| *** anniec has quit IRC | 00:00 | |
| *** epim has quit IRC | 00:03 | |
| rloo | NobodyCam: it is fine to forget me;) | 00:04 |
|---|---|---|
| NobodyCam | rloo: not yiu found a good issue last look | 00:04 |
| NobodyCam | keep it up... I luv it :) | 00:05 |
| NobodyCam | s/yiu/you/ | 00:05 |
| *** matsuhashi has joined #openstack-ironic | 00:06 | |
| rloo | NobodyCam: ha ha. I'll take a quick look but am too tired to do a detailed review tonight. | 00:07 |
| NobodyCam | lol :) its only 51 lines of code... and about 100 of tests :-p | 00:07 |
| * NobodyCam *DUCKS* | 00:08 | |
| *** anniec has joined #openstack-ironic | 00:08 | |
| rloo | I am short; ducking won't help ;) | 00:08 |
| NobodyCam | doh :-p | 00:08 |
| NobodyCam | hehehe | 00:08 |
| * NobodyCam Jumps?!? | 00:09 | |
| NobodyCam | :-p | 00:09 |
| rloo | did you land yet? | 00:09 |
| NobodyCam | lol | 00:09 |
| NobodyCam | yep | 00:09 |
| rloo | so i know it was mentioned about assuming the task just has one node, but... how is anyone supposed to know that if they look at this code change? | 00:10 |
| NobodyCam | so a TODO: add multi node support? | 00:11 |
| rloo | I *think* deva and max_lobur mentioned not providing multinode support at all. | 00:11 |
| devananda | NobodyCam: I just edited https://wiki.openstack.org/wiki/Ironic#Using_Disk_Image_Builder a bit | 00:12 |
| NobodyCam | :) | 00:12 |
| devananda | so, yes | 00:12 |
| devananda | dropping multi-node support ++ | 00:12 |
| rloo | NobodyCam, devananda. dropping multi-node support for this one method, or throughout ironic code? | 00:13 |
| devananda | that derailed into an architectural discussion about schedudling and group oeprations and failt domains | 00:13 |
| NobodyCam | should Include anyhting about it on the review? | 00:13 |
| devananda | ooh sweet - my undercloud node failed to boot with | 00:14 |
| devananda | util.py: failed to resize filesystem | 00:14 |
| NobodyCam | out of disk space? | 00:15 |
| devananda | disk is 16G | 00:15 |
| NobodyCam | free space on host? | 00:16 |
| devananda | >50% free | 00:16 |
| *** epim has joined #openstack-ironic | 00:16 | |
| NobodyCam | humm | 00:16 |
| *** matsuhashi has quit IRC | 00:21 | |
| *** matsuhashi has joined #openstack-ironic | 00:21 | |
| *** matsuhashi has quit IRC | 00:25 | |
| *** matsuhas_ has joined #openstack-ironic | 00:25 | |
| *** anniec has quit IRC | 00:30 | |
| *** anniec has joined #openstack-ironic | 00:34 | |
| *** rloo has quit IRC | 00:35 | |
| devananda | hmmm | 00:38 |
| devananda | so i'm reminded of a bug i filed a bit ago | 00:39 |
| devananda | think of this situation | 00:39 |
| devananda | deploy starts | 00:39 |
| devananda | it locks a node, fetches the node and its ports from the db, and passes those objects to driver to work on them | 00:40 |
| devananda | here's where the wierdness of single vs multi node task is currently being felt | 00:40 |
| devananda | but also | 00:40 |
| devananda | what if another request comes in to update one of those ports? | 00:40 |
| devananda | like -- to change it's MAC | 00:40 |
| devananda | right now, ports don't have a "reservation" field. | 00:40 |
| devananda | there's no locking done around updating ports at all. let alone cross-resource lock awareness | 00:41 |
| devananda | like you can't update this port because the Node that it belons to is locked | 00:41 |
| devananda | but we obviously shouldn't | 00:41 |
| devananda | allow the port's mac to be changed mid-deploy | 00:41 |
| devananda | NobodyCam ^ | 00:42 |
| NobodyCam | yes | 00:43 |
| devananda | i was going to point rloo at that, but she's gone | 00:43 |
| NobodyCam | ya | 00:44 |
| NobodyCam | also digambar_ has started on https://bugs.launchpad.net/ironic/+bug/1276393 | 00:44 |
| NobodyCam | which seem to be along the same lines | 00:44 |
| NobodyCam | might be a good time to update the bug :-p out from under him | 00:45 |
| devananda | oh | 00:45 |
| devananda | that's the bug i filed :) | 00:46 |
| NobodyCam | :) | 00:46 |
| devananda | yep that's fine | 00:46 |
| devananda | that solution is fine | 00:46 |
| devananda | needs a new RPC method | 00:46 |
| devananda | that's probably all | 00:47 |
| devananda | API should call update_port like it calls update_node today | 00:47 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 00:50 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 00:51 |
| NobodyCam | devananda: updated per rloo's and max's comemnts | 00:52 |
| *** epim has quit IRC | 00:53 | |
| devananda | :) | 00:54 |
| *** nosnos has joined #openstack-ironic | 01:01 | |
| NobodyCam | ok /me calles it a day. devananda tomorrow ^^^ I will work ^^^ until it land tomorrow! | 01:07 |
| devananda | :) | 01:12 |
| devananda | romcheg: any progress on the nova_bm -> ironic db migration stuff? | 01:12 |
| NobodyCam | :) | 01:12 |
| *** john-n-seattle1 has quit IRC | 01:13 | |
| *** jbjohnso has joined #openstack-ironic | 01:14 | |
| *** anniec has quit IRC | 01:17 | |
| *** anniec has joined #openstack-ironic | 01:38 | |
| *** killer_prince has quit IRC | 01:55 | |
| *** anniec has quit IRC | 02:03 | |
| *** jbjohnso has quit IRC | 02:34 | |
| *** john-n-seattle has joined #openstack-ironic | 02:44 | |
| *** anniec has joined #openstack-ironic | 02:45 | |
| *** anniec has quit IRC | 02:58 | |
| *** anniec has joined #openstack-ironic | 03:03 | |
| *** matsuhas_ has quit IRC | 03:39 | |
| *** dkehn__ has joined #openstack-ironic | 04:00 | |
| *** dkehn has quit IRC | 04:03 | |
| *** dkehn_ has quit IRC | 04:03 | |
| *** dkehn has joined #openstack-ironic | 04:05 | |
| *** dkehn__ is now known as dkehn_ | 04:12 | |
| *** anniec has quit IRC | 04:27 | |
| *** killer_prince has joined #openstack-ironic | 04:33 | |
| *** shawal has joined #openstack-ironic | 04:43 | |
| *** harlowja is now known as harlowja_away | 04:44 | |
| *** shawal has quit IRC | 04:44 | |
| *** shawal has joined #openstack-ironic | 04:44 | |
| *** matsuhashi has joined #openstack-ironic | 04:52 | |
| *** lazy_prince has joined #openstack-ironic | 05:31 | |
| *** aignatov_ is now known as aignatov | 05:36 | |
| *** nosnos_ has joined #openstack-ironic | 05:43 | |
| *** nosnos_ has quit IRC | 05:44 | |
| *** nosnos_ has joined #openstack-ironic | 05:44 | |
| *** nosnos has quit IRC | 05:46 | |
| *** anniec has joined #openstack-ironic | 06:02 | |
| *** saju_m has joined #openstack-ironic | 06:02 | |
| *** nosnos_ has quit IRC | 06:04 | |
| *** nosnos has joined #openstack-ironic | 06:04 | |
| *** anniec_ has joined #openstack-ironic | 06:05 | |
| *** anniec has quit IRC | 06:06 | |
| *** anniec_ is now known as anniec | 06:06 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/71192 | 06:06 |
| *** hstimer has quit IRC | 06:12 | |
| *** dkehn__ has joined #openstack-ironic | 06:14 | |
| *** dkehn__ has quit IRC | 06:16 | |
| *** dkehn__ has joined #openstack-ironic | 06:17 | |
| *** dkehn has quit IRC | 06:17 | |
| *** dkehn_ has quit IRC | 06:17 | |
| *** shawal has quit IRC | 06:17 | |
| *** dkehn has joined #openstack-ironic | 06:19 | |
| *** lifeless has quit IRC | 06:34 | |
| lazy_prince | hi all, | 06:35 |
| *** dkehn_ has joined #openstack-ironic | 06:36 | |
| lazy_prince | I am using devstack and i am getting error when nova-api is starting. http://paste.openstack.org/show/64530/ | 06:37 |
| lazy_prince | can someone please help.. | 06:38 |
| *** dkehn has quit IRC | 06:38 | |
| *** dkehn__ has quit IRC | 06:39 | |
| *** dkehn has joined #openstack-ironic | 06:40 | |
| *** max_lobur_afk has quit IRC | 06:40 | |
| *** max_lobur_afk has joined #openstack-ironic | 06:41 | |
| *** anniec has quit IRC | 06:41 | |
| *** romcheg has quit IRC | 06:42 | |
| *** romcheg has joined #openstack-ironic | 06:42 | |
| *** dshulyak has quit IRC | 06:44 | |
| *** max_lobur_afk has quit IRC | 06:44 | |
| *** mdurnosvistov has quit IRC | 06:44 | |
| *** romcheg has quit IRC | 06:44 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Make ironic-api not single threaded https://review.openstack.org/72851 | 06:44 |
| *** max_lobur_afk has joined #openstack-ironic | 06:44 | |
| *** mdurnosvistov has joined #openstack-ironic | 06:44 | |
| *** max_lobur_afk has quit IRC | 06:45 | |
| *** mdurnosvistov has quit IRC | 06:45 | |
| *** mdurnosvistov has joined #openstack-ironic | 06:45 | |
| *** max_lobur_afk has joined #openstack-ironic | 06:45 | |
| *** romcheg has joined #openstack-ironic | 06:45 | |
| *** dshulyak has joined #openstack-ironic | 06:45 | |
| devananda | lazy_prince: nova-api? i think you may be in the wrong channel | 06:48 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add samples to serve API through Apache mod_wsgi https://review.openstack.org/72069 | 06:53 |
| lazy_prince | devananda: my mistake. i didnt notice that i am in wrong channel.. | 06:56 |
| *** aignatov is now known as aignatov_ | 07:00 | |
| *** dkehn__ has joined #openstack-ironic | 07:01 | |
| *** aignatov_ is now known as aignatov | 07:02 | |
| *** dkehn has quit IRC | 07:04 | |
| *** dkehn_ has quit IRC | 07:04 | |
| *** dkehn has joined #openstack-ironic | 07:06 | |
| *** athomas has joined #openstack-ironic | 07:17 | |
| *** aignatov is now known as aignatov_ | 07:19 | |
| *** coolsvap has joined #openstack-ironic | 07:21 | |
| *** coolsvap1 has joined #openstack-ironic | 07:26 | |
| *** coolsvap has quit IRC | 07:28 | |
| *** coolsvap1 has quit IRC | 07:36 | |
| *** coolsvap has joined #openstack-ironic | 07:36 | |
| GheRivero | morning all | 08:09 |
| *** mdurnosvistov_ has joined #openstack-ironic | 08:10 | |
| *** romcheg1 has joined #openstack-ironic | 08:16 | |
| *** dkehn_ has joined #openstack-ironic | 08:24 | |
| Haomeng | GheRivero: morning:) | 08:24 |
| *** pquerna has quit IRC | 08:25 | |
| *** dkehn_ has quit IRC | 08:26 | |
| *** pquerna has joined #openstack-ironic | 08:26 | |
| *** dkehn_ has joined #openstack-ironic | 08:26 | |
| *** dkehn has quit IRC | 08:26 | |
| *** dkehn__ has quit IRC | 08:27 | |
| *** dkehn has joined #openstack-ironic | 08:28 | |
| *** pquerna has quit IRC | 08:49 | |
| *** pquerna has joined #openstack-ironic | 08:58 | |
| *** coolsvap1 has joined #openstack-ironic | 08:59 | |
| *** mdurnosvistov_ has quit IRC | 09:00 | |
| *** coolsvap has quit IRC | 09:01 | |
| *** ifarkas has joined #openstack-ironic | 09:02 | |
| *** aignatov_ is now known as aignatov | 09:02 | |
| *** jistr has joined #openstack-ironic | 09:04 | |
| *** jistr is now known as jistr|sick | 09:04 | |
| openstackgerrit | Yongli He proposed a change to openstack/ironic: Use six.moves.urllib.parse instead of urlparse https://review.openstack.org/72889 | 09:09 |
| *** lucasagomes has joined #openstack-ironic | 09:15 | |
| *** yuriyz has joined #openstack-ironic | 09:19 | |
| *** tatyana has joined #openstack-ironic | 09:32 | |
| *** athomas has quit IRC | 09:36 | |
| Haomeng | lucasagomes: morning:) | 09:38 |
| lucasagomes | Haomeng, morning :) | 09:46 |
| Haomeng | lucasagomes: :) | 09:46 |
| Haomeng | lucasagomes: I have an idea about auto-discory and register ironic ipmi node utility, just ping ipmi port to discovery it and get arp list to find the mac by ip, how do you think about this idea, can we create a bp for this new feature? | 09:46 |
| Haomeng | lucasagomes: auto-discovery | 09:47 |
| lucasagomes | Haomeng, nice! yea there's a bp already in the "discussion" state | 09:47 |
| lucasagomes | so we could update it with such infos | 09:47 |
| * lucasagomes grabs the link | 09:47 | |
| lucasagomes | https://blueprints.launchpad.net/ironic/+spec/discovery-ramdisk | 09:47 |
| Haomeng | lucasagomes: I know this bp, but it is in-band | 09:47 |
| Haomeng | lucasagomes: my idea is try to find ipmi physical node | 09:48 |
| Haomeng | lucasagomes: does not depend on ramdisk | 09:48 |
| lucasagomes | Haomeng, ahh ack, yea that bp is in-band for a more generic type of discovery | 09:48 |
| lucasagomes | sure feel free to put up a blueprint for that :) | 09:48 |
| Haomeng | lucasagomes: this discovery tool can be run from our ironic conductor node | 09:49 |
| Haomeng | lucasagomes: ok, thank you:) | 09:49 |
| lucasagomes | yea that would be pretty useful to have | 09:49 |
| lucasagomes | thank u! | 09:49 |
| Haomeng | lucasagomes: and I want to get it with a discovery-template with default ipmi user/password and other parameters to be populated to our ironic node when creating | 09:50 |
| Haomeng | lucasagomes: :) | 09:50 |
| lucasagomes | nice! btw I read something re ipmi / ibm yesterday | 09:51 |
| Haomeng | lucasagomes: to auto-config our node driver info | 09:51 |
| lucasagomes | lemme try to find the link | 09:51 |
| lucasagomes | a guy was reverse engineering the protocol | 09:51 |
| Haomeng | lucasagomes: ok | 09:51 |
| lucasagomes | http://mjg59.dreamwidth.org/29210.html | 09:51 |
| Haomeng | lucasagomes: got, thk | 09:51 |
| lucasagomes | Haomeng, nice, yea auto-discovery is one of the most requested feature in Ironic | 09:52 |
| lucasagomes | we def gotta include it in J | 09:52 |
| Haomeng | lucasagomes: ok | 09:52 |
| Haomeng | lucasagomes: let me prepare the draft first:) | 09:52 |
| lucasagomes | Haomeng, sure :) thanks | 09:52 |
| Haomeng | lucasagomes: and welcome your comments once the bp is created:) | 09:52 |
| lucasagomes | ^^ will do | 09:53 |
| Haomeng | lucasagomes: :) | 09:53 |
| Haomeng | lucasagomes: nice day, I will be away for a while:) | 09:53 |
| lucasagomes | Haomeng, cheers, for u too | 09:53 |
| Haomeng | lucasagomes: :) | 09:54 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix API inconsistence when changing node's states https://review.openstack.org/67506 | 10:00 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix API inconsistence when changing node's states https://review.openstack.org/67506 | 10:16 |
| *** martyntaylor has joined #openstack-ironic | 10:33 | |
| *** martyntaylor has quit IRC | 10:33 | |
| *** jbjohnso has joined #openstack-ironic | 10:38 | |
| *** matsuhashi has quit IRC | 10:41 | |
| *** aignatov is now known as aignatov_ | 10:41 | |
| *** jbjohnso has quit IRC | 10:50 | |
| *** martyntaylor has joined #openstack-ironic | 10:52 | |
| *** saju_m has quit IRC | 10:57 | |
| *** jistr|sick has quit IRC | 10:57 | |
| *** max_lobur_afk is now known as max_lobur | 11:03 | |
| max_lobur | morning Everyone! | 11:03 |
| *** lifeless has joined #openstack-ironic | 11:07 | |
| lazy_prince | Hi.. I am planning to test https://review.openstack.org/#/c/64542. I have an esxi host registered to vCenter. do I need to provide creds for the the host or for the vCenter. i believe i have to use host creds.. do i need to specify any additional properties or any config change? | 11:11 |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/70382 | 11:13 |
| lucasagomes | hi lazy_prince yes you will need the host creds | 11:13 |
| lucasagomes | lazy_prince, you might want to take a look at https://etherpad.openstack.org/p/IronicDeployDevstack | 11:13 |
| lucasagomes | it's a guide to deploy a vm using ironic with pxe_ssh driver | 11:14 |
| lucasagomes | note we are using "ssh_virt_type=virsh" so you'll have o change it to "vmware" for ur test | 11:14 |
| *** dkehn__ has joined #openstack-ironic | 11:18 | |
| *** john-n-seattle2 has joined #openstack-ironic | 11:18 | |
| *** dkehn__ has quit IRC | 11:20 | |
| *** dkehn__ has joined #openstack-ironic | 11:20 | |
| *** dkehn has quit IRC | 11:20 | |
| *** aignatov_ is now known as aignatov | 11:21 | |
| *** dkehn_ has quit IRC | 11:21 | |
| lazy_prince | lucasagomes: thanks. I am also assuming that for this to work the ESXi host should not be in locked-in mode. let me test it and get back to you | 11:21 |
| lucasagomes | lazy_prince, ack | 11:21 |
| *** john-n-seattle has quit IRC | 11:21 | |
| *** dkehn has joined #openstack-ironic | 11:22 | |
| *** aignatov is now known as aignatov_ | 11:37 | |
| *** saju_m has joined #openstack-ironic | 11:40 | |
| *** matsuhashi has joined #openstack-ironic | 11:44 | |
| max_lobur | tempest check for each comment seems to bring some sense | 11:44 |
| max_lobur | merge conflicts discovered on early stage | 11:44 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix API inconsistence when changing node's states https://review.openstack.org/67506 | 11:46 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add topic to the change_node_maintenance_mode() RPC method https://review.openstack.org/72724 | 11:47 |
| max_lobur | does anybody know how to properly apply this https://review.openstack.org/#/c/72069/2 ? :) | 11:49 |
| max_lobur | I'm not very familiar with apache | 11:49 |
| max_lobur | but would like to test | 11:49 |
| *** saju_m has quit IRC | 11:49 | |
| max_lobur | what should I do? I tried to copy to etc/apache2/ironic and restart apache, but it does not respond to after this (basically the api is not running) | 11:50 |
| max_lobur | brb | 11:54 |
| devananda | morning, all | 11:59 |
| max_lobur | morning devananda :) | 12:01 |
| max_lobur | you're early today | 12:01 |
| devananda | indeed. and i was up late | 12:02 |
| lucasagomes | morning max_lobur devananda | 12:06 |
| max_lobur | morning Lucas | 12:06 |
| yuriyz | Morning Ironic | 12:13 |
| *** coolsvap_ has joined #openstack-ironic | 12:14 | |
| max_lobur | morning yuriyz 8) | 12:14 |
| yuriyz | devananda, should "if self.host != mapped_hosts[0]" be here https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L364 ? | 12:15 |
| yuriyz | otherwise sync tasks duplicated IMO | 12:16 |
| *** coolsvap1 has quit IRC | 12:17 | |
| *** dkehn_ has joined #openstack-ironic | 12:18 | |
| *** dkehn has quit IRC | 12:20 | |
| *** dkehn__ has quit IRC | 12:21 | |
| *** dkehn has joined #openstack-ironic | 12:22 | |
| devananda | yuriyz: if hash_distribution_replicas > 1, then you're correct | 12:23 |
| devananda | yuriyz: actually, I think, instead of only checking if self.host == mapped_hosts[0], we should get list of dead conductors and exclude them, and see if self.host is the first one still alive | 12:25 |
| *** aignatov_ is now known as aignatov | 12:26 | |
| devananda | yuriyz: that is the intent of hash_distribution_replicas > 1 -- let next conductor in the ring take over for a temporarily down conductor, before a full rebalance is initiated | 12:26 |
| *** jbjohnso has joined #openstack-ironic | 12:27 | |
| devananda | until we finish implementing take over / rebalance things, perhaps hash distribution replicas should default to 1 | 12:36 |
| *** lucasagomes is now known as lucas-hungry | 12:41 | |
| devananda | max_lobur: as you're testing your workerpool stuff, be sure you're using a threaded api server | 12:41 |
| lucas-hungry | devananda, left some comments on the threaded api server review | 12:41 |
| devananda | max_lobur: eg, 72851 or 72069 | 12:41 |
| devananda | lucas-hungry: ack, just about to fix them | 12:41 |
| lucas-hungry | ack | 12:42 |
| lucas-hungry | there's one thing that happened on my env right now when using that patch | 12:42 |
| lucas-hungry | I gotta take a better look | 12:42 |
| lucas-hungry | going to grab some food first | 12:42 |
| *** Alexei_987 has joined #openstack-ironic | 12:45 | |
| *** aignatov is now known as aignatov_ | 12:45 | |
| max_lobur | devananda, k, actually I didn't notice 72851 so far | 12:45 |
| max_lobur | will take a look | 12:45 |
| max_lobur | the 72851 and 72069 are alternatives to each other, right? | 12:46 |
| devananda | max_lobur: right | 12:47 |
| *** aignatov_ is now known as aignatov | 12:48 | |
| max_lobur | devananda, and about 72069 - are there other steps except copy to /etc and restart apache to make it running? | 12:48 |
| max_lobur | or there are much more? | 12:48 |
| *** lucas-hungry has quit IRC | 12:50 | |
| openstackgerrit | Martyn Taylor proposed a change to openstack/ironic: Remove POST calls in tests for resource creation https://review.openstack.org/64674 | 12:54 |
| *** mgoddard has quit IRC | 13:00 | |
| lazy_prince | regarding https://review.openstack.org/#/c/64542, I have a small doubt, will SSHPower driver create the baremetal vm on the ESXi host for me or do I have to create a VM manually and then create ironic node and set its MAC id.. | 13:01 |
| *** lucas-hungry has joined #openstack-ironic | 13:03 | |
| devananda | max_lobur: copy that and edit as needed for your enviroment (paths, etc) and install mod_wsgi | 13:04 |
| max_lobur | devananda, ack | 13:05 |
| *** saju_m has joined #openstack-ironic | 13:05 | |
| devananda | max_lobur: greenthreaded api is sufficient for testing. production folks will want the apache mod_wsgi | 13:05 |
| *** saju_m has quit IRC | 13:06 | |
| *** saju_m has joined #openstack-ironic | 13:07 | |
| devananda | yuriyz: https://bugs.launchpad.net/ironic/+bug/1279331 | 13:08 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Make ironic-api not single threaded https://review.openstack.org/72851 | 13:12 |
| *** mgoddard has joined #openstack-ironic | 13:15 | |
| lazy_prince | So I have tried testing https://review.openstack.org/#/c/64542 and the VM is not coming up. the logs and the node details are placed at http://paste.openstack.org/show/64631/ . can someone please take a look and help me figure out what did i miss.. | 13:19 |
| devananda | lucas-hungry: what exception are you triggering? I'm seeing proper error returns for things | 13:21 |
| devananda | lazy_prince: the error is coming from http://git.openstack.org/cgit/openstack/ironic/tree/ironic/drivers/modules/ssh.py#n168 | 13:23 |
| devananda | lazy_prince: there should be some ERROR in the logs before that EXCEPTION. it may also help to turn on DEBUG logging | 13:24 |
| devananda | lazy_prince: the SSH power driver is not able to find the VM you specified. This is the method that tries to search for the VM name based on MAC address of the node's port(s) -- http://git.openstack.org/cgit/openstack/ironic/tree/ironic/drivers/modules/ssh.py#n213 | 13:26 |
| devananda | lazy_prince: have you associated a Port with the VM node? | 13:26 |
| lazy_prince | yes.. the mac id is 00:50:56:9c:30:ea | 13:27 |
| *** jdob has joined #openstack-ironic | 13:28 | |
| lazy_prince | devananda: I am using devstack, how do i turn debug logs.. | 13:30 |
| devananda | lazy_prince: edit /etc/ironic/ironic.conf and uncomment this line | 13:32 |
| devananda | #debug=true | 13:32 |
| devananda | then restart ir-cond service | 13:32 |
| lazy_prince | devananda: thanks.. let me try this. | 13:33 |
| *** mdurnosvistov has quit IRC | 13:37 | |
| *** mdurnosvistov has joined #openstack-ironic | 13:38 | |
| devananda | agordeev: ping | 13:52 |
| devananda | agordeev: question on https://review.openstack.org/#/c/70348/5/lib/ironic -- are you testing this with dnsmasq? I see your #TODOs about it, but I am not aware that Ironic works with dnsmasq today | 13:53 |
| lazy_prince | devananda: so i noticed that debug is already enabled in my case and the complete log along with node details, port details are available at http://paste.openstack.org/show/64641/ | 13:56 |
| *** lucas-hungry is now known as lucasagomes | 13:56 | |
| max_lobur | bbiaw | 13:58 |
| lucasagomes | devananda, so afair, the exceptions was raised in the conductor but weren't send be to the api | 13:59 |
| lucasagomes | so the api was hanging on the request, I will take a better look at it | 14:00 |
| lucasagomes | I'm also testing the console patch on top of ur patch so it's a bit messy | 14:00 |
| lazy_prince | devananda: I am using esxi 5.1 build 1065491 and i notices that some of the commands that was coded for vmware does not work on the esxi shell | 14:01 |
| *** coolsvap_ has quit IRC | 14:02 | |
| devananda | lazy_prince: ahhh. ok! that's good to know -- i don't have access to vmwre/esxi shell to test | 14:03 |
| devananda | lazy_prince: can provide feedback on the review with details about those commands? | 14:03 |
| devananda | lucasagomes: exception in what RPC method? | 14:04 |
| devananda | lucasagomes: oh heh, nvm, you said you're also working on the console stuff | 14:05 |
| lucasagomes | devananda, the exception was a RPC method which is not in trunk, I was testing the console patch | 14:05 |
| lucasagomes | yea | 14:05 |
| devananda | :) | 14:05 |
| *** rloo has joined #openstack-ironic | 14:07 | |
| lazy_prince | devananda: sure.. but before that i wanted to ensure that my setup is proper and there is no user mistake. since i am not able to execute certain commands like *vim-cmd list --all*, it makes me feel that its a bug in the code and i might be able to find what is the exact command that will work. but again it may depend on esxi version as well so I would like to know what version of esxi was used during development. | 14:08 |
| lazy_prince | devananda: in the worst may be, we will end up using different ssh_virt_type for different versions of esxi. | 14:08 |
| devananda | lazy_prince: hm | 14:10 |
| devananda | this is a copy/paste error: "list --all|grep running|awk -v qc='\"' -F\" \" '{print qc$2qc}'", | 14:10 |
| devananda | that is the command for virsh, not for vmware | 14:10 |
| devananda | lazy_prince: i'm going to post another rev of that patch. it's clearly a copy/paste error.... here is the original Nova patch with proper vmware commands: https://review.openstack.org/#/c/64542/1/ironic/drivers/modules/ssh.py | 14:12 |
| devananda | bad paste. i mean https://review.openstack.org/#/c/58940/9/nova/virt/baremetal/virtual_power_driver_settings.py | 14:13 |
| lazy_prince | devananda: okay.. thanks for the update. I will add my comments to the review and I will test it once its updated | 14:14 |
| devananda | Haomeng: ping | 14:14 |
| *** saju_m has quit IRC | 14:21 | |
| *** aignatov is now known as aignatov_ | 14:25 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: SSHPower driver support VMware ESXi https://review.openstack.org/64542 | 14:27 |
| devananda | Haomeng: I have updated your patch ^ | 14:27 |
| devananda | lazy_prince: hope that helps | 14:28 |
| *** aignatov_ is now known as aignatov | 14:28 | |
| *** matty_dubs|gone is now known as matty_dubs | 14:34 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Add ability to break TaskManager locks via REST API https://review.openstack.org/71212 | 14:37 |
| *** saju_m has joined #openstack-ironic | 14:38 | |
| lazy_prince | devananda: I have tested the 2nd patch set and it seems to be working in my case. | 14:52 |
| devananda | lazy_prince: that's great :) | 14:53 |
| agordeev | devananda: ping. hello! | 14:55 |
| max_lobur | back | 14:55 |
| agordeev | morning/evening, guys! | 14:55 |
| lazy_prince | devananda: my test just included creating a node in ironic and provisioning instance on the baremetal VM. is there any specific testing that I need to perform. please let me know. | 14:56 |
| agordeev | devananda: i'm using https://etherpad.openstack.org/p/IronicDeployDevstack as the source. It suggests to use dnsmasq :) | 14:57 |
| lazy_prince | i too am using dnsmasq as described in https://etherpad.openstack.org/p/IronicDeployDevstack and it works for me. | 14:59 |
| devananda | agordeev: ahhh... right. I blame lucasagomes for that :) | 14:59 |
| lucasagomes | hah | 14:59 |
| lucasagomes | lemme read | 14:59 |
| lucasagomes | agordeev, oh yea my bad, I mean, it's a temporary solution | 15:00 |
| devananda | lucasagomes: have you gotten deploy with the noav driver to work with neutron? That shouldn't be dependent on neutron support in our PXE driver | 15:00 |
| lucasagomes | as the neutron integration is not yet complete (almost there) | 15:00 |
| lucasagomes | devananda, not yet :( but we have a guy testing it in the moment | 15:00 |
| lucasagomes | ifarkas, | 15:00 |
| devananda | k | 15:00 |
| NobodyCam | Good morning Ironic, says the man making coffee | 15:00 |
| devananda | NobodyCam: morning! | 15:00 |
| * ifarkas reading back | 15:01 | |
| devananda | lazy_prince: so as a power driver, if you are able to start/stop/reboot an ESXi VM using ironic with the SSH power driver, then I think that is sufficient tests for that patch | 15:01 |
| agordeev | devananda: also a question. lib/baremetal uses bm_poseur to maintain VMs and bridge. Should I reuse it? The code seems too obsolete and outdated, so I'm in doubt and introduced mine functions to lib/ironic which are duplicating lib/baremetal functionality | 15:02 |
| devananda | agordeev: please don't reuse it, heh | 15:02 |
| devananda | agordeev: i think it would be better to use the code from tripleo-incubator, or if it's not that much, just use your own | 15:02 |
| lazy_prince | devananda: I deleted the node using ironic node-delete but the VM is not powered off. is that expected..? or do i have to use a separate cmd to power off the VM.. | 15:03 |
| devananda | lazy_prince: oh, huh. that could be a bug. please try all three commansd: power on, power off, reboot | 15:03 |
| devananda | lazy_prince: also please test it with more than 1 VM to make sure that it is correctly identifying the right VM | 15:04 |
| lazy_prince | devananda: can you please help me with the cmds, i am kinda new to these stuff... | 15:04 |
| * lucasagomes adds a not to the pad saying that the dnsmasq step is a temporary solution | 15:04 | |
| lucasagomes | note* | 15:04 |
| devananda | lazy_prince: actually -- yes,that is expected right now. if you merely issue "ironic node-delete", it will not attempt to change the power state of the node prior to deleting it | 15:04 |
| NobodyCam | max_lobur: fyi: on 70468 I got this for port in task.node.ports: | 15:05 |
| NobodyCam | AttributeError: 'Node' object has no attribute 'ports' | 15:05 |
| lazy_prince | devananda: okay... | 15:05 |
| max_lobur | NobodyCam, lemme see | 15:05 |
| devananda | lazy_prince: once you have enrolled with node-creaet and node-update, then issue "ironic node-set-power-state $NODE $state" where state is one of "on", "off", and "reboot" | 15:06 |
| max_lobur | NobodyCam, you're right | 15:06 |
| max_lobur | sorry | 15:06 |
| max_lobur | my bad | 15:06 |
| max_lobur | task.node - is a reference to UOM node | 15:06 |
| max_lobur | not to the node resource | 15:06 |
| max_lobur | therefore will have to leave task.resources[0].ports | 15:06 |
| max_lobur | sorry for confusion | 15:06 |
| NobodyCam | :) | 15:07 |
| devananda | max_lobur: thus why i'd like to refactor the resource manager class :) | 15:08 |
| max_lobur | :) | 15:08 |
| max_lobur | so far we have: | 15:08 |
| max_lobur | 1. Node SqlAlchemy model | 15:08 |
| max_lobur | 2. Node UOM | 15:08 |
| max_lobur | 3. Node resource | 15:09 |
| lazy_prince | devananda: okay.. I will do that tomorrow. getting late.. will do rest of the SSHpower testing tomorrow. | 15:09 |
| max_lobur | 4. Node object in the REST api | 15:09 |
| max_lobur | may be confusing for beginning :) | 15:09 |
| devananda | lazy_prince: thanks much! | 15:09 |
| NobodyCam | max_lobur: it is confusing for me at times | 15:10 |
| max_lobur | going to grab some food | 15:10 |
| max_lobur | bbiaw | 15:10 |
| NobodyCam | :) | 15:10 |
| max_lobur | devananda, https://review.openstack.org/#/c/72851/ - works for me | 15:10 |
| devananda | max_lobur: dont forget node the actual hardware :) | 15:10 |
| max_lobur | though I'd like to wait while Lucas doing his investigation | 15:11 |
| max_lobur | devananda, ++ :D | 15:11 |
| devananda | max_lobur: why +1 not +2? | 15:11 |
| max_lobur | because I'm waiting for Lucas :) | 15:11 |
| devananda | max_lobur: so +2 it :) | 15:11 |
| max_lobur | k ;D | 15:12 |
| devananda | max_lobur: then if lucas finishes his testing, he can +2/+A | 15:12 |
| max_lobur | right | 15:12 |
| max_lobur | done :D | 15:12 |
| devananda | hahaha | 15:13 |
| *** lazy_prince has quit IRC | 15:13 | |
| max_lobur | :D | 15:14 |
| max_lobur | bbiaw | 15:14 |
| *** max_lobur is now known as max_lobur_afk | 15:14 | |
| devananda | max_lobur: you can always say "+2 but i want to give $someone time to test it" and other cores will honor that and not approve .... | 15:14 |
| *** observer_ has joined #openstack-ironic | 15:18 | |
| NobodyCam | brb | 15:20 |
| *** observer_ has left #openstack-ironic | 15:21 | |
| openstackgerrit | Andrey Kurilin proposed a change to openstack/python-ironicclient: Sync latest code and reuse exceptions from oslo https://review.openstack.org/71500 | 15:24 |
| *** aignatov is now known as aignatov_ | 15:33 | |
| *** dkehn__ has joined #openstack-ironic | 15:37 | |
| *** coolsvap has joined #openstack-ironic | 15:37 | |
| *** dkehn has quit IRC | 15:38 | |
| *** dkehn_ has quit IRC | 15:38 | |
| *** dkehn has joined #openstack-ironic | 15:40 | |
| *** dkehn_ has joined #openstack-ironic | 15:43 | |
| *** dkehn has quit IRC | 15:45 | |
| *** dkehn__ has quit IRC | 15:47 | |
| *** dkehn__ has joined #openstack-ironic | 15:47 | |
| *** aignatov_ is now known as aignatov | 15:47 | |
| *** dkehn has joined #openstack-ironic | 15:48 | |
| *** tatyana_ has joined #openstack-ironic | 15:48 | |
| *** davidlenwell_ has joined #openstack-ironic | 15:49 | |
| *** MIDENN_ has joined #openstack-ironic | 15:50 | |
| *** SpamapS_ has joined #openstack-ironic | 15:51 | |
| devananda | https://review.openstack.org/#/c/63937/ needs another core to review it | 15:54 |
| romcheg1 | I'm on it | 15:54 |
| *** thedodd has joined #openstack-ironic | 15:55 | |
| *** dkehn_ has quit IRC | 15:56 | |
| *** martyntaylor has quit IRC | 15:56 | |
| *** tatyana has quit IRC | 15:56 | |
| *** SpamapS has quit IRC | 15:56 | |
| *** davidlenwell has quit IRC | 15:56 | |
| *** mordred has quit IRC | 15:56 | |
| *** tatyana_ is now known as tatyana | 15:56 | |
| *** ifarkas has quit IRC | 15:58 | |
| lucasagomes | devananda, start/stop of the console, async or sync? | 15:59 |
| lucasagomes | devananda, I thought about async, but I don't think it that a task like that | 15:59 |
| lucasagomes | would take a long time | 15:59 |
| lucasagomes | (unless ipmi times out etc) | 16:00 |
| lucasagomes | hmm | 16:00 |
| lucasagomes | NobodyCam, ^ | 16:00 |
| devananda | lucasagomes: hah. don't count on anything with ipmi being synchronous | 16:00 |
| devananda | lucasagomes: async it must be | 16:00 |
| lucasagomes | ack | 16:00 |
| devananda | lucasagomes: also, other drivers .... | 16:00 |
| devananda | we need to broadly adopt the model that all operations on nodes are async | 16:01 |
| lucasagomes | hah thanks | 16:01 |
| lucasagomes | indeed | 16:01 |
| *** coolsvap1 has joined #openstack-ironic | 16:01 | |
| devananda | we already basicaly do that, but it's not explicitly documented anywhere that that is our policy | 16:01 |
| lucasagomes | what holds me on the async is always the feedback for the user | 16:01 |
| lucasagomes | to the* | 16:02 |
| devananda | yea, i get that -- we want instant gratification | 16:02 |
| devananda | but this is a cloud :) | 16:02 |
| devananda | everything is async, because sync does not scale | 16:02 |
| lucasagomes | devananda, yea, for e.g the console... should we use the last_error as well? | 16:02 |
| devananda | yep | 16:02 |
| lucasagomes | ack | 16:02 |
| lucasagomes | we might want to include a flag on states telling whether the console is enabled/disabled/not supported | 16:03 |
| *** coolsvap has quit IRC | 16:03 | |
| *** martyntaylor has joined #openstack-ironic | 16:04 | |
| *** mordred has joined #openstack-ironic | 16:04 | |
| devananda | lucasagomes: validate() :) | 16:04 |
| devananda | also, states/console should indeed indicate whether console is on/off | 16:05 |
| lucasagomes | devananda, yea | 16:05 |
| lucasagomes | today we have GET /nodes/<uuid>/states | 16:06 |
| lucasagomes | which show the current states + last_error | 16:06 |
| lucasagomes | so as we r using last_error I thought about having a field indicating whether the console is on or off there | 16:06 |
| *** rloo has quit IRC | 16:08 | |
| *** rloo has joined #openstack-ironic | 16:08 | |
| devananda | not sure I follow | 16:09 |
| devananda | i would expect console to have an entry in /states just like power and deploy do | 16:09 |
| lucasagomes | ack yea that was the question heh having an entry or not :) | 16:10 |
| devananda | romcheg1: i've seen you post a few comments asking for LOG(_('message %s') % var) to be converted to LOG(_('message %s'), var) | 16:10 |
| devananda | romcheg1: is there a particular reason or precedent for this? | 16:10 |
| *** mrda_away is now known as mrda | 16:11 | |
| mrda | morning all! | 16:11 |
| *** anniec has joined #openstack-ironic | 16:11 | |
| devananda | mornin! | 16:11 |
| romcheg1 | devananda: I've seen a lot of comments like that. That is the main reason I think. I could not find that in the Hacking but for me it looks better | 16:11 |
| NobodyCam | morning mrda | 16:12 |
| devananda | romcheg1: ok, i would prefer not to spend review cycles on individual stylistic preferences -- we all have them, and they're all different :) | 16:12 |
| *** nosnos has quit IRC | 16:12 | |
| max_lobur_afk | back, but have to run again :/ what a day | 16:13 |
| devananda | romcheg1: if it's not enforced by flake8 / hacking, we shouldn't really worry about it. or if it's worth fixing, then pls add it to hacking so that it doesn't take human eyes spotting non-compliance | 16:13 |
| max_lobur_afk | will join later from home | 16:13 |
| devananda | cause we'll invariably miss a few ... that's what flake is for :) | 16:13 |
| romcheg1 | devananda: There were a bigger problem. Otherwise I would not mentioned and -1ed the patch | 16:14 |
| rloo | devananda, romcheg1. icbw, but I thought someone mentioned that using string interpolation (%) was less efficient, cuz it was always done even if the string wasn't going to be logged. | 16:14 |
| devananda | rloo: if folks are concerned about the performance impact of that, i'd expect it to end up in hacking sooner or later | 16:15 |
| *** anniec_ has joined #openstack-ironic | 16:16 | |
| *** anniec has quit IRC | 16:16 | |
| *** anniec_ is now known as anniec | 16:16 | |
| romcheg1 | devananda: I never -1 problems like PEP257 or interpolation in log messages, if there are no other problems. If the patch is going to get a new change set anyway, why not mention PEP257? | 16:16 |
| NobodyCam | oh sure | 16:16 |
| rloo | devananda, ido you know why lifeless asked you to make this change: https://review.openstack.org/#/c/70468/2/ironic/common/neutron.py | 16:16 |
| NobodyCam | :-p | 16:16 |
| devananda | rloo: hm, i thought it had to do with handling of LOG.exception, but IMBW too :) | 16:18 |
| devananda | nova has a mix of both % and , | 16:20 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 16:23 |
| NobodyCam | max_lobur_afk: romcheg1 lucasagomes devananda ^^^ please ... if you have time | 16:23 |
| *** rloo has quit IRC | 16:23 | |
| *** rloo has joined #openstack-ironic | 16:23 | |
| * devananda looks | 16:24 | |
| lucasagomes | NobodyCam, ack | 16:24 |
| NobodyCam | :) | 16:24 |
| devananda | NobodyCam: reviewed | 16:28 |
| NobodyCam | lol devananda -1 his own patch lol | 16:31 |
| devananda | heh | 16:31 |
| * NobodyCam waits to push up fixes until other look too... | 16:33 | |
| *** matty_dubs is now known as matty_dubs|lunch | 16:33 | |
| devananda | rloo: sorry for the confusion, I think it stemmed from my misunderstanding lifeless' comment | 16:38 |
| devananda | i've replied on that rev | 16:38 |
| rloo | thx devananda. I'm curious to know what he says. It may be that you interpreted his comment correctly ;) Will see... | 16:39 |
| *** BLZbubba has joined #openstack-ironic | 16:40 | |
| *** yuriyz has quit IRC | 16:41 | |
| *** aignatov is now known as aignatov_ | 16:44 | |
| *** coolsvap1 has quit IRC | 16:44 | |
| NobodyCam | oh patch set #2 | 16:45 |
| NobodyCam | :-p | 16:45 |
| devananda | hehehe | 16:45 |
| NobodyCam | mornig walkies.. bbaifm | 16:48 |
| BLZbubba | hi guys, n00b here to the bare metal provisioning system(s) but I an very interesting in setting up a proof of concept | 16:53 |
| BLZbubba | just one quick question here, will it be possible to snapshot at least the /dev/sda ? | 16:54 |
| BLZbubba | if it loads on the local disk instead of that crazy & unusable pxe/boot/iscsi stack | 16:55 |
| BLZbubba | or could I use qlogic hba's mayabe | 16:55 |
| BLZbubba | maybe too :P | 16:55 |
| NobodyCam | BLZbubba: their have talks about suck plans but right now that is not possiable | 16:56 |
| NobodyCam | and Welcome to Ironic | 16:57 |
| BLZbubba | admittedly it is a tough problem to grab sda through ipmi | 16:57 |
| BLZbubba | or however else you might try it | 16:57 |
| BLZbubba | should i try ironic or the havana version? | 16:58 |
| BLZbubba | ironic is a great project name | 16:58 |
| NobodyCam | if you are setting up now. I would point you to Nova-Baremetal | 16:59 |
| NobodyCam | Thank you :) | 16:59 |
| NobodyCam | look for us after Icehouse release | 16:59 |
| *** rsacharya has joined #openstack-ironic | 17:00 | |
| BLZbubba | ok cool, i'll lurk here again after I have havana working | 17:00 |
| NobodyCam | :) | 17:00 |
| *** coolsvap has joined #openstack-ironic | 17:00 | |
| devananda | BLZbubba: nova-baremetal should be considered experimental, and is going to be deprecated in the Icehouse release (~2mo away) | 17:03 |
| devananda | BLZbubba: if you're looking to contribute, i'd say, just start with ironic | 17:03 |
| * lucasagomes thinking about the snapshot | 17:04 | |
| devananda | BLZbubba: otoh, if you want to install havana from packages and learn how to administer openstack in general, yea, go with havana & nova baremetal | 17:04 |
| lucasagomes | there's no hypervisor, so no snapshot | 17:04 |
| devananda | lucasagomes: we could do a cold snapshot -- boot into separate ramdisk | 17:04 |
| lucasagomes | devananda, yea | 17:04 |
| lucasagomes | well it's more a copy the disk into glance | 17:04 |
| lucasagomes | I'm fine calling it snapshot tho | 17:05 |
| devananda | right. that's a snapshot :) | 17:05 |
| lucasagomes | but it's not really | 17:05 |
| lucasagomes | :P | 17:05 |
| devananda | hehe | 17:05 |
| devananda | if the node uses a Cinder volume instead of local storage, then snapshotting that should be possible | 17:05 |
| lucasagomes | devananda, we could have a snapshot like fs snapshot | 17:05 |
| devananda | via the Cidner API | 17:05 |
| * NobodyCam pulls out his old norton ghost disk | 17:05 | |
| lucasagomes | or using devicemapper | 17:05 |
| lucasagomes | with the snapshot target | 17:05 |
| lucasagomes | devananda, ahh cinder sounds good | 17:06 |
| devananda | lucasagomes: devmapper stil lrequires stopping the instance, booting into a separate environment | 17:06 |
| lucasagomes | yea that's correct | 17:07 |
| *** dkehn__ is now known as dkehn_ | 17:08 | |
| romcheg1 | lucasagomes: around? | 17:09 |
| lucasagomes | romcheg1, yes | 17:09 |
| romcheg1 | Can you please take a look at test_maintenance_nodes_error https://review.openstack.org/#/c/63937/9/ironic/tests/api/test_nodes.py | 17:09 |
| romcheg1 | Ah, no | 17:10 |
| romcheg1 | N questions :) | 17:10 |
| lucasagomes | romcheg1, seems grand, since maintenance is a boolean | 17:11 |
| lucasagomes | blah is a wrong value inputed | 17:11 |
| lucasagomes | so 400 (badrequest) | 17:11 |
| romcheg1 | lucasagomes: For a few seconds I confused 4** and 5** | 17:11 |
| lucasagomes | ack :D | 17:11 |
| NobodyCam | bbt..brb | 17:11 |
| * lucasagomes always gets confused by http codes as well | 17:12 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: API: Expose a way to start/stop the console https://review.openstack.org/72998 | 17:17 |
| *** anniec has quit IRC | 17:20 | |
| *** tatyana has quit IRC | 17:25 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: API: Expose a way to start/stop the console https://review.openstack.org/72998 | 17:25 |
| *** jbjohnso has quit IRC | 17:29 | |
| *** athomas has joined #openstack-ironic | 17:29 | |
| NobodyCam | devananda: should I wait for lifeless's reply b4 push up new rev of neutron patch? | 17:32 |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic: Implement API to get driver properties https://review.openstack.org/73005 | 17:33 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove _load_one_plugin fallback https://review.openstack.org/73006 | 17:33 |
| devananda | NobodyCam: regarding interpolation of params in LOG? nope. either way is fine according to hacking | 17:35 |
| NobodyCam | :) | 17:35 |
| devananda | NobodyCam: stick with what ever you have for now. if we need , instead of % it should go in hacking | 17:35 |
| NobodyCam | :) | 17:36 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove _load_one_plugin fallback https://review.openstack.org/73006 | 17:36 |
| *** lucasagomes is now known as lucas-afk | 17:37 | |
| * lucas-afk gotta go get a haircut | 17:38 | |
| *** matty_dubs|lunch is now known as matty_dubs | 17:38 | |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Implement _update_neutron in PXE driver https://review.openstack.org/70468 | 17:38 |
| NobodyCam | lucas-afk: will you be back? | 17:38 |
| lucas-afk | NobodyCam, yup | 17:38 |
| NobodyCam | :) | 17:38 |
| lucas-afk | NobodyCam, it will be late, but I will read the scrollback | 17:38 |
| NobodyCam | lucas-afk: for me its just https://review.openstack.org/70468 | 17:38 |
| NobodyCam | that my project for the day | 17:39 |
| NobodyCam | (or until it lands) | 17:39 |
| lucas-afk | NobodyCam, ack, awesome yea I have to give neutron another go | 17:39 |
| lucas-afk | NobodyCam, thanks that! I will test it probably tomorrow | 17:39 |
| NobodyCam | :) | 17:40 |
| *** matsuhashi has quit IRC | 17:48 | |
| *** saju_m has quit IRC | 17:50 | |
| *** saju_m has joined #openstack-ironic | 17:51 | |
| *** matsuhashi has joined #openstack-ironic | 17:51 | |
| *** matsuhashi has quit IRC | 17:55 | |
| *** saju_m has quit IRC | 17:58 | |
| *** saju_m has joined #openstack-ironic | 17:59 | |
| devananda | romcheg1: any progress on the db migration tool? | 18:01 |
| *** pleia2 has quit IRC | 18:02 | |
| *** mdurnosvistov_ has joined #openstack-ironic | 18:02 | |
| romcheg1 | devananda: I was going to publish it but discovered a conceptual problem | 18:03 |
| romcheg1 | Now it's almost ready | 18:03 |
| romcheg1 | to be published | 18:03 |
| *** pleia2 has joined #openstack-ironic | 18:04 | |
| devananda | romcheg1: I am going to send a status report today on our progress towards graduation | 18:04 |
| devananda | romcheg1: that BP is an essential piece, and it looks pretty bad to have no code up still ... | 18:04 |
| devananda | even with conceptual problems, that would still show progress :) | 18:04 |
| romcheg1 | That will be fixed this week | 18:04 |
| *** rsacharya has quit IRC | 18:12 | |
| *** hstimer has joined #openstack-ironic | 18:21 | |
| *** marty- has joined #openstack-ironic | 18:23 | |
| *** hstimer has quit IRC | 18:23 | |
| *** marty- has quit IRC | 18:24 | |
| *** hstimer has joined #openstack-ironic | 18:24 | |
| *** aignatov_ is now known as aignatov | 18:27 | |
| *** harlowja_away is now known as harlowja | 18:28 | |
| *** dkehn__ has joined #openstack-ironic | 18:32 | |
| *** davidlenwell_ is now known as davidlenwell | 18:32 | |
| *** dkehn_ has quit IRC | 18:35 | |
| *** dkehn has quit IRC | 18:35 | |
| *** aignatov is now known as aignatov_ | 18:36 | |
| lifeless | devananda: I don't think every API use can or should be resolved to a HACKING statement | 18:36 |
| lifeless | devananda: thats kindof saying 'don't do reviews, just check HACKING approves' | 18:37 |
| *** dkehn has joined #openstack-ironic | 18:37 | |
| *** dkehn__ is now known as dkehn_ | 18:37 | |
| NobodyCam | bbiafm... running to starbucks for real coffee | 18:39 |
| *** Alexei_987 has quit IRC | 18:49 | |
| *** saju_m has quit IRC | 18:51 | |
| *** saju_m has joined #openstack-ironic | 19:00 | |
| *** ndipanov has joined #openstack-ironic | 19:00 | |
| NobodyCam | ahh real coffee :) | 19:04 |
| NobodyCam | well almost real :-p | 19:04 |
| *** martyntaylor has left #openstack-ironic | 19:05 | |
| NobodyCam | ok rev 13 of https://review.openstack.org/#/c/70468 just passed mr J's inspection... | 19:32 |
| NobodyCam | max_lobur_afk: are you back? | 19:32 |
| devananda | NobodyCam: lgtm, but it's my patch so I won't +2 :p | 19:42 |
| NobodyCam | :) w00t | 19:43 |
| devananda | NobodyCam: also have you tested it with neutron to make sure it /actually/ works? :) | 19:43 |
| NobodyCam | devananda: can I get a +1 for the record | 19:43 |
| *** aignatov_ is now known as aignatov | 19:43 | |
| NobodyCam | just got my undercloud back up | 19:43 |
| devananda | NobodyCam: i couldn't get seed to work this morning, was going to test it ... | 19:43 |
| NobodyCam | working towards that atm | 19:44 |
| NobodyCam | also working on the patch bash script | 19:44 |
| NobodyCam | lol keep getting distracted | 19:44 |
| *** tatyana has joined #openstack-ironic | 19:47 | |
| *** saju_m has quit IRC | 19:48 | |
| *** max_lobur has joined #openstack-ironic | 20:19 | |
| max_lobur | back | 20:19 |
| NobodyCam | Wb max_lobur | 20:20 |
| max_lobur | thx, looking to https://review.openstack.org/#/c/70468 | 20:21 |
| NobodyCam | :) hehehe | 20:21 |
| NobodyCam | Ty max_lobur | 20:21 |
| NobodyCam | Side question anyone linaggo lately? | 20:22 |
| rloo | NobodyCam: have you tried sending her email? | 20:23 |
| NobodyCam | should have been a seen in htere | 20:23 |
| NobodyCam | no.. but I will now | 20:23 |
| max_lobur | NobodyCam: +2, ready for merge IMO | 20:34 |
| NobodyCam | max_lobur: w00t!!! | 20:34 |
| NobodyCam | now just need romcheg or lucas-afk to eye it over :) | 20:35 |
| NobodyCam | I think lucas-afk wants to test it localy | 20:35 |
| romcheg1 | NobodyCam: I will be available a bit later. Have a class now | 20:36 |
| NobodyCam | :) | 20:36 |
| NobodyCam | Ty romcheg1 :) | 20:36 |
| NobodyCam | stay in school | 20:36 |
| NobodyCam | :-p | 20:36 |
| NobodyCam | hehehe | 20:36 |
| max_lobur | :) | 20:37 |
| openstackgerrit | Sirushti Murugesan proposed a change to openstack/ironic: Adds support for deploying Disk Images https://review.openstack.org/73054 | 20:46 |
| *** harlowja is now known as harlowja_away | 20:48 | |
| devananda | shortstop: thanks! | 20:56 |
| *** tatyana has quit IRC | 20:56 | |
| *** jdob has quit IRC | 20:58 | |
| *** jdob has joined #openstack-ironic | 20:58 | |
| *** SpamapS_ is now known as SpamapS | 20:58 | |
| *** SpamapS has quit IRC | 20:58 | |
| *** SpamapS has joined #openstack-ironic | 20:58 | |
| shortstop | devananda, np :) | 21:00 |
| NobodyCam | humm I didn't get a ssh key for power on my undercloud | 21:02 |
| devananda | NobodyCam: see #tripleo. seems like several ppl have hit the same DBError today | 21:05 |
| devananda | i dunno about the ssh key tho | 21:05 |
| NobodyCam | atp i'm thinking its something I did incorrectly | 21:06 |
| *** jdob has quit IRC | 21:08 | |
| *** jdob has joined #openstack-ironic | 21:08 | |
| openstackgerrit | A change was merged to openstack/ironic: SSHPower driver support VMware ESXi https://review.openstack.org/64542 | 21:11 |
| *** derekh has joined #openstack-ironic | 21:13 | |
| devananda | just in case some of you don't often look at the mailing list, there are two emails up that i'd like to draw ya'll's attention to... | 21:15 |
| *** ndipanov has quit IRC | 21:15 | |
| max_lobur | :) | 21:22 |
| NobodyCam | well its not db err and i did include nova-ironic element in dib build line, however dib did not even try to include the nova-ironic element, didn't even log anything about it | 21:24 |
| rloo | devananda: what is involved in installation/deployer docs for ironic? can we copy/paste some existing docn? | 21:26 |
| devananda | rloo: what's needed is akin to this ... | 21:27 |
| * devananda hunts for the link | 21:27 | |
| devananda | something like this for all the config options | 21:28 |
| devananda | http://docs.openstack.org/havana/config-reference/content/ch_configuring-openstack-compute.html | 21:28 |
| devananda | and something like this in general for running / administering ironic | 21:29 |
| devananda | http://docs.openstack.org/admin-guide-cloud/content/ch_introduction-to-openstack-compute.html | 21:29 |
| rloo | eek. hmm. devananda. when does this doc need to be done? | 21:29 |
| devananda | by when ever the graduation deadline is (which i'm not exactly clear on, tbh, but it's before the summit) | 21:30 |
| devananda | lemme ping ttx real quick | 21:30 |
| rloo | ok, would be good if this isn't needed til after icehouse 3 or whenever the last code freeze is. | 21:30 |
| devananda | right | 21:31 |
| devananda | i'm sure it's after i3 | 21:31 |
| devananda | one of the points of the FF period is to give time to documentation team :) | 21:31 |
| devananda | i'm just not sure on the exact date | 21:31 |
| rloo | so the documentation team will write this docn then? | 21:31 |
| devananda | well... no, heh | 21:31 |
| devananda | it's a bit of chicken-and-egg. openstack-doc team doesn't have a responsibility to write docs for non integrated project | 21:32 |
| devananda | but we have a responsibility to provide docs | 21:32 |
| * devananda grumbles at the number of chickens and eggs | 21:32 | |
| rloo | ha ha. i can help with it i guess if no one else volunteers. Let me know when you're desparate. (I don't like writing docn.) | 21:32 |
| NobodyCam | thinks of chicken and waffles ... | 21:33 |
| JoshNang | hey guys, i'm working on https://bugs.launchpad.net/ironic/+bug/1272185. I'm running into the issue that if the partial uuid is an integer, add_identity_filter() in db/sqlalchemy/api.py will assume the uuid is a DB id. thoughts? | 21:38 |
| JoshNang | (because add_identify_filter can filter based on id or uuid) | 21:39 |
| max_lobur | JoshNang: Hi! | 21:40 |
| JoshNang | max_lobur: hi! | 21:40 |
| max_lobur | thanks for your efforts | 21:40 |
| max_lobur | I thought about where uuid like statement to solve this | 21:41 |
| max_lobur | but currently I don't see how to distinguis id and uuid prefix | 21:41 |
| max_lobur | we probably need some bigger refactoring to achieve this | 21:41 |
| max_lobur | also I've put those bug to wish list | 21:41 |
| max_lobur | this means we probably won't pay much attention to the patch now.. | 21:42 |
| JoshNang | that's fine. I was hoping it'd be an easy one to fix quickly and move on. a good way to get familiar with the code base. | 21:42 |
| max_lobur | lemme see if I find better bug for this | 21:43 |
| devananda | max_lobur: why do you think a partial uuid is valid at all? | 21:43 |
| max_lobur | do you mean we would need to break our validation to acheive this? | 21:44 |
| max_lobur | "where uuid startswith" I meant. | 21:45 |
| *** dkehn__ has joined #openstack-ironic | 21:46 | |
| russell_h | allowing a partial UUID seems a little odd | 21:46 |
| russell_h | although I've always appreciated short git hashes | 21:46 |
| devananda | short SHA is one thing | 21:47 |
| max_lobur | russell_h: + | 21:47 |
| max_lobur | I thought about git as well | 21:47 |
| devananda | truncated UUID is not a UUID | 21:47 |
| devananda | UUID has a very specific format | 21:47 |
| *** dkehn__ has quit IRC | 21:48 | |
| *** dkehn__ has joined #openstack-ironic | 21:48 | |
| *** dkehn has quit IRC | 21:48 | |
| *** dkehn_ has quit IRC | 21:49 | |
| JoshNang | right, I was looking at it as a way to make the CLI a bit more friendly, like the bug says. | 21:50 |
| devananda | JoshNang: adding names / aliases to Nodes would accomplish that | 21:50 |
| russell_h | if this happens, it feels like it needs to happen at the API layer | 21:50 |
| *** dkehn has joined #openstack-ironic | 21:50 | |
| russell_h | otherwise you can't use an index | 21:50 |
| devananda | I just commented on the bug. | 21:50 |
| max_lobur | https://bugs.launchpad.net/ironic/+bug/1277062 devananda , have you seen this | 21:50 |
| devananda | max_lobur: ahh good point, yes | 21:54 |
| devananda | max_lobur: see my emails to the dev list a bit ago? :) | 21:54 |
| max_lobur | seem no, about what? | 21:55 |
| max_lobur | devananda: are we ok to give https://bugs.launchpad.net/ironic/+bug/1277062 to JoshNang? | 21:56 |
| devananda | max_lobur: subject: [Ironic] [TripleO] Goal setting // progress towards integration | 21:57 |
| devananda | it pertains to 1277062 ... | 21:57 |
| max_lobur | ah, of course :) | 21:59 |
| *** pleia2 has quit IRC | 22:01 | |
| *** pleia2 has joined #openstack-ironic | 22:07 | |
| max_lobur | devananda: I understand this is low priority, but he wants a low-hanging-fruit to jump in, also if we would have review sessions like you proposed it will probably take 10 minutes of our common time to review it and merge | 22:07 |
| max_lobur | Everyone, how about to have the first review session tomorrow ? | 22:07 |
| NobodyCam | can we push 8 to 8:30 | 22:08 |
| NobodyCam | :-p | 22:08 |
| NobodyCam | more time for coffee | 22:08 |
| max_lobur | NobodyCam: that's OK to me as well :) | 22:09 |
| NobodyCam | :) | 22:10 |
| devananda | max_lobur: yep, makes sense | 22:10 |
| max_lobur | cool, JoshNang, would you please take https://bugs.launchpad.net/ironic/+bug/1277062 ? | 22:10 |
| *** lucas-afk is now known as lucasagomes | 22:11 | |
| JoshNang | max_lobur: certainly! | 22:11 |
| max_lobur | If you push patch today than it's highly probably it will be reviewed tomorrow :) | 22:11 |
| max_lobur | thanks :) | 22:11 |
| devananda | max_lobur: hm... i think we should start at the end of the queue | 22:12 |
| devananda | it might take more than 3 hours to work through it :p | 22:12 |
| devananda | afk a few... bbiah or so | 22:13 |
| max_lobur | :) | 22:13 |
| max_lobur | I hope there are a lot of patches already +1 and +2ed, so we'll spend no more than 10 minutes on them :) | 22:14 |
| lucasagomes | :) yea nice idea to dedicate a time to review stuff | 22:16 |
| lucasagomes | NobodyCam, cool, btw, the neutron patch... it looks ok I'm grand +a that | 22:16 |
| lucasagomes | have you tested on ur environment? | 22:17 |
| openstackgerrit | Ruby Loo proposed a change to openstack/python-ironicclient: Fix Iterface misspelling from node-validate cmd https://review.openstack.org/73082 | 22:18 |
| NobodyCam | lucasagomes: not yet... been tring | 22:24 |
| NobodyCam | :-p | 22:24 |
| NobodyCam | to many things | 22:24 |
| lucasagomes | NobodyCam, hah yea | 22:24 |
| lucasagomes | NobodyCam, np, I will give it a shot tomorrow morning | 22:24 |
| lucasagomes | but code-wise it looks good | 22:24 |
| NobodyCam | awesome !!!! | 22:25 |
| romcheg1 | lucasagomes, devananda: Did we came up with a consensus about this one https://review.openstack.org/#/c/71212 ? | 22:27 |
| lucasagomes | romcheg1, hmm idk really, but I liked the idea of using DELETE | 22:28 |
| lucasagomes | and suggested that on the review as well | 22:28 |
| romcheg1 | I cannot remember this topic was discussed on the meeting, although it was planned | 22:28 |
| lucasagomes | romcheg1, yea it wasn't | 22:28 |
| lucasagomes | but there are comments on that patch | 22:29 |
| *** athomas has quit IRC | 22:29 | |
| romcheg1 | lucasagomes: I like DELETE as well. But I think we need more opinions | 22:29 |
| lucasagomes | romcheg1, yeah | 22:29 |
| lucasagomes | that was mentioned as well | 22:29 |
| lucasagomes | the review itself will be a good way to get some opnions | 22:29 |
| lucasagomes | since now it's using the delete operation | 22:29 |
| romcheg1 | Yup, I just wanted to raise the question again :) | 22:30 |
| *** mdurnosvistov_ has quit IRC | 22:30 | |
| lucasagomes | romcheg1, ack, well delete is +1 for me I liked | 22:31 |
| lucasagomes | and saw some other apis also using it | 22:31 |
| lucasagomes | like etcd | 22:31 |
| lucasagomes | (using it to release locks) | 22:31 |
| *** matty_dubs is now known as matty_dubs|gone | 22:35 | |
| max_lobur | I like how it's implemented, but I'm not sure about the idea | 22:36 |
| max_lobur | as I recall there was agrs that reservation should be broken only by another conductor | 22:36 |
| max_lobur | that going to pick that node | 22:36 |
| lucasagomes | max_lobur, yea, automatic take over | 22:37 |
| *** jdob has quit IRC | 22:37 | |
| lucasagomes | which needs to be implemented as well | 22:37 |
| max_lobur | It might be misleading - to let user break the reservation, get control over the node, start some action .. | 22:37 |
| romcheg1 | That is quite dangerous | 22:37 |
| lucasagomes | +1 | 22:37 |
| max_lobur | and then conductor will take it back | 22:37 |
| lucasagomes | it's powerful indeed, and should be used with caution | 22:38 |
| max_lobur | yep, that's kinda hack Ironic | 22:38 |
| max_lobur | I tend to have some API gateway to force takeover | 22:38 |
| romcheg1 | If a conductor is not sending heartbeats that might mean that it's dead or that the network is dead but the conductor is performing some actions against the node | 22:38 |
| max_lobur | or like that | 22:38 |
| max_lobur | to push the whole mechanism | 22:39 |
| max_lobur | but not changing particular links | 22:39 |
| romcheg1 | In that case breaking the lock might lead to unpredictable consequences | 22:39 |
| lucasagomes | yea | 22:40 |
| lucasagomes | maybe it's not that useful to have in the api | 22:40 |
| lucasagomes | thinking that | 22:40 |
| lucasagomes | our api users are admins | 22:40 |
| lucasagomes | they also might have access to the db | 22:40 |
| max_lobur | yup | 22:40 |
| lucasagomes | and if the _really_ wants to break the lock | 22:40 |
| lucasagomes | they can go and update that field on the db | 22:40 |
| romcheg1 | Admins can at least monitor the state of the conductor | 22:40 |
| max_lobur | it's just a column in db | 22:41 |
| lucasagomes | don't need us to expose that via an api resource | 22:41 |
| romcheg1 | Automatic breaking system cannot | 22:41 |
| max_lobur | I thought it can.. isn't it | 22:41 |
| max_lobur | once heartbeat stopped - other conductor takes control | 22:42 |
| max_lobur | until the ring is fully rebalanced | 22:42 |
| max_lobur | though I'm not quite familiar with this part of code | 22:42 |
| romcheg1 | Then we have to make sure the old conductor is not doing anything to that node | 22:42 |
| romcheg1 | Two conductors will have conflict | 22:42 |
| lucasagomes | yea | 22:43 |
| lucasagomes | it's hard to map all the corner cases | 22:43 |
| lucasagomes | there's some cases where the conductor is not even dead | 22:43 |
| lucasagomes | but the network failed | 22:43 |
| lucasagomes | and a take over happened | 22:43 |
| romcheg1 | yeah | 22:43 |
| romcheg1 | It might be either problems with networking or heartbeat delivery | 22:44 |
| lucasagomes | yea | 22:44 |
| openstackgerrit | Ruby Loo proposed a change to openstack/python-ironicclient: Return node_uuid from a port-show cmd https://review.openstack.org/73091 | 22:45 |
| romcheg1 | Humans looks more reliable in that case | 22:45 |
| lucasagomes | heh | 22:46 |
| romcheg1 | They can call to the datacenter to be 100% sure the conductor is not working :D | 22:46 |
| lucasagomes | kinda, depends on how many things the human has to look at :P | 22:46 |
| lucasagomes | cloud is about automation | 22:46 |
| lucasagomes | so we gotta try to automate as much as can | 22:46 |
| romcheg1 | agree | 22:46 |
| lucasagomes | as we can* | 22:46 |
| max_lobur | lucasagomes: + | 22:46 |
| romcheg1 | and as safe as we can | 22:47 |
| lucasagomes | romcheg1, +1 | 22:47 |
| max_lobur | better to focus on takeover enhancement | 22:47 |
| romcheg1 | max_lobur: +1 | 22:47 |
| romcheg1 | Nope, +2 :) | 22:47 |
| max_lobur | +A | 22:47 |
| max_lobur | Starting gate jobs... | 22:47 |
| lucasagomes | yea! | 22:48 |
| lucasagomes | hah | 22:48 |
| openstackgerrit | A change was merged to openstack/ironic: Implement a multiplexed VendorPassthru example https://review.openstack.org/70863 | 22:53 |
| *** harlowja_away is now known as harlowja | 22:54 | |
| *** wusWWW has joined #openstack-ironic | 23:00 | |
| *** aignatov is now known as aignatov_ | 23:03 | |
| lucasagomes | devananda, saw ur email about the ephemeral disks bp | 23:04 |
| lucasagomes | I will fix the two patches I've tomorrow | 23:04 |
| *** rwsu has quit IRC | 23:06 | |
| *** rwsu has joined #openstack-ironic | 23:08 | |
| max_lobur | http://logs.openstack.org/79/71879/3/check/gate-ironic-python26/8f4ebfe LOST | 23:16 |
| max_lobur | gah | 23:16 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Removing unused get_node_power_state https://review.openstack.org/73097 | 23:19 |
| lucasagomes | max_lobur, heh will recheck | 23:22 |
| lucasagomes | btw, it's quite late here | 23:22 |
| lucasagomes | I'm gonna go :) | 23:22 |
| lucasagomes | have a good night everybody | 23:22 |
| *** lucasagomes has quit IRC | 23:23 | |
| max_lobur | night Lucas | 23:24 |
| NobodyCam | night lucas | 23:29 |
| NobodyCam | missed him :-p | 23:29 |
| openstackgerrit | A change was merged to openstack/ironic: Add parameter for filtering nodes by maintenance mode https://review.openstack.org/63937 | 23:30 |
| *** thedodd has quit IRC | 23:31 | |
| max_lobur | devananda: if you have time today, could you please take a look at the https://review.openstack.org/#/c/69135/ (race bug)? If you don't spot any *conceptual* problems I'll make the same for vendor action | 23:39 |
| max_lobur | i'm going sleep too, night folks | 23:41 |
| NobodyCam | night max_lobur | 23:42 |
| *** max_lobur has quit IRC | 23:42 | |
| *** derekh has quit IRC | 23:44 | |
| NobodyCam | brb... quick walkies | 23:56 |
| russell_h | I'm thinking about how I could make a deploy agent autodiscover its node_uuid | 23:57 |
| russell_h | we've got an existing system where we inventory the node and submit that to an API, which then finds a matching server in the database | 23:57 |
| russell_h | I'd like to do something similar here. It feels like there are two main options: I expose some sort of filtering on the nodes API, so the agent can try to find a matching node then use that | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!