42 | |
| *** igor_ has quit IRC | 20:44 | |
| lifeless | rloo: the end - thats for LVM | 20:47 |
| lifeless | rloo: and some bios assisted raids that write metadata to the end not the strt | 20:47 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic-specs: Support for different boot loaders and locations https://review.openstack.org/94925 | 20:48 |
| rloo | lifeless: the end of the disk? | 20:48 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Workaround for conflicting volume labels https://review.openstack.org/93133 | 20:50 |
| *** romcheg has joined #openstack-ironic | 20:50 | |
| NobodyCam | gah white space | 20:51 |
| lifeless | rloo: yes | 20:51 |
| rloo | lifeless: so we want to erase the end of the disk for other reasons besides GPT. | 20:52 |
| lifeless | yes | 20:52 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Workaround for conflicting volume labels https://review.openstack.org/93133 | 20:52 |
| rloo | NobodyCam will kill me if I ask him to update the comment again. | 20:52 |
| * NobodyCam wont | 20:52 | |
| NobodyCam | :-p | 20:52 |
| lifeless | generally we want to erase any metadata from *any* partitioning / disk management system | 20:53 |
| lifeless | what we're trying to do at that point is to emulate 'brand new empty disk' | 20:53 |
| lifeless | without writing to the entire disk | 20:53 |
| lifeless | which is why I proposed 4K top and bottom | 20:53 |
| rloo | lifeless, NobodyCam: maybe that's a different patch then. | 20:53 |
| rloo | lifeless: aha. I knew I wanted to ask you something. So that explains the 4k. | 20:53 |
| NobodyCam | that where the patch startedat | 20:54 |
| lifeless | not because there is a specific X,Y,Z that use 4K, but because we don't want to be in the business of updating our code every time some new metadata management layer turns out to violate our assumptions | 20:54 |
| rloo | lifeless: how do you know 4k is the magic number? | 20:54 |
| lifeless | and 4K seems large enough to thoroughly wipe most any signature scheme I can imagine | 20:54 |
| lifeless | rloo: I don't | 20:54 |
| lifeless | rloo: we could do a 1M or 1'track' equally | 20:55 |
| lifeless | I chose 4K because: | 20:55 |
| lifeless | - its the sector size in modern disks | 20:55 |
| lifeless | - its the page size in most OSes we'll encounter | 20:55 |
| lifeless | - its larger than the location of the metadata schemes I've had personal deep-dive experience with | 20:56 |
| rloo | lifeless, NobodyCam: my feedback was based on the 'destroy_mbr_gpt' function. Is the code as-is, good to go for now (if someone needs it or to make NobodyCam happy). And then do another patch to generalize to 4k and change destroy_mbr_gpt to something else? | 20:56 |
| NobodyCam | lifeless: can you please take a look at the current state of 93133 before I push up rloo next change (heheh poke poke.... j/k rloo :) ) | 20:57 |
| lifeless | sure give me a sec | 20:58 |
| NobodyCam | rloo: it would be easy for me to change back to 4k just bump up count and take a liitle more away from the total block count | 20:59 |
| rloo | NobodyCam: up to you. in addition to the math, I'd ask you to update the docstring and maybe change the name of the function. Or at least make it clear in the docstring, that it isn't to just delete MBR or GPT but any other *stuff* that might be there. | 21:01 |
| lifeless | NobodyCam: I think my earlier review feedback on the structure got lost | 21:02 |
| rloo | NobodyCam: and why it is safe to delete 1M from the start & end of the drive. | 21:03 |
| rloo | NobodyCam: sorry, I mean 4k not 1M. | 21:03 |
| lifeless | NobodyCam: did you restart this patch from a different id or something ? | 21:04 |
| NobodyCam | no | 21:04 |
| lifeless | huh | 21:05 |
| lifeless | ok, so I'm going to throw a spanner in the works | 21:05 |
| lifeless | sorry | 21:05 |
| NobodyCam | rev 17 is what https://etherpad.openstack.org/p/tripleo-end-to-end-automatic-bm calls for! | 21:05 |
| NobodyCam | oh no :-p | 21:06 |
| rloo | spanner == wrench I think ;) | 21:06 |
| lifeless | this patch is interacting with some poor structure in the preserve ephemeral patch | 21:07 |
| lifeless | I was damn sure I had put this in a review somewhere. | 21:07 |
| lifeless | putting it in this one now | 21:07 |
| * NobodyCam recalls a word derived from http://www.re-enactmentshop.com/sites/default/files/imagecache/lightbox_fullscreen/global/shop/napoleonic_wars/NAP-clogs_klompen_re-enactmentshop_dot_com.jpg | 21:08 | |
| *** mrda-away is now known as mrda | 21:09 | |
| *** romcheg has quit IRC | 21:09 | |
| mrda | Morning Ironic! | 21:10 |
| NobodyCam | good morning mrda | 21:10 |
| boris-42 | NobodyCam mrda lol why you are working during holidays?) | 21:10 |
| mrda | Hi NobodyCam - was there a meeting this morning? The wiki wasn't updated, and I wasn't going to get up at 4:30am if it wasn't on :) | 21:10 |
| NobodyCam | everyday is a holiday | 21:11 |
| NobodyCam | yea it got updated late this morning | 21:11 |
| NobodyCam | I'll also try and keep an eye out for that too | 21:12 |
| mrda | rats | 21:12 |
| * mrda goes to read scrollback | 21:12 | |
| NobodyCam | :) mrda lots of folks on holiday | 21:12 |
| mrda | of course - didn't even think of that | 21:12 |
| rloo | mrda: It is admirable of you to even contemplate getting up at 4:30am for a meeting! | 21:16 |
| rloo | NobodyCam: I'll touch base with you tomorrow about 93133. See where it gets to by then... | 21:17 |
| mrda | I checked the wiki, and it wasn't updated, so I wasn't going to take the risk. But if I know it's on, I'll get up early. Afterall - this is the project where I want to contribute :) | 21:17 |
| *** rloo is now known as rloo_away | 21:18 | |
| lifeless | NobodyCam: reviewed | 21:18 |
| NobodyCam | rloo: ack ... have a good night | 21:18 |
| mrda | but thanks for the kind words rloo_away :) | 21:18 |
| * devananda updates wiki to reference next week | 21:18 | |
| * NobodyCam is hesitant to click refresh :-p | 21:18 | |
| NobodyCam | ty lifeless :) | 21:19 |
| * devananda saves | 21:24 | |
| NobodyCam | lifeless: I should be able to get those addressed tomorrow. | 21:24 |
| sysexit | hi, i have a question. In which file is ssh key injection currently implemented. I can't find it. | 21:25 |
| sysexit | (my previous nick was geertj by the way :) | 21:25 |
| mrda | thanks devananda for updating the wiki! Appreciated :) | 21:25 |
| devananda | sysexit: that was never supported by ironic | 21:25 |
| sysexit | devananda: ah.. OK so obviously I can't find it. | 21:26 |
| lifeless | NobodyCam: yeah none of them are huge | 21:26 |
| sysexit | devananda: say you wanted to do a full undercloud + overcloud deployment with say tripleO. Wouldn't it need the ssh keys to the bare metal VMs to install the openstack components? | 21:27 |
| sysexit | s/bare metal VMs/bare metal hosts/ | 21:27 |
| devananda | sysexit: wouldn't the overcloud images have been built with those software components already present, so that the services could be configured during first-boot? | 21:28 |
| lifeless | sysexit: you might prefer to ask about tripleo in #tripleo :) | 21:28 |
| devananda | sysexit: nothing in ironic (or, afaik, tripleo) should SSH into a host and *install* something. also, what lifeless said :) | 21:28 |
| sysexit | devananda: what if s/tripleo/devstack/? Doesn't DevStack ssh into the nodes? | 21:29 |
| devananda | sysexit: the only situation in which Ironic requires an SSH key is when you use the SSHPowerDriver to control the power state of virtual machines which are used in testing to emulate hardware | 21:30 |
| devananda | sysexit: in production, you should be using physical machines with Ironic,a nd using the IPMIPowerDRiver to control their state | 21:30 |
| NobodyCam | sysexit: colud-init / os-apply-config give use this file: https://github.com/openstack/tripleo-image-elements/blob/master/elements/ironic-conductor/os-apply-config/mnt/state/var/lib/ironic/virtual-power-key | 21:30 |
| lifeless | NobodyCam: we're about to stop using that actually | 21:31 |
| NobodyCam | :-p | 21:31 |
| lifeless | since we got the better interface merged | 21:31 |
| lifeless | just need to get time to get a patch up | 21:31 |
| devananda | sysexit: devstack does not ssh into the nodes -- it ssh's to the devstack _host_ to start/stop the VMs which are emulating hardware, eg. in a test environment | 21:31 |
| sysexit | devananda: actually it does in multi-node mode. But I just noticed that it requires you to have the nodes pre-configured with ssh access. | 21:32 |
| lifeless | nothing devstack does should be considered relevant for production :) | 21:33 |
| lifeless | its a development tool | 21:33 |
| devananda | sysexit: devstack should never be SSH'ing into the Ironic nodes. When testing with Tempest, *tempest* will ssh into the resulting instances, regardless of whether you use ironic or not. | 21:33 |
| sysexit | lifeless: heh, you're right :) But our company is targeting developers at the moment :) | 21:33 |
| lifeless | and as devananda says, it shouldn't ever ssh to the emulated baremetal nodes | 21:33 |
| lifeless | in fact if ironic or devstack *can* ssh to the emulated baremetal nodes, you have a huge security hole in place | 21:34 |
| sysexit | good point... | 21:35 |
| sysexit | mind if I explain what I'm trying to do? | 21:35 |
| lifeless | please do | 21:36 |
| sysexit | I'm trying to create an Ironic driver for Ravello (www.ravellosystems.com) | 21:36 |
| sysexit | The idea is to create a "virtual undercloud" | 21:37 |
| sysexit | The VMs in Ravello have Intel VT + Layer 2 networking. | 21:37 |
| sysexit | So they should be a good basis for running a regular OpenStack deployment on top. | 21:37 |
| sysexit | Mostly for dev/test. | 21:38 |
| lifeless | you may want to read https://review.openstack.org/#/c/92642/ | 21:38 |
| devananda | sysexit: perhaps i'm missing something obvious -- what does that have to do with bare metal provisioning? | 21:39 |
| sysexit | devandana: well, it's sort of "virtual bare metal" | 21:40 |
| *** igor_ has joined #openstack-ironic | 21:40 | |
| sysexit | lifeless: that review request is interesting. | 21:41 |
| lifeless | devananda: its conceptually equivalent to the ssh driver or the openstack driver we discussed a couple weeks back. | 21:42 |
| lifeless | devananda: with all the same pro/cons around it :) | 21:42 |
| devananda | lifeless: "conceptually equivalent to the openstack driver" -- yes, and so I have the same objections to it, with one more -- ravello looks like a product, not a test plan | 21:43 |
| lifeless | devananda: I expected you would when sysexit explained :) | 21:44 |
| devananda | lifeless: and the ssh driver is quite separate in both use and function. it's not layering on top of another cloud | 21:44 |
| lifeless | devananda: it is, its just a very primitive cloud | 21:44 |
| *** igor_ has quit IRC | 21:45 | |
| sysexit | lifeless/devananda: one way to think about it is a virtual data center | 21:45 |
| *** romcheg has joined #openstack-ironic | 21:45 | |
| sysexit | we thought it could be useful | 21:45 |
| sysexit | rather than scrounging around for hardware, or having the slow performance of emulated hypervisors, you could create a fast substrate for running a real OpenStack | 21:46 |
| sysexit | devananda: can you explain your concerns? | 21:46 |
| sysexit | or maybe you think it's not useful? | 21:47 |
| devananda | sysexit: scope. Ironic is, IMNSHO, geared towards provisioning physical machines | 21:47 |
| devananda | sysexit: if you want to provision virtual machines, or manage a virtual data center, why not have a nova driver? | 21:48 |
| sysexit | devananda: I guess it's possible. But I was expecting most OpenStack deployment tooling (like TripleO and Devstack) to use the Ironic API directly at some point. | 21:49 |
| sysexit | So that's why I thought I'd be a good idea to start with Ironic | 21:49 |
| devananda | sysexit: that is an incorrect assumption :) | 21:49 |
| lifeless | sysexit: TripleO depends on Nova, not Ironic - we do use Ironic if thats the chosen backend, but only to register nodes | 21:49 |
| *** igor_ has joined #openstack-ironic | 21:50 | |
| devananda | lifeless: fwiw, having a nova driver that talks to a remote openstack cloud makes far mroe sense to me (than having an ironic driver do that) | 21:50 |
| devananda | lifeless: and there have been a few discussions about enabling a workflow in nova similar to that for federated hypervisors // federated clouds | 21:51 |
| lifeless | devananda: In order to test Ironics behaviour at scale, we need it to be an ironic driver. | 21:51 |
| lifeless | devananda: My intent is to run O(1000) node deployment tests in the next couple of months, in virt | 21:51 |
| devananda | lifeless: to test ironic's control plane behavior at that scale, in the absense of hardware, I would suggest creating a fake driver that has the same performance characteristics | 21:52 |
| lifeless | devananda: we also want to test heat's behaviour | 21:52 |
| lifeless | devananda: and novas | 21:53 |
| lifeless | devananda: and os-collect-* polling overheads | 21:53 |
| devananda | sure. again ^ | 21:53 |
| lifeless | devananda: no, you miss the point | 21:53 |
| lifeless | you're suggesting a functional test. | 21:53 |
| devananda | oh. got it now | 21:53 |
| lifeless | I want an integration test. | 21:53 |
| devananda | you want the instances to be real, cause their init scripts generate load in the cloud that deployed them | 21:53 |
| *** romcheg has quit IRC | 21:53 | |
| lifeless | I want to be able to enumerate a *small* number of differences vs a real datacentre. | 21:53 |
| lifeless | To isolate the places where bugs may lurk. | 21:54 |
| devananda | even if that load is coming from traffic on the data center's ingress network, which may be an order of magnitude slower than the internal network, and thus throttle all the requests | 21:54 |
| lifeless | physical switch -> virtual switch and IPMI -> Nova API | 21:54 |
| *** igor_ has quit IRC | 21:54 | |
| lifeless | and node performance on 1-disk physical -> VMs contending on 2-disk mirror sets | 21:54 |
| lifeless | for instance | 21:55 |
| lifeless | is a short list | 21:55 |
| lifeless | still scary | 21:55 |
| lifeless | but much less scary than 'well we have functional tests for all the things that broke last time' | 21:55 |
| mrda | lifeless: sounds like fun! | 21:55 |
| sysexit | lifeless: would you run VMs on those O(1000) nodes? If so would you use emulation (or nested virt?) | 21:55 |
| sysexit | lifeless: and regarding to your point about TripleO using Ironic to register nodes: at what level is that? Register bare-metal nodes so that TripleO can provision them through nova? If so where do you get the inventory from? | 21:56 |
| lifeless | sysexit: at this stage we're not trying to assess the behaviour of the deployed cloud; so we would, but they would likely be docker or lxc containers for sanity | 21:56 |
| lifeless | sysexit: from nodes.json | 21:56 |
| sysexit | got it | 21:56 |
| *** Mikhail_D_ltp has quit IRC | 21:56 | |
| lifeless | devananda: what ingress network? All load for these scenarios is internally generated | 21:56 |
| devananda | lifeless: see the ML thread about non-CI'd ironic drivers? I'm increasingly concerned about ironic carrying "test" drivers in tree (at least in the current location) and folks thinking they are meant for production | 21:57 |
| lifeless | devananda: I did see it. | 21:57 |
| devananda | lifeless: the functionality you're proposing could point to /any/ openstack cloud -- not just the one that tripleo-ci deploys on the adjascent rack | 21:57 |
| lifeless | devananda: the quintupleo spec you mean? | 21:58 |
| devananda | yes | 21:58 |
| lifeless | you're worried that someone would choose to use Ironic pointing to e.g. HP public cloud rather than using HP Public cloud directly ? | 21:58 |
| devananda | or that, seeing that driver in tree, think it's a sane thing to do | 21:59 |
| devananda | and implement a similar driver for AWS | 21:59 |
| devananda | for example | 21:59 |
| lifeless | I think you're borrowing a problem you don't have | 22:00 |
| lifeless | I'd be more worried about internal API pressure from such drivers | 22:00 |
| devananda | i've learned that, if we put something in tree, folks will use it. even if it's experimental/unsupported/not-meant-for-that | 22:00 |
| lifeless | than their existence | 22:00 |
| lifeless | if you put it out of tree they will use it too | 22:00 |
| lifeless | and they will whinge either way | 22:00 |
| lifeless | what concrete problem are you trying to avoid? e.g. 'I don't want users disappointed when we change something that was a test tool and break their use case' | 22:01 |
| lifeless | because - if its that, honestly, our dev community are users too, and I wouldn't want us to break them. | 22:02 |
| NobodyCam | lifeless: just fyi: 90890 fixed my issue :) | 22:05 |
| lifeless | NobodyCam: great, once it passes CI (its failing on ironic-virtual...) we can land it | 22:06 |
| lifeless | ah its passed | 22:06 |
| *** romcheg has joined #openstack-ironic | 22:09 | |
| sysexit | lifeless/devananda: similar to QuintupleO I think I only need a power driver (PXE will already work on our SDN). Setting aside if this should be in Ironic or not, do you think this will be useful at all? | 22:09 |
| sysexit | lifeless: in your comment to the review request you specially state poor performance of QEmu.. | 22:09 |
| lifeless | I do indeed, because I'm referring to our ability to test quintupleo itself within a cloud. | 22:11 |
| lifeless | that would be septupleo | 22:12 |
| lifeless | openstack on openstack on openstack on openstack | 22:12 |
| sysexit | No problem in computer science that can't be addressed by another layer of indirection :) | 22:13 |
| lifeless | do I think the ability to run a full-stack tripleo in an elastic cloud will be useful? certainly, for folk with access to that cloud | 22:14 |
| lifeless | it would be a great way to get past the scale limits of a single machine | 22:14 |
| sysexit | lifeless: we support the HP cloud by the way | 22:14 |
| sysexit | Ravello does not operate its own cloud. | 22:14 |
| sysexit | We are nested as well :) | 22:14 |
| lifeless | http://www.ravellosystems.com/ says AWS and Google | 22:14 |
| sysexit | HP too | 22:15 |
| lifeless | sure | 22:15 |
| lifeless | but it says AWS and Google | 22:15 |
| lifeless | hard in an openstack context to be excited by a product that doesn't list openstack | 22:15 |
| sysexit | I think they keyword optimized it. | 22:15 |
| lifeless | openstack isn't present on that page at all | 22:15 |
| sysexit | But the product has it in it, albeit the older version. | 22:15 |
| lifeless | nor HP nor Rackspace | 22:16 |
| lifeless | just saying, there is no way I can point $random tripleo dev at that | 22:16 |
| lifeless | they'd think I'd gone crazy | 22:16 |
| lifeless | devananda: would you object to making your specs be siblings, not a stack? | 22:17 |
| lifeless | devananda: I rather suspect the logging one, for instance, is less contentious than the all-async one | 22:17 |
| devananda | lifeless: not at all. in fact, i meant to do that. lemme fix the doc/source/index.rst change | 22:20 |
| sysexit | lifeless: we can make HP and OpenStack more public. We had Rackspace at some point but remove it. But do we really think it's need to support every major cloud before an OpenStack dev would use this? | 22:21 |
| lifeless | sysexit: I think most folk would think twice about spending money with a non-OpenStack cloud to do dev on OpenStack | 22:22 |
| lifeless | I don't think you need to support every major cloud, not at all | 22:22 |
| devananda | lifeless: actualy i need food. haven't eaten and it's 3pm. will bbl | 22:22 |
| lifeless | ttyl | 22:23 |
| sysexit | lifeless: go it, makes sense | 22:23 |
| sysexit | s/go/got/ | 22:23 |
| sysexit | lifeless: Once I have a power driver, is there any specific TripleO test case or workflow you could point me to? | 22:27 |
| sysexit | Just to see if it works.. | 22:27 |
| lifeless | well, tripleo-incubator's devtest would be a good start :) | 22:28 |
| -openstackstatus- NOTICE: stopping gerrit briefly to rebuild its search index in an attempt to fix post-rename oddities (will update with notices every 10 minutes until completed) | 22:29 | |
| *** ChanServ changes topic to "stopping gerrit briefly to rebuild its search index in an attempt to fix post-rename oddities (will update with notices every 10 minutes until completed)" | 22:29 | |
| sysexit | Cool, I'll give that a try. I assume it already has all the right Ironic stuff so from a high level I only need to set the driver to "pxe_ravello"? | 22:29 |
| *** max_lobur1 has joined #openstack-ironic | 22:29 | |
| * NobodyCam steps afk for a while | 22:29 | |
| sysexit | lifeless: many thanks for the help so far. Logging off now (it's past midnight :). | 22:32 |
| *** sysexit is now known as sysexit_away | 22:32 | |
| *** max_lobur has quit IRC | 22:33 | |
| lifeless | sysexit_away: broadly speaking :) | 22:33 |
| *** radsy has joined #openstack-ironic | 22:34 | |
| *** radsy has joined #openstack-ironic | 22:34 | |
| *** ChanServ changes topic to "OpenStack Bare Metal Provisioning | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic | Status: https://etherpad.openstack.org/p/IronicWhiteBoard" | 22:35 | |
| *** max_lobur1 has quit IRC | 22:36 | |
| *** aboutGod has joined #openstack-ironic | 22:37 | |
| *** sysexit_away has quit IRC | 22:37 | |
| openstackgerrit | A change was merged to openstack/ironic-specs: Updated .gitreview file for repo rename https://review.openstack.org/95295 | 22:39 |
| *** aboutGod has left #openstack-ironic | 22:42 | |
| *** radsy has quit IRC | 22:45 | |
| *** igor_ has joined #openstack-ironic | 22:51 | |
| *** igor_ has quit IRC | 22:55 | |
| *** lucas-dinner has quit IRC | 23:09 | |
| *** romcheg has left #openstack-ironic | 23:18 | |
| *** dvorak has quit IRC | 23:29 | |
| *** radsy has joined #openstack-ironic | 23:33 | |
| *** rloo_away has quit IRC | 23:39 | |
| lifeless | dtantsur|afk: so - is your patch being moved forward? | 23:46 |
| lifeless | dtantsur|afk: I'd like to downgrade the ironic virtual job to nonvoting on dib if its going to be more than a day or so unfixed. | 23:47 |
| lifeless | devananda: ^ | 23:47 |
| *** romcheg has joined #openstack-ironic | 23:50 | |
| *** igor_ has joined #openstack-ironic | 23:51 | |
| *** romcheg has left #openstack-ironic | 23:52 | |
| *** igor_ has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
