| *** naohirot has joined #openstack-ironic | 00:01 | |
| *** dkingshott has quit IRC | 00:04 | |
| naohirot | good morning ironic | 00:04 |
|---|---|---|
| NobodyCam | morning naohirot :) | 00:04 |
| naohirot | NobodyCam: good evening :) | 00:05 |
| NobodyCam | :) | 00:05 |
| *** alexpilotti has quit IRC | 00:10 | |
| *** mlsoule has quit IRC | 00:12 | |
| rloo | hi NobodyCam -- just reading https://review.openstack.org/152734. seems like it should be two separate patches, one to revert, one to add the config? | 00:15 |
| NobodyCam | I thought about that but there was the config option we added withthe last patch so I just adjusted it | 00:16 |
| *** andreykurilin_ has quit IRC | 00:16 | |
| NobodyCam | and because I left all the logging the other patch added in it wasn't a real revert | 00:16 |
| NobodyCam | and dprince and chatted about the config option ^^^^ in scrollback | 00:17 |
| *** BadCub_ has quit IRC | 00:18 | |
| rloo | NobodyCam: oh. hmm. well, see what others think. to me, the config is separate but anyway. | 00:18 |
| NobodyCam | rloo: I can break them up | 00:18 |
| NobodyCam | :) | 00:18 |
| NobodyCam | but I'll do that tomorrow :-p | 00:18 |
| NobodyCam | hehehe | 00:18 |
| rloo | NobodyCam: I used to think it was ok, but somehow I seem to have been brainwashed. argh. | 00:19 |
| NobodyCam | :) | 00:19 |
| *** alex_xu_ has quit IRC | 00:23 | |
| *** davideagnello has joined #openstack-ironic | 00:24 | |
| *** davideagnello has quit IRC | 00:29 | |
| *** jerryz has joined #openstack-ironic | 00:38 | |
| *** chenleji has quit IRC | 00:40 | |
| *** rloo has quit IRC | 00:51 | |
| *** david-lyle is now known as david-lyle_afk | 00:57 | |
| openstackgerrit | Kyle Stevenson proposed openstack/ironic-python-agent: Add log and return results to IPA extensions https://review.openstack.org/152755 | 01:04 |
| openstackgerrit | Kyle Stevenson proposed openstack/ironic-python-agent: Add additional logging and return results for IPA extensions https://review.openstack.org/152755 | 01:06 |
| kylestev | JayF: ^ when you get a chance :) | 01:07 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add documentation on parallel argument for driver periodic tasks https://review.openstack.org/152756 | 01:11 |
| jroll | dtantsur|afk: ^ approved your patch and put this on top | 01:12 |
| jroll | kylestev: wrap your commit messages bro | 01:12 |
| kylestev | jroll: doing that atm lol | 01:12 |
| jroll | :P | 01:13 |
| jroll | 72 characters is the norm iirc | 01:13 |
| openstackgerrit | Kyle Stevenson proposed openstack/ironic-python-agent: Add additional logging and return results for IPA extensions https://review.openstack.org/152755 | 01:14 |
| jroll | whee | 01:14 |
| kylestev | ahhh much better | 01:14 |
| kylestev | russell_h: ^^^ btw | 01:18 |
| *** achanda_ has joined #openstack-ironic | 01:19 | |
| *** achanda has quit IRC | 01:22 | |
| *** achanda_ has quit IRC | 01:24 | |
| *** jmanko has joined #openstack-ironic | 01:38 | |
| *** rameshg87 has joined #openstack-ironic | 01:41 | |
| *** jmankov has quit IRC | 01:42 | |
| *** jmanko has quit IRC | 01:42 | |
| *** jmanko has joined #openstack-ironic | 01:42 | |
| *** anderbubble has joined #openstack-ironic | 01:45 | |
| *** Marga_ has quit IRC | 01:51 | |
| *** Marga_ has joined #openstack-ironic | 01:52 | |
| *** chenglch has joined #openstack-ironic | 01:52 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add documentation on parallel argument for driver periodic tasks https://review.openstack.org/152756 | 01:53 |
| *** Marga_ has quit IRC | 02:07 | |
| *** rwsu is now known as rwsu-afk | 02:11 | |
| openstackgerrit | Merged openstack/ironic: Add support for driver-specific periodic tasks https://review.openstack.org/147523 | 02:20 |
| *** yog_ has quit IRC | 02:31 | |
| *** ramineni has joined #openstack-ironic | 02:42 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: iRMC Virtual Media Deploy Driver for Ironic https://review.openstack.org/134865 | 02:44 |
| *** jerryz has quit IRC | 02:45 | |
| *** yog_ has joined #openstack-ironic | 02:45 | |
| *** jerryz has joined #openstack-ironic | 02:48 | |
| *** spandhe has quit IRC | 02:50 | |
| *** eghobo has quit IRC | 03:04 | |
| *** ijw has joined #openstack-ironic | 03:09 | |
| *** yog_ has quit IRC | 03:18 | |
| *** rameshg87 has quit IRC | 03:27 | |
| *** harlowja is now known as harlowja_away | 03:28 | |
| *** stendulker has joined #openstack-ironic | 03:29 | |
| *** anderbubble has quit IRC | 03:32 | |
| *** yog_ has joined #openstack-ironic | 03:34 | |
| openstackgerrit | jiangfei proposed openstack/python-ironicclient: cli support --os-endpoint https://review.openstack.org/152765 | 03:37 |
| *** jmankov has joined #openstack-ironic | 03:41 | |
| *** jmank has joined #openstack-ironic | 03:42 | |
| *** jmanko has quit IRC | 03:44 | |
| *** jmankov has quit IRC | 03:45 | |
| *** Haomeng|2 has joined #openstack-ironic | 03:50 | |
| *** Haomeng has quit IRC | 03:51 | |
| *** mitz_ has quit IRC | 03:56 | |
| *** yuanying has quit IRC | 03:56 | |
| *** ijw has quit IRC | 03:56 | |
| *** yuanying_ has joined #openstack-ironic | 03:56 | |
| *** mitz has joined #openstack-ironic | 03:56 | |
| *** ijw has joined #openstack-ironic | 03:58 | |
| *** achanda has joined #openstack-ironic | 04:20 | |
| *** achanda has quit IRC | 04:22 | |
| *** achanda has joined #openstack-ironic | 04:22 | |
| *** eghobo has joined #openstack-ironic | 04:59 | |
| *** Nisha has joined #openstack-ironic | 05:02 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: Add logical name support to Ironic https://review.openstack.org/141737 | 05:04 |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Initial Support to RIS https://review.openstack.org/152780 | 05:06 |
| *** davideagnello has joined #openstack-ironic | 05:09 | |
| *** david-lyle_afk has quit IRC | 05:11 | |
| *** achanda has quit IRC | 05:11 | |
| *** david-lyle_afk has joined #openstack-ironic | 05:11 | |
| *** harlowja_away has quit IRC | 05:19 | |
| *** davideagnello has quit IRC | 05:21 | |
| *** alex_xu has joined #openstack-ironic | 05:24 | |
| *** bradjones has quit IRC | 05:27 | |
| *** achanda has joined #openstack-ironic | 05:27 | |
| *** davideagnello has joined #openstack-ironic | 05:28 | |
| *** bradjones has joined #openstack-ironic | 05:30 | |
| *** davideagnello has quit IRC | 05:33 | |
| *** pradipta_away is now known as pradipta | 05:35 | |
| russell_h | kylestev: nice | 05:42 |
| openstackgerrit | Michael Davies proposed openstack/ironic: Add logical name support to Ironic https://review.openstack.org/141737 | 05:44 |
| *** eghobo has quit IRC | 05:44 | |
| *** killer_prince is now known as lazy_prince | 05:49 | |
| *** Marga_ has joined #openstack-ironic | 05:53 | |
| *** Marga_ has quit IRC | 05:53 | |
| *** Marga_ has joined #openstack-ironic | 05:54 | |
| *** coolsvap_ is now known as coolsvap | 05:57 | |
| *** achanda has quit IRC | 06:00 | |
| *** achanda has joined #openstack-ironic | 06:05 | |
| *** achanda has quit IRC | 06:13 | |
| *** jerryz has quit IRC | 06:34 | |
| *** ijw has quit IRC | 06:34 | |
| *** achanda has joined #openstack-ironic | 06:35 | |
| *** yog_ has quit IRC | 06:45 | |
| *** mrda is now known as mrda-away | 06:47 | |
| *** yog_ has joined #openstack-ironic | 06:54 | |
| *** chlong has joined #openstack-ironic | 07:06 | |
| *** yog_ has quit IRC | 07:09 | |
| *** ukalifon has joined #openstack-ironic | 07:09 | |
| *** Haomeng has joined #openstack-ironic | 07:20 | |
| *** Haomeng|2 has quit IRC | 07:20 | |
| *** Nisha has quit IRC | 07:21 | |
| *** dtantsur|afk is now known as dtantsur | 07:22 | |
| dtantsur | Morning Ironic | 07:22 |
| *** chlong has quit IRC | 07:28 | |
| *** yog_ has joined #openstack-ironic | 07:30 | |
| GheRivero | morning ironic | 07:34 |
| *** davideagnello has joined #openstack-ironic | 07:38 | |
| *** davideagnello has quit IRC | 07:43 | |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: Add hacking to test-requirements https://review.openstack.org/152817 | 07:46 |
| *** achanda has quit IRC | 07:46 | |
| dtantsur | GheRivero, morning | 07:54 |
| *** jcoufal has joined #openstack-ironic | 07:54 | |
| *** jerryz has joined #openstack-ironic | 07:57 | |
| *** achanda has joined #openstack-ironic | 08:01 | |
| openstackgerrit | Tan Lin proposed openstack/ironic: Update agent driver with new field driver_internal_info https://review.openstack.org/152055 | 08:08 |
| *** ifarkas has joined #openstack-ironic | 08:13 | |
| *** achanda has quit IRC | 08:26 | |
| *** stelfer has joined #openstack-ironic | 08:26 | |
| *** jistr has joined #openstack-ironic | 08:29 | |
| *** lucasagomes has joined #openstack-ironic | 08:29 | |
| *** kozhukalov has joined #openstack-ironic | 08:37 | |
| *** yuanying has joined #openstack-ironic | 08:47 | |
| *** yuanying_ has quit IRC | 08:50 | |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: Add hacking and fix pep8 issues https://review.openstack.org/152817 | 08:53 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor ilo.deploy._get_single_nic_with_vif_port_id() https://review.openstack.org/152506 | 09:00 |
| openstackgerrit | Zhenzan Zhou proposed openstack/ironic: Add policy show_password to mask passwords in driver_info https://review.openstack.org/150688 | 09:03 |
| *** athomas has joined #openstack-ironic | 09:08 | |
| dtantsur | ifarkas, FYI driver-specific periodic tasks were merged | 09:09 |
| ifarkas | dtantsur, excellent, thanks! | 09:10 |
| openstackgerrit | Marco Morais proposed openstack/ironic: Rename provision_state to power_state in test_manager.py https://review.openstack.org/152833 | 09:20 |
| kozhukalov | guys, am i right when think that Ironic does some background stuff using green threads and if I have many background tasks which are to be run at the same time (for example ssh some server and run some a little bit too long running command like 10 minutes) it is not a problem at all because green threads do cost almost nothing and potentially I could | 09:25 |
| kozhukalov | configure workers_pool_size=10000 instead of 100 by default? | 09:25 |
| dtantsur | I remember people having some problems with too larger workers pool, but I'm not aware of details | 09:26 |
| dtantsur | but yes, these are green threads | 09:26 |
| dtantsur | you better ask devananda when he's available | 09:27 |
| *** derekh has joined #openstack-ironic | 09:27 | |
| *** bauzas has quit IRC | 09:29 | |
| kozhukalov | dtantsur: thanx, devananda is now sitting next to me ) hope he'll find couple minutes to answer | 09:30 |
| dtantsur | heh that should work :) | 09:30 |
| *** bauzas has joined #openstack-ironic | 09:31 | |
| kozhukalov | im interesting in that in the context of using Ironic together with kind of agent which does not have any kind of async API but just a set of executable entry points and i am going to run them via ssh | 09:32 |
| *** igordcard has joined #openstack-ironic | 09:49 | |
| *** MattMan has joined #openstack-ironic | 09:49 | |
| *** jcoufal_ has joined #openstack-ironic | 09:52 | |
| *** jcoufal has quit IRC | 09:53 | |
| *** igordcard has quit IRC | 09:54 | |
| *** naohirot has quit IRC | 09:54 | |
| *** yog_ has quit IRC | 09:55 | |
| *** Haomeng|2 has joined #openstack-ironic | 09:58 | |
| *** Haomeng has quit IRC | 09:59 | |
| *** stelfer has quit IRC | 09:59 | |
| *** davideagnello has joined #openstack-ironic | 10:05 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Conductor errors if enabled_drivers are not found https://review.openstack.org/152844 | 10:08 |
| *** saripurigopi has joined #openstack-ironic | 10:08 | |
| *** andreykurilin_ has joined #openstack-ironic | 10:08 | |
| *** yog_ has joined #openstack-ironic | 10:09 | |
| openstackgerrit | Merged stackforge/proliantutils: Add hacking and fix pep8 issues https://review.openstack.org/152817 | 10:11 |
| *** dtantsur is now known as dtantsur|bbl | 10:13 | |
| *** dlpartain has joined #openstack-ironic | 10:21 | |
| *** mgoddard has joined #openstack-ironic | 10:34 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Initial Support to RIS https://review.openstack.org/152780 | 10:34 |
| *** Nisha has joined #openstack-ironic | 10:37 | |
| *** dlpartain has left #openstack-ironic | 10:39 | |
| *** chenglch has quit IRC | 10:40 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Initial Support to RIS https://review.openstack.org/152780 | 10:40 |
| *** chenleji has joined #openstack-ironic | 10:41 | |
| chenleji | anybody could show me more about network in ironic deployment? thx~ | 10:43 |
| *** yog_ has quit IRC | 10:44 | |
| *** athomas has quit IRC | 10:44 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add support for local boot https://review.openstack.org/146189 | 10:45 |
| *** alexpilotti has joined #openstack-ironic | 10:47 | |
| *** pelix has joined #openstack-ironic | 10:48 | |
| *** athomas has joined #openstack-ironic | 10:51 | |
| *** dtantsur|bbl is now known as dtantsur | 10:56 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: Add logical name support to Ironic https://review.openstack.org/141737 | 11:00 |
| Nisha | dtantsur, | 11:00 |
| Nisha | hi | 11:00 |
| Nisha | dtantsur, i am facing some rpc issue | 11:00 |
| Nisha | dtantsur, i am getting this error "Endpoint does not support RPC method inspect_hardware" | 11:01 |
| Nisha | it goes to rpcapi but is not able to resolve it to manager.py in conductor even though the function exists there | 11:02 |
| dtantsur | Nisha, something wrong with RPC versions? | 11:03 |
| dtantsur | did you update it in both places? | 11:03 |
| Nisha | yes | 11:03 |
| Nisha | i have updated RPC_API_VERSION correctly | 11:03 |
| dtantsur | Nisha, what's the patch? do you have it up to date now? I'll have a look once I finish with the current call | 11:04 |
| Nisha | It is not uptodate...i will just post the patches... | 11:04 |
| dtantsur | ok | 11:05 |
| *** ramineni has quit IRC | 11:05 | |
| *** yog_ has joined #openstack-ironic | 11:07 | |
| *** yog_ has quit IRC | 11:12 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add states required for node-inspection https://review.openstack.org/147857 | 11:17 |
| *** andreykurilin_ has quit IRC | 11:20 | |
| openstackgerrit | Tan Lin proposed openstack/ironic: Add AMT-PXE-Driver Common Library https://review.openstack.org/135184 | 11:22 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Add AMT-PXE-Driver Power&Management Interface https://review.openstack.org/152857 | 11:22 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Enable AMT driver to set boot device in pxe._continue_deploy. https://review.openstack.org/152858 | 11:22 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Add AMT-PXE driver doc https://review.openstack.org/152859 | 11:22 |
| lucasagomes | Nisha, hi there, any updates on https://review.openstack.org/#/c/141012/ ? | 11:30 |
| lucasagomes | can I update the patch myself? | 11:30 |
| Nisha | lucasagomes, on this patch we need to handle the operators as supported in extra_specs_op.py in nova/scheduler/filters/ | 11:33 |
| Nisha | so i did changes for others but how to handle <or> operator i was confused | 11:33 |
| Nisha | i will update the patch today | 11:33 |
| *** yog_ has joined #openstack-ironic | 11:36 | |
| *** saripurigopi has quit IRC | 11:39 | |
| Nisha | lucasagomes, please go ahead and update the patch. | 11:41 |
| Nisha | will u be handling the operators also? | 11:41 |
| *** stendulker has quit IRC | 11:49 | |
| *** Nisha_away has joined #openstack-ironic | 11:54 | |
| *** Nisha has quit IRC | 11:54 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Conductor errors if enabled_drivers are not found https://review.openstack.org/152844 | 11:57 |
| lucasagomes | Nisha_away, hi sorry for the delay. Right, hmm I wasn't planning on updating the next one | 12:01 |
| *** vdrok has joined #openstack-ironic | 12:01 | |
| *** david-lyle_afk is now known as david-lyle | 12:01 | |
| lucasagomes | Nisha_away, I will update the #141012 and ping some nova people see if I get someone to look at it | 12:01 |
| lucasagomes | because that's blocking the local boot stuff | 12:01 |
| Nisha_away | i am updating it but as of now not handling or operator... | 12:01 |
| lucasagomes | I can help out with the operator one too, but I've to update myself about it (as reading the spec and undestanding the use case, etc) | 12:01 |
| lucasagomes | Nisha_away, sure, right | 12:02 |
| Nisha_away | will update for that by night | 12:02 |
| lucasagomes | Nisha_away, right. Thanks! that should be quick update | 12:02 |
| *** lynxman has quit IRC | 12:02 | |
| lucasagomes | a quick* | 12:02 |
| lucasagomes | Nisha_away, thanks... | 12:02 |
| *** lucasagomes is now known as lucas-hungry | 12:03 | |
| lucas-hungry | I will grab some lunch | 12:03 |
| *** Nisha_away is now known as Nisha | 12:03 | |
| *** bradjones has quit IRC | 12:11 | |
| Nisha | lucasgomes i am getting following pep8 error for json.dumps jsonutils.dumps must be used instead of json.dumps | 12:12 |
| Nisha | lucas-hungry, ^^^^ | 12:12 |
| *** lynxman has joined #openstack-ironic | 12:13 | |
| *** alexpilotti has quit IRC | 12:13 | |
| *** bradjones has joined #openstack-ironic | 12:14 | |
| *** vdrok has quit IRC | 12:14 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add MANAGEABLE state and associated transitions https://review.openstack.org/150073 | 12:14 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add states required for node-inspection https://review.openstack.org/147857 | 12:14 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add InspectInterface for node-introspection https://review.openstack.org/149817 | 12:14 |
| dtantsur | Nisha, is it good that you keep rebasing https://review.openstack.org/#/c/150073/ ? | 12:15 |
| *** trown|call has joined #openstack-ironic | 12:16 | |
| *** vdrok has joined #openstack-ironic | 12:16 | |
| Nisha | dtantsur, means? | 12:16 |
| Nisha | for my testing i am not using 150073 | 12:16 |
| dtantsur | <openstackgerrit> Nisha Agarwal proposed openstack/ironic: Add MANAGEABLE state and associated transitions https://review.openstack.org/150073 | 12:16 |
| dtantsur | this ^^^ | 12:17 |
| dtantsur | it's possible you step on devananda's shoes when doing it | 12:17 |
| Nisha | i am just raising review for my changes which depends on the deva's patch | 12:17 |
| *** henry_chen has joined #openstack-ironic | 12:17 | |
| *** kevinbenton_ has joined #openstack-ironic | 12:17 | |
| Nisha | just raising git review is causing it | 12:18 |
| dtantsur | Nisha, you have to make sure your local copy of his patch is up-to-date before calling git reivew | 12:18 |
| *** rodrigod` has joined #openstack-ironic | 12:18 | |
| dtantsur | Nisha, always start with getting the latest patchset of this patch, then rebasing your work on top of it with 'git rebase -i' | 12:18 |
| henry_chen | who can tell me the network for ironic? | 12:18 |
| dtantsur | henry_chen, sorry, what do you mean by "network for ironic"? | 12:19 |
| henry_chen | I can't understand the traffic network and management network in ironic. | 12:19 |
| *** subscope has quit IRC | 12:19 | |
| *** Haomeng has joined #openstack-ironic | 12:19 | |
| Nisha | i did pull latest changes i.e. patch set 14 | 12:19 |
| *** Haomeng|2 has quit IRC | 12:20 | |
| *** subscope has joined #openstack-ironic | 12:20 | |
| *** yog_ has quit IRC | 12:20 | |
| *** erwan_taf has quit IRC | 12:20 | |
| *** trown has quit IRC | 12:20 | |
| *** kevinbenton has quit IRC | 12:20 | |
| *** rodrigods has quit IRC | 12:20 | |
| *** sbadia has quit IRC | 12:20 | |
| Nisha | dtantsur, i am following https://wiki.openstack.org/wiki/Gerrit_Workflow | 12:22 |
| *** rodrigod` is now known as rodrigods | 12:22 | |
| *** rodrigods has quit IRC | 12:22 | |
| *** rodrigods has joined #openstack-ironic | 12:22 | |
| henry_chen | dtantsur, I read the code, found ironic conductor use the "dd" command to copy the image into disk via management network. | 12:25 |
| *** yog_ has joined #openstack-ironic | 12:26 | |
| henry_chen | meanwhile, when the server is power on, does the physical server invoke the ironic API to enter venderpassthr? | 12:27 |
| *** sbadia has joined #openstack-ironic | 12:27 | |
| *** EmilienM|afk is now known as EmilienM | 12:27 | |
| dtantsur | henry_chen, you mean during deployment? yes, deploy ramdisk calls back Ironic once it prepared iscsi | 12:28 |
| *** erwan_taf has joined #openstack-ironic | 12:28 | |
| henry_chen | dtantsur, in the stage of deployment, the bare metal server must pass the management network. but how can i get the metadata info when the server booting. | 12:31 |
| henry_chen | i means they are in different network, the management network and traffic network. | 12:32 |
| dtantsur | henry_chen, probably your question is partly related to https://review.openstack.org/#/c/143480/ | 12:33 |
| henry_chen | dtantsur> thanks! i will read it righ now. | 12:34 |
| *** davideagnello has quit IRC | 12:34 | |
| *** david-lyle is now known as david-lyle_afk | 12:36 | |
| *** david-lyle_afk is now known as david-lyle | 12:36 | |
| *** david-lyle is now known as david-lyle_afk | 12:44 | |
| *** david-lyle_afk is now known as david-lyle | 12:45 | |
| *** vipul has quit IRC | 12:59 | |
| *** NobodyCam has quit IRC | 13:00 | |
| *** BadCub has quit IRC | 13:01 | |
| *** henry_chen has quit IRC | 13:01 | |
| lucas-hungry | Nisha, will take a look... | 13:01 |
| *** lucas-hungry is now known as lucasagomes | 13:02 | |
| *** vipul has joined #openstack-ironic | 13:03 | |
| *** naohirot has joined #openstack-ironic | 13:03 | |
| *** dprince has joined #openstack-ironic | 13:07 | |
| *** BadCub has joined #openstack-ironic | 13:07 | |
| *** NobodyCam has joined #openstack-ironic | 13:07 | |
| lucasagomes | Nisha, what is that scope = split() ? That wasn't part of the previous patch-set | 13:10 |
| lucasagomes | do we support space on the option value? | 13:10 |
| dtantsur | is it for operators? | 13:10 |
| lucasagomes | well that spec doesn't cover that | 13:11 |
| lucasagomes | afaik | 13:11 |
| *** Nisha has quit IRC | 13:13 | |
| *** pradipta has quit IRC | 13:13 | |
| *** rameshg87 has joined #openstack-ironic | 13:15 | |
| lucasagomes | dtantsur, I'm a bit lost actually... you know where is the spec for operators? | 13:16 |
| dtantsur | no idea, I got to know about their existence after your today's talk with Nisha :) | 13:16 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Enable AMT driver to set boot device in pxe deploy https://review.openstack.org/152858 | 13:18 |
| lucasagomes | dtantsur, heh yeah | 13:18 |
| lucasagomes | also this thing about having "operator value" | 13:18 |
| lucasagomes | separated by a space, it looks bad | 13:18 |
| dtantsur | lucasagomes, if it's something official in Nova, we have to support it | 13:19 |
| lucasagomes | I mean, having capabilities as string key:value,key:value... is already bad | 13:19 |
| dtantsur | otherwise we better no | 13:19 |
| lucasagomes | dtantsur, right yeah I need the spec to find that out really | 13:19 |
| lucasagomes | I think that for that patch, we should keep it minimum as was before | 13:19 |
| dtantsur | ++ | 13:19 |
| dtantsur | lucasagomes, related http://specs.openstack.org/openstack/nova-specs/specs/kilo/approved/add-all-in-list-operator-to-extra-spec-ops.html | 13:20 |
| lucasagomes | dtantsur, oh right... ok... that change should be part of that ^ spec then | 13:21 |
| lucasagomes | the scope() bit | 13:21 |
| dtantsur | still, I would love to see it split | 13:21 |
| lucasagomes | dtantsur, right, like into 2 patches one implementing "pass-flavor-capabilities ..." and another for the "add-all-in ..." ? | 13:24 |
| dtantsur | definitely | 13:24 |
| lucasagomes | or having that split() as part of the #141012? | 13:24 |
| lucasagomes | ok cool yeah +1 for that | 13:24 |
| lucasagomes | I will update the patch | 13:24 |
| *** Nisha has joined #openstack-ironic | 13:28 | |
| *** mjturek has joined #openstack-ironic | 13:28 | |
| rameshg87 | lucasagomes, dtantsur, can you please have a look at raid configuration spec: https://review.openstack.org/#/c/135899/ | 13:35 |
| lucasagomes | rameshg87, +1 | 13:35 |
| * lucasagomes do now | 13:35 | |
| *** chenleji has quit IRC | 13:35 | |
| rameshg87 | lucasagomes, thanks | 13:35 |
| dtantsur | will do after the call | 13:35 |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Correct the default log level https://review.openstack.org/152893 | 13:40 |
| *** mlsoule has joined #openstack-ironic | 13:43 | |
| *** trown|call is now known as trown | 13:43 | |
| *** jjohnson2 has joined #openstack-ironic | 13:47 | |
| *** rameshg87 has quit IRC | 13:49 | |
| *** rameshg87 has joined #openstack-ironic | 13:49 | |
| *** rloo has joined #openstack-ironic | 13:53 | |
| *** rameshg87 has quit IRC | 13:54 | |
| Nisha | lucasagomes, | 13:54 |
| lucasagomes | Nisha, I've updated the patch | 13:54 |
| lucasagomes | it didn't look right to implement 2 specs in 1 patch | 13:55 |
| Nisha | lucasagomes, i saw u updated the patch | 13:55 |
| Nisha | its not two specs | 13:55 |
| Nisha | the other spec is abandoned. | 13:55 |
| Nisha | This spec is to populate the flavor capabilities value into instance_info | 13:56 |
| Nisha | the other spec was to support multiple values of a capability i.e. say BootModes='uefi bios' | 13:57 |
| Nisha | lucasagomes, the other spec was to support multiple values of a capability i.e. say BootModes='uefi bios' in the node.properties so that if on flavor side one capability value is given i.e. either bios or uefi, then the node is selected for deploy | 13:58 |
| Nisha | lucasagomes, to supprt this nova has a syntax like extra_specs = {'capabilities:BootModes' : '<in> uefi'} | 13:59 |
| Nisha | lucasagomes, this format of flavor will select the node whose properties has capabilities as {'capabilities' : 'BootModes:bios uefi'} | 14:00 |
| Nisha | lucasagomes, that means when we copy the flavor values to instance_info we have to strip off the opertaors | 14:01 |
| Nisha | in above supported case the operator is <in> | 14:01 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add documentation on parallel argument for driver periodic tasks https://review.openstack.org/152756 | 14:01 |
| jroll | morning all :) | 14:01 |
| dtantsur | jroll, morning! | 14:01 |
| jroll | dtantsur: ^^ fixed that up | 14:01 |
| jroll | ohai :) | 14:01 |
| dtantsur | many thanks! | 14:02 |
| Nisha | lucasagomes, ^^^ | 14:02 |
| lucasagomes | Nisha, sorry, yeah reading | 14:02 |
| lucasagomes | Nisha, right... so in order to have that <in> to work we need the spec that supports multiple values of capabilities? | 14:04 |
| lucasagomes | which is now abandoned | 14:04 |
| lucasagomes | right? | 14:04 |
| Nisha | no no | 14:04 |
| lucasagomes | the other spec is abandoned.... he other spec was to support multiple values | 14:04 |
| lucasagomes | how many specs we are talking about? | 14:04 |
| lucasagomes | 3 | 14:04 |
| lucasagomes | ? | 14:04 |
| lucasagomes | Nisha, can you point me to all the specs that are approved (not abandoned) ? | 14:06 |
| Nisha | the spec which is abandoned was to chhose a node when flavor gives capabilities as capabilities:BootModes = 'uefi' | 14:06 |
| Nisha | and node has multiple values fr same capability | 14:07 |
| Nisha | in that case computcapbility filter does exact match | 14:07 |
| lucasagomes | right, so nodes can't have multiple vlaues for the same capability... | 14:08 |
| lucasagomes | so the test is that, if the node has capability X specified, and the flavor also has capability X we match it | 14:08 |
| Nisha | so now if we give in flavr as value as '<op> val' | 14:08 |
| Nisha | then <op> acts on the node capability values and selects if flavor val is in the node capability | 14:09 |
| lucasagomes | Nisha, oh right... so that operator is actually part of the ComputeCapabilitiesFilter ? | 14:09 |
| Nisha | yes | 14:10 |
| lucasagomes | gotcha | 14:10 |
| lucasagomes | ok... I completed misunderstood that, and the fact that none of the other patch-sets were tackling that operator use case | 14:10 |
| lucasagomes | makes it a bit more confusing for me | 14:10 |
| lucasagomes | Nisha, I will re-add that bit then | 14:10 |
| Nisha | so now it need to strip off operators | 14:10 |
| Nisha | thanks | 14:11 |
| lucasagomes | cool, yeah I see it | 14:11 |
| lucasagomes | Nisha, thanks for explaining | 14:11 |
| Nisha | it still need to handle <or> operator | 14:11 |
| lucasagomes | Nisha, right, do we care what is the operator? I mean, the ComputeCapabilitiesFilter does right? | 14:11 |
| Nisha | but we need to in ironic virt driver again | 14:12 |
| Nisha | bec filter just selects the node based on operator | 14:12 |
| Nisha | it doesnt strip off the oprator | 14:13 |
| *** stendulker has joined #openstack-ironic | 14:14 | |
| lucasagomes | Nisha, right... I'm not 100% sure how it will look like to work with <or> or any other operator | 14:15 |
| lucasagomes | so i will re-add that bit like before and mark the patch as WIP | 14:15 |
| lucasagomes | Nisha, you plan to continue working on the patch to add support for other operators soon? | 14:15 |
| Nisha | yes | 14:16 |
| lucasagomes | Nisha, cool thanks | 14:16 |
| Nisha | thanks | 14:17 |
| *** Marga_ has quit IRC | 14:17 | |
| dtantsur | Nisha, can we have 2 patches: what we have now and a separate patch for operators | 14:21 |
| dtantsur | don't really want to hold the future we need badly (local boot) due to feature I personally don't use at all (operators) | 14:21 |
| Nisha | dtantsur, but the first patch will be partial fix | 14:22 |
| dtantsur | well, ok | 14:22 |
| jroll | dtantsur: +1 | 14:23 |
| *** BadCub_ has joined #openstack-ironic | 14:23 | |
| BadCub_ | Morning Ironic | 14:23 |
| dtantsur | BadCub_, morning | 14:24 |
| jroll | heya BadCub_ :) | 14:24 |
| Nisha | dtantsur, i wil try to post the patch as soon as possible may be by today | 14:24 |
| dtantsur | Nisha, still, I think it should be 2 patches. | 14:25 |
| BadCub_ | heya jroll and dtantsur | 14:26 |
| Nisha | dtantsur, ok | 14:27 |
| *** chenleji has joined #openstack-ironic | 14:29 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Use prolianutils module for ilo driver tests https://review.openstack.org/151674 | 14:30 |
| NobodyCam | good morning Ironic | 14:31 |
| jroll | mornin NobodyCam :) | 14:31 |
| dtantsur | NobodyCam, morning | 14:33 |
| *** mgoddard has quit IRC | 14:40 | |
| *** lucasagomes has quit IRC | 14:42 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add iRMC Virtual Media Deploy module for iRMC Driver https://review.openstack.org/151958 | 14:42 |
| NobodyCam | morning jroll and dtantsur :) | 14:43 |
| *** lucasagomes has joined #openstack-ironic | 14:44 | |
| *** mgoddard has joined #openstack-ironic | 14:45 | |
| rloo | jroll: just saw your comment for 150688. 'bump the API version'. You mean the microversion? | 14:45 |
| jroll | rloo: yes | 14:45 |
| rloo | jroll: to be honest, the microversioning (use of etc) isn't clear to me. Are we following the nova spec for that? (I only skimmed it. maybe I should read it.) | 14:46 |
| NobodyCam | morning rloo :) | 14:46 |
| rloo | morning NobodyCam, jroll, and everyone else! | 14:46 |
| NobodyCam | :) | 14:47 |
| jroll | rloo: basically, they have extra things for api extensions and whatnot that we don't care about | 14:47 |
| rloo | jroll: does it describe when we up the microversion? | 14:48 |
| rloo | jroll: is it only for API changes, or any change/feature? | 14:48 |
| rloo | jroll: can we add a spec referring to nova spec, on how we're going to use microversioning? | 14:48 |
| jroll | rloo: should be just API changes, which also includes adding a db column etc | 14:49 |
| jroll | rloo: a spec pointing at a spec? | 14:49 |
| rloo | jroll: so this is a change to policy.json. Or do you mean cuz the password gets bleeped out, it affects the API so it is an api change? | 14:49 |
| jroll | right | 14:49 |
| jroll | the response may be different | 14:49 |
| rloo | jroll: adding microversion seems like a big thing (ie a feature) and we don't have a spec for it :-( | 14:49 |
| rloo | jroll: it doesn't hurt to up the microversion if we aren't sure if we should or not? | 14:50 |
| jroll | rloo: no, it doesn't hurt, it only is a pain point if the minimum allowed version changes afaict | 14:51 |
| jroll | (though we should try to only bump it when there's an actual change) | 14:52 |
| rloo | jroll: i was thinking we'd add something to bump the microversion for each merged patch, just in case :D | 14:52 |
| jroll | >.> | 14:52 |
| *** derekh has quit IRC | 14:53 | |
| *** absubram has left #openstack-ironic | 14:59 | |
| * naohirot good night > all | 14:59 | |
| *** naohirot has quit IRC | 14:59 | |
| NobodyCam | night naotokl | 14:59 |
| NobodyCam | doh | 14:59 |
| NobodyCam | tab fail | 14:59 |
| * NobodyCam has lost his coffee.. | 15:00 | |
| *** derekh has joined #openstack-ironic | 15:01 | |
| kylestev | to rerun tests on review.openstack.org do I enter a comment starting with “recheck”? | 15:04 |
| jroll | yep | 15:04 |
| openstackgerrit | Merged openstack/ironic-specs: Update amt-pxe-driver spec https://review.openstack.org/141269 | 15:04 |
| kylestev | gah, alright. for some reason I wrote “!recheck” earlier… derp | 15:05 |
| jroll | lol | 15:06 |
| kylestev | :P alright bbiab | 15:06 |
| NobodyCam | lucasagomes: have a quick second | 15:07 |
| *** ijw has joined #openstack-ironic | 15:09 | |
| *** jcoufal has joined #openstack-ironic | 15:09 | |
| *** zz_jgrimm is now known as jgrimm | 15:09 | |
| *** jcoufal_ has quit IRC | 15:12 | |
| *** EmilienM has quit IRC | 15:14 | |
| *** EmilienM has joined #openstack-ironic | 15:16 | |
| lucasagomes | NobodyCam, hey, hmm we are leaving here now | 15:19 |
| lucasagomes | NobodyCam, what's up? | 15:19 |
| NobodyCam | lucasagomes: ack enjoy | 15:19 |
| NobodyCam | oh I have a idea (from one of the internal guys) | 15:19 |
| NobodyCam | on the iSCSI stuff | 15:20 |
| lucasagomes | NobodyCam, right | 15:20 |
| NobodyCam | the suggestion was to check /dev/disk/by-ip/ | 15:20 |
| NobodyCam | I may try that today | 15:20 |
| lucasagomes | NobodyCam, right, yeah I think it would be good to tackle the problem | 15:20 |
| lucasagomes | instead of working around it again | 15:20 |
| lucasagomes | NobodyCam, if we want to make it very safe, I'm also fine with the sleep(10) | 15:21 |
| NobodyCam | ya :) I could tell by the comments :-p | 15:21 |
| *** PaulCzar has joined #openstack-ironic | 15:21 | |
| *** Nisha has quit IRC | 15:21 | |
| lucasagomes | NobodyCam, cool, thanks for looking into it | 15:21 |
| NobodyCam | :) | 15:21 |
| NobodyCam | enjoy you night lucasagomes :) | 15:22 |
| lucasagomes | NobodyCam, thanks! I ttyl | 15:22 |
| NobodyCam | :) | 15:22 |
| *** lucasagomes has quit IRC | 15:22 | |
| *** mgoddard has quit IRC | 15:23 | |
| rloo | hmm, sphinx seems to be complaining. i am wondering if some package got updated. | 15:24 |
| rloo | http://logs.openstack.org/88/150688/9/check/gate-ironic-docs/3580d47/ | 15:24 |
| rloo | if anyone else is seeing that error, we should probably look into it... | 15:25 |
| rloo | dtantsur: you around? wrt 152756, the parallel=True in the decorator. I meant that in base.py, we should elaborate on the effects of parallel=False. | 15:27 |
| rloo | dtantsur: but I can do that. A simple copy/paste from jroll's excellent documentation. | 15:28 |
| dtantsur | rloo, hi! makes sense to me (though we might change it when switching to Oslo) | 15:28 |
| rloo | dtantsur: yeah, that's what I wanted to know, how it was going to change in oslo. but then i thought we don't know when that might happen | 15:28 |
| rloo | dtantsur: so probably best to document how it works 'now'. | 15:29 |
| dtantsur | unfortunately we don't, you're right | 15:29 |
| dtantsur | yeah, let's do | 15:29 |
| rloo | dtantsur: you want me to do it or do you want to? | 15:29 |
| dtantsur | rloo, please do, if you have some time :) thanks | 15:29 |
| rloo | dtantsur: yeah, just about to do it, so I'll do it :-) | 15:30 |
| *** EmilienM is now known as EmilienM|afk | 15:35 | |
| *** stendulker_ has joined #openstack-ironic | 15:37 | |
| *** stendulker has quit IRC | 15:39 | |
| *** ChuckC has joined #openstack-ironic | 15:40 | |
| *** ChuckC_ has quit IRC | 15:41 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic: update docstring for driver_periodic_task's parallel param https://review.openstack.org/152956 | 15:42 |
| *** davideagnello has joined #openstack-ironic | 15:46 | |
| dtantsur | agent_ssh is voting now! | 15:49 |
| NobodyCam | w00t | 15:50 |
| *** digimonde has joined #openstack-ironic | 15:52 | |
| rloo | whoa, have we made it to the big times? :-) | 15:53 |
| rloo | we need an agent_007 driver... | 15:54 |
| dtantsur | oh, awesome idea! | 15:54 |
| dtantsur | we need to find a vendor with something named 007... | 15:54 |
| NobodyCam | rloo: I love it | 15:54 |
| NobodyCam | !!! | 15:54 |
| openstack | NobodyCam: Error: "!!" is not a valid command. | 15:54 |
| dtantsur | lol | 15:54 |
| rloo | :D | 15:55 |
| dtantsur | !such command very refure not wow | 15:55 |
| openstack | dtantsur: Error: "such" is not a valid command. | 15:55 |
| NobodyCam | oh does the agent support "BOND"ing interfaces | 15:55 |
| NobodyCam | and I guess we need to change all the ipa folks name to James | 15:56 |
| rloo | +1 for James and BOND | 15:57 |
| NobodyCam | :-p | 15:59 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: update docstring for driver_periodic_task's parallel param https://review.openstack.org/152956 | 16:03 |
| *** Marga_ has joined #openstack-ironic | 16:03 | |
| jlvillal | rloo: On https://review.openstack.org/152956 What are cases where people would/should use parallel=False? | 16:05 |
| jlvillal | I was wondering if examples for the 'parallel=False' case would be useful. | 16:06 |
| rloo | jlvillal: none, as far as I am concerned. but dtantsur is the one that added the code, I'm just a concerned citizen ;) | 16:06 |
| jlvillal | rloo: Thanks :) | 16:06 |
| rloo | jlvillal: I wanted that there so that reviewers would be reminded to question any use of False ;) | 16:07 |
| jroll | NobodyCam: lol, it will :D | 16:07 |
| NobodyCam | nice James | 16:07 |
| NobodyCam | lol | 16:07 |
| NobodyCam | heheheheh | 16:07 |
| jroll | (support bonding) | 16:07 |
| jroll | :P | 16:07 |
| jroll | I guess it's the network driver that does that but hey, close enough | 16:08 |
| NobodyCam | :p | 16:08 |
| *** ChuckC_ has joined #openstack-ironic | 16:14 | |
| *** ChuckC has quit IRC | 16:17 | |
| *** EmilienM|afk is now known as EmilienM | 16:18 | |
| *** ukalifon has quit IRC | 16:23 | |
| jlvillal | When doing an 'importutils.try_import('blah.blah") Should the return value be checked? Or is it okay to assume success? | 16:23 |
| *** erwan_taf has quit IRC | 16:24 | |
| *** coolsvap is now known as coolsvap_ | 16:28 | |
| rloo | jlvillal: taking a look at the code indicates that it doesn't necessarily return success | 16:29 |
| rloo | jlvillal: https://github.com/openstack/oslo.utils/blob/master/oslo_utils/importutils.py#L68 | 16:29 |
| jlvillal | rloo: Thanks. I was doing a review and notice most times the return is checked. But this instance it was not. I will add a comment. | 16:30 |
| rloo | jlvillal: thx! | 16:31 |
| jlvillal | rloo: Maybe this is a bike shedding question :) Do we care about 'if (not foo):' vs 'if not foo:' I personally prefer with out the parenthesis for simple case. But wondering if that is just up to each user to decide. | 16:31 |
| * jlvillal assumes each user... | 16:32 | |
| rloo | jlvillal: personally, i don't care. i don't know what others think. | 16:32 |
| jlvillal | Thanks I will ignore :) | 16:32 |
| rloo | jlvillal: you can mention it as a nit i guess. or someone else could/might. i agree it is cleaner w/o () | 16:33 |
| dtantsur | I pretty much hate redundant brackets, but usually don't -1 only for them | 16:33 |
| jlvillal | rloo: Thanks | 16:33 |
| rloo | jlvillal: see, a nit then ;) | 16:33 |
| jlvillal | Okay, I'll mention it once... | 16:33 |
| *** ijw has quit IRC | 16:36 | |
| *** Marga_ has quit IRC | 16:39 | |
| *** rwsu-afk is now known as rwsu | 16:40 | |
| jlvillal | rloo: I will work on my very first patch to say that 'spacing=5' means every 5 seconds... I will wait for that other patch to get merged though. | 16:42 |
| rloo | thx jlvillal | 16:44 |
| stendulker_ | NobodyCam: Hi | 16:50 |
| *** jerryz has quit IRC | 16:50 | |
| NobodyCam | morning stendulker_ :) | 16:50 |
| stendulker_ | NobodyCam: Can you please have a look at spec for Ironic Management Interfaces to support UEFI Secure Boot https://review.openstack.org/#/c/135845/ | 16:50 |
| NobodyCam | stendulker_: sure.. but its going to take me a few minutes to get to it :) | 16:51 |
| stendulker_ | NobodyCam: Had to post the patch to address comments from Ruby Loo. They are to her satisfaction and she has approved the spec | 16:51 |
| NobodyCam | :) | 16:52 |
| NobodyCam | anyone know if this is a valid mock? @mock.patch.object(os, 'path.exists') | 16:52 |
| stendulker_ | NobodyCam: Thank you :) | 16:52 |
| *** anderbubble has joined #openstack-ironic | 16:53 | |
| dtantsur | NobodyCam, @mock.patch.object(os.path, 'exists') | 16:53 |
| NobodyCam | :) TY dtantsur :) | 16:53 |
| lazy_prince | jroll: since the network provider interface spec has merged, are you planning to open reviews for https://review.openstack.org/#/c/139687/ anytime soon..? | 16:55 |
| *** davideagnello has quit IRC | 16:56 | |
| *** davideagnello has joined #openstack-ironic | 16:57 | |
| jlvillal | So how's the sprint going in Grenoble? | 17:00 |
| *** lazy_prince is now known as killer_prince | 17:00 | |
| *** dtantsur is now known as dtantsur|afk | 17:01 | |
| dtantsur|afk | g'night! | 17:02 |
| NobodyCam | night dtantsur|afk :) | 17:02 |
| rloo | night dtantsur|afk | 17:03 |
| *** kozhukalov has quit IRC | 17:03 | |
| *** chenleji has quit IRC | 17:05 | |
| *** absubram has joined #openstack-ironic | 17:07 | |
| *** rameshg87 has joined #openstack-ironic | 17:09 | |
| *** anderbubble has quit IRC | 17:15 | |
| *** anderbubble has joined #openstack-ironic | 17:16 | |
| *** jlvillal has quit IRC | 17:17 | |
| *** jlvillal has joined #openstack-ironic | 17:18 | |
| *** spandhe has joined #openstack-ironic | 17:20 | |
| *** Marga_ has joined #openstack-ironic | 17:21 | |
| *** ijw has joined #openstack-ironic | 17:23 | |
| *** athomas has quit IRC | 17:23 | |
| *** jistr has quit IRC | 17:25 | |
| *** jcoufal has quit IRC | 17:26 | |
| *** xxj has quit IRC | 17:29 | |
| *** xxj has joined #openstack-ironic | 17:29 | |
| jroll | killer_prince: I forgot that had a +2, thanks for the bump | 17:32 |
| jroll | s/+/-/ | 17:33 |
| *** GheRivero has quit IRC | 17:34 | |
| *** GheRivero has joined #openstack-ironic | 17:34 | |
| openstackgerrit | Merged openstack/ironic: Add documentation on parallel argument for driver periodic tasks https://review.openstack.org/152756 | 17:35 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: improve iSCSI connection check https://review.openstack.org/152734 | 17:38 |
| *** spandhe has quit IRC | 17:38 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic-specs: New driver interface for RAID configuration https://review.openstack.org/135899 | 17:40 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: improve iSCSI connection check https://review.openstack.org/152734 | 17:40 |
| rameshg87 | JoshNang, hi | 17:41 |
| JoshNang | rameshg87: o/ | 17:41 |
| rameshg87 | JoshNang, had a question on zapping | 17:41 |
| JoshNang | ask away! | 17:41 |
| rameshg87 | JoshNang, for a zap method which requires agent to be booted on the bare metal first | 17:42 |
| rameshg87 | JoshNang, are you planning to add some common utility method for it ? | 17:42 |
| rameshg87 | JoshNang, i mean it's a common problem that zap methods using agent should have ramdisk booted and heartbeating | 17:42 |
| rameshg87 | JoshNang, right ? | 17:42 |
| JayF | I'd assume it's "free" | 17:43 |
| JayF | i.e. if you direct something to do an in-band zapping action | 17:43 |
| JoshNang | ^ right | 17:43 |
| JayF | Ironic knows how to boot an agent on something, and will do it | 17:43 |
| JoshNang | you'll tell the api to do a zap step, api tells conductor, conductor tells agent driver, agent driver tells agent | 17:43 |
| rameshg87 | JayF, JoshNang, okay | 17:44 |
| rameshg87 | so what if for example node is powered off | 17:44 |
| JoshNang | it'll power on and boot the agent. | 17:45 |
| rameshg87 | agent driver will prepare pxe config, and power on the node | 17:45 |
| rameshg87 | and then on the next heart beat the method will be called, is it that way ? | 17:45 |
| JoshNang | sounds about right | 17:46 |
| rameshg87 | basically for a person writing inband agent zapping methods, they can directly start sending commands to agent, right ? | 17:46 |
| rameshg87 | i mean in the method, they can directly start sending commands to agent | 17:47 |
| jroll | I'm inclined to think that when a server is put in "zapping" state, it should just go ahead and boot the ramdisk | 17:47 |
| JayF | jroll: I disagree; because there are Zapping tasks that may be done out of band, and that may want to be done with a powered off machine | 17:48 |
| JayF | jroll: i.e. "Upgrade the firmware on my BMC via OOB" | 17:48 |
| JoshNang | i mean, you pass in the list of zap steps when you go to zapping | 17:48 |
| jroll | riiight. | 17:48 |
| JoshNang | the conductor will know at that point if it should boot the agent or not | 17:48 |
| jroll | yeah | 17:48 |
| JoshNang | rameshg87: what do you mean, inband agent zapping methods? | 17:48 |
| jroll | "step" authors shouldn't need to think much about if a ramdisk is up | 17:49 |
| JoshNang | the agent driver won't really have methods for zapping. it'll have the same execute_step function it uses in cleaning. the zap steps for the agent come from the agent. | 17:49 |
| rameshg87 | JoshNang, if i have a zap method for raid configuration | 17:49 |
| rameshg87 | JoshNang, i don't need to worry about if ramdisk is up or not. i can directly send the command to do the raid configuration (assuming ramdisk supports it) | 17:50 |
| rameshg87 | i guess this is what jroll too said above ^^ | 17:50 |
| jroll | rameshg87: the agent asks ironic which steps it needs to execute | 17:50 |
| JoshNang | right | 17:50 |
| jroll | or something | 17:50 |
| jroll | idk how anything works anymore | 17:50 |
| *** GheRivero has quit IRC | 17:51 | |
| *** GheRivero has joined #openstack-ironic | 17:51 | |
| JoshNang | rameshg87: you don't really send a command, or need to write anything in ironic to make agent steps work. the agent has a list of steps it can execute, those are exposed to the zapping api. if you want to add raid config to the agent, the code for all that goes into IPA | 17:51 |
| *** eghobo has joined #openstack-ironic | 17:51 | |
| rameshg87 | JoshNang, i think there is a small disconnect somewhere then .. | 17:52 |
| JoshNang | rameshg87: i think so too | 17:53 |
| rameshg87 | JoshNang, please have a look at https://review.openstack.org/#/c/135899/18 and https://review.openstack.org/#/c/147803/ | 17:53 |
| rameshg87 | JoshNang, drac wants to do raid configuration oob and ilo wants to do the raid configuration inband | 17:53 |
| JoshNang | rameshg87: ok, let me review these | 17:54 |
| JayF | Why is that a problem? | 17:54 |
| JayF | Conductor could still tell at time of zapping steps being passed in if it was inband or oob | 17:54 |
| rameshg87 | JayF, it's not a problem | 17:55 |
| rameshg87 | JayF, but i guess it will be different from other agent zapping tasks.. | 17:55 |
| JayF | How so? It can't be, can it? | 17:55 |
| JayF | There's one agent. | 17:55 |
| *** killer_prince is now known as lazy_prince | 17:55 | |
| JoshNang | yeah i'm really confused about this raidmanagement interface | 17:56 |
| rameshg87 | JayF, JoshNang, in parent spec https://review.openstack.org/#/c/135899 | 17:56 |
| *** spandhe has joined #openstack-ironic | 17:57 | |
| rameshg87 | we specified a single zap method 'raid.apply_configuration' | 17:57 |
| jroll | jlvillal: I guess you're john villalovos? I've started seeing you in reviews, welcome to the project :) | 17:57 |
| *** coolsvap_ is now known as coolsvap | 17:57 | |
| jlvillal | jroll: Yes I am :) Thanks! I just started this week. I'm looking forward to it. Trying to learn right now. | 17:57 |
| jroll | rameshg87: an entire new interface? another thing to mix in to drivers? :/ | 17:57 |
| rameshg87 | and drivers implementing RAIDInterface can implement the methods create_configuration | 17:58 |
| rameshg87 | jroll, yes .. | 17:58 |
| rameshg87 | jroll, please have a look at the spec https://review.openstack.org/#/c/135899 | 17:58 |
| jroll | jlvillal: awesome, glad to have you! where are you based / what's your goal with ironic? do you know about the midcycles happening this week and next? | 17:58 |
| jroll | rameshg87: I'm looking | 17:58 |
| rameshg87 | jroll, since there is interest with 2 vendors (drac and ilo) on the same thing, we thought of proposing an interface for it | 17:59 |
| JoshNang | i'm confused why this is it's own interface rather than zap steps on drac/agent/ilo existing interfaces. | 17:59 |
| jlvillal | jroll: Based in Hillsboro, Oregon. Intel guy. Goal is to help wherever I can. I am trying to figure out a way to attend the mid-cycle in SF. For some reason they said no to Grenoble ;) | 17:59 |
| NobodyCam | hey hey dprince you around? | 17:59 |
| jroll | rameshg87: why can't it just be part of the management interface? or zap steps stuff? | 17:59 |
| JoshNang | jlvillal: awesome! hope you can make the SF midcycle! | 18:00 |
| jroll | jlvillal: awesome! we (rackspace) have a couple guys from your neck of the woods | 18:00 |
| jlvillal | JoshNang: Thanks :) | 18:00 |
| dprince | NobodyCam: hi | 18:00 |
| NobodyCam | hey hey dprince :) | 18:01 |
| jlvillal | jroll: Sweet. Maybe I'll see them sometime. | 18:01 |
| rameshg87 | jroll, i agree management interface is an alternative. | 18:01 |
| *** lazy_prince is now known as killer_prince | 18:02 | |
| NobodyCam | would you be able to test: https://review.openstack.org/152734 on Fedora.. I only ubuntu here :( | 18:02 |
| *** derekh has quit IRC | 18:02 | |
| rameshg87 | ifarkas, are you around ? | 18:05 |
| JoshNang | i do like the additional api that this adds, but if we have a raid interface, next we'll have a firmware interface, etc | 18:07 |
| NobodyCam | huh? http://logs.openstack.org/34/152734/3/check-tripleo/check-tripleo-ironic-undercloud-precise-nonha/7507bcc/console.html#_2015-02-04_18_05_09_816 | 18:08 |
| rameshg87 | JoshNang, jroll, yeah i think it can be moved to management interface | 18:09 |
| jroll | rameshg87: woot. | 18:09 |
| jroll | NobodyCam: fun | 18:10 |
| NobodyCam | thats a new one at least for me | 18:10 |
| *** harlowja has joined #openstack-ironic | 18:11 | |
| rameshg87 | JoshNang, jroll, the extra api that it adds allows to do some validation on the input | 18:11 |
| jroll | NobodyCam: yeah, seems like it's not us breaking it though :) | 18:11 |
| *** dkingshott has joined #openstack-ironic | 18:11 | |
| jlvillal | NobodyCam: To test https://review.openstack.org/152734 , would I need a full OpenStack install. I'm not there yet... I have the developer setup with virtualenv. I'm not yet sure how many computers I need to setup a full OpenStack setup. | 18:11 |
| jroll | rameshg87: at input time? that's fair, though you could do some special casing in the PATCH code | 18:12 |
| NobodyCam | oh jlvillal that was directed ar dprince :) | 18:12 |
| rameshg87 | JoshNang, special casing ? what does it mean ? | 18:12 |
| jlvillal | NobodyCam: Okay :) | 18:12 |
| NobodyCam | but thank you :) | 18:12 |
| rameshg87 | jroll, i meant ^^ | 18:12 |
| rameshg87 | jroll, what does "special casing" mean ? | 18:13 |
| jroll | rameshg87: in the api for node-update, if key == '/raid_stuff': _validate_raid_things() | 18:13 |
| jroll | though I guess that would be driver-specific, would need to be in the conductor | 18:13 |
| dprince | NobodyCam: are you asking me to test this patch? | 18:13 |
| *** Nisha has joined #openstack-ironic | 18:14 | |
| rameshg87 | jroll, the spec proposes PUT /nodes/<uuid>/raid/configuration and GET /nodes/<uuid>/raid/configuration for the same | 18:14 |
| NobodyCam | dprince: yes ... if you have a Fedora setup. :) | 18:14 |
| rameshg87 | jroll, and a command line "ironic node-set-raid-configuration" | 18:14 |
| NobodyCam | dprince: I do not have real hardware | 18:15 |
| jroll | rameshg87: oh, right, we could still do that | 18:15 |
| rameshg87 | jroll, we can validate the json information passed and put it straight to node.driver_internal_info | 18:15 |
| jroll | yep | 18:15 |
| jroll | ++ | 18:15 |
| jroll | rameshg87: we can do that *and* keep it in ManagementInterface | 18:15 |
| dprince | NobodyCam: sure. If you could just block it until I get a chance to do that. Perhaps later today. Maybe tomorrow... | 18:15 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add InspectInterface for node-introspection https://review.openstack.org/149817 | 18:16 |
| rameshg87 | jroll, yeah keep those methods in ManagementInterface with base.py implementation raising NotImplementedError | 18:16 |
| NobodyCam | dprince: its WIP'd now I will not remove until hearing from you... Thank you :) | 18:16 |
| jroll | rameshg87: yep, awesome | 18:17 |
| rameshg87 | jroll, thank you | 18:17 |
| rameshg87 | JoshNang, thanks | 18:17 |
| JoshNang | np! | 18:17 |
| JoshNang | i'll leave a more detailed review on the raid management spec | 18:17 |
| rameshg87 | JoshNang, great .. thanks | 18:19 |
| * jlvillal starts watching https://www.youtube.com/watch?v=vjGPuBfZ87A :) | 18:19 | |
| *** coolsvap is now known as coolsvap_ | 18:20 | |
| rameshg87 | jroll, one question | 18:24 |
| jroll | rameshg87: shoot | 18:25 |
| rameshg87 | jroll, going through the spec https://review.openstack.org/#/c/135899 we have 4 methods proposed in RAIDInterface | 18:25 |
| rameshg87 | jroll, RAIDInterface.create_configuration(), RAIDInterface.delete_configuration(), RAIDInterface.get_logical_disk_properties() are supposed to be implemented by a driver interested in doing raid configuration | 18:26 |
| jroll | rameshg87: I see three, but yes, go on | 18:26 |
| rameshg87 | jroll, and RAIDInterface.apply_configuration() which will be the zapping task | 18:27 |
| jroll | ah, right | 18:27 |
| rameshg87 | jroll, in short 4 methods need to be moved to management interface | 18:27 |
| rameshg87 | jroll, does it still sound right for management interface to hold those ? | 18:27 |
| JoshNang | rameshg87: why separate create and apply? | 18:27 |
| jroll | rameshg87: it seems fine to me, idk what others think. I'm also fine with a separate interface if there's a good reason why | 18:28 |
| rameshg87 | JoshNang, create_configuration() is supposed to be implemented by a driver (ex, drac) | 18:28 |
| rameshg87 | JoshNang, this is not a zap task | 18:28 |
| rameshg87 | JoshNang, apply_configuration() is a zap method | 18:29 |
| JoshNang | right, i'm saying that apply/create could be one zap method | 18:29 |
| *** Marga_ has quit IRC | 18:30 | |
| *** Marga_ has joined #openstack-ironic | 18:31 | |
| rameshg87 | JoshNang, apply needs to call create 2 times - first with the root_volume and second with all non_root_volumes | 18:31 |
| JoshNang | sure, but zap takes a list of steps | 18:31 |
| *** achanda has joined #openstack-ironic | 18:31 | |
| rameshg87 | JoshNang, so how can same zap method be called with 2 different arguments ? | 18:33 |
| rameshg87 | may be still i am not getting it completely | 18:34 |
| rameshg87 | JoshNang, https://github.com/openstack/ironic-specs/blob/master/specs/kilo/implement-zapping-states.rst says zap steps are pulled in automatically from interfaces | 18:35 |
| JoshNang | right. when you call zapping, you pass in a list of steps. i don't see any reason you couldn't call the same step with different (or the same if you really want) args | 18:36 |
| rameshg87 | JoshNang, but since we had an api for passing raid config data (PUT /nodes/<uuid>/raid/configuration) and we dumped the information to driver_internal_info | 18:39 |
| rameshg87 | JoshNang, we would need the zap method to pick up data from there instead | 18:40 |
| JoshNang | if you combined create and apply you wouldn't have to store it at all | 18:41 |
| *** pelix has quit IRC | 18:42 | |
| rameshg87 | JoshNang, i agree. | 18:42 |
| rameshg87 | JoshNang, i thought you meant to get the data while specifying zap step, right ? | 18:42 |
| rameshg87 | JoshNang, yeah but still it can be combined | 18:44 |
| JoshNang | ok | 18:44 |
| rameshg87 | JoshNang, i would check with ifarkas if it seems okay to him | 18:44 |
| *** anderbubble has quit IRC | 18:44 | |
| JoshNang | right, i'm reviewing right now | 18:44 |
| rameshg87 | JoshNang, so apply, delete, get_logical_disk_properties would be there .. | 18:45 |
| rameshg87 | JoshNang, okay .. | 18:45 |
| JoshNang | would be on the raid management interface? | 18:48 |
| rameshg87 | JoshNang, a new RAIDInterface or on the management interface | 18:48 |
| JoshNang | gotcha | 18:49 |
| rameshg87 | JoshNang, getting late night around here..please leave your comments on the spec .. | 18:51 |
| rameshg87 | JoshNang, will check it tomorrow | 18:51 |
| JoshNang | rameshg87: ok! | 18:51 |
| rameshg87 | JoshNang, bye .. | 18:51 |
| JoshNang | have a good night! :) | 18:51 |
| rameshg87 | have a good day :) | 18:51 |
| *** rameshg87 has quit IRC | 18:52 | |
| *** wanyen has joined #openstack-ironic | 18:55 | |
| *** Marga_ has quit IRC | 18:58 | |
| *** Marga_ has joined #openstack-ironic | 18:58 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Hardware Inspection https://review.openstack.org/149823 | 18:59 |
| openstackgerrit | Merged openstack/ironic: Rename provision_state to power_state in test_manager.py https://review.openstack.org/152833 | 19:01 |
| *** stendulker_ has quit IRC | 19:09 | |
| *** kan_ has quit IRC | 19:25 | |
| *** kan_ has joined #openstack-ironic | 19:26 | |
| *** achanda has quit IRC | 19:35 | |
| *** Marga_ has quit IRC | 19:38 | |
| *** Marga_ has joined #openstack-ironic | 19:39 | |
| *** ironic has joined #openstack-ironic | 19:40 | |
| *** Nisha has quit IRC | 19:43 | |
| *** davideagnello has quit IRC | 19:44 | |
| *** achanda has joined #openstack-ironic | 19:45 | |
| *** harlowja is now known as harlowja_away | 19:50 | |
| *** wanyen has quit IRC | 19:50 | |
| ironic | Hello, Have anyone tried to deploy ironic using an external dhcp server? | 19:51 |
| *** davideagnello has joined #openstack-ironic | 19:52 | |
| *** afaranha has joined #openstack-ironic | 19:52 | |
| *** EmilienM is now known as EmilienM|afk | 19:53 | |
| afaranha | I deployed a server using Ironic and the OS Image (Fedora and Ubuntu) was installed, but the OpenStack didn't passed the user credentials to the server (Cloud init and keypair). | 19:53 |
| afaranha | When I read the logs, the server was trying to get these information from the DHCP server, instead of the Ironic server. Do anyone have any idea why this is happening? | 19:53 |
| *** ironic has left #openstack-ironic | 19:53 | |
| devananda | afaranha: that is normal behavior for cloud-init | 19:54 |
| devananda | afaranha: it's looking for a metadata service to pull the data from. are you running nova as well? | 19:55 |
| *** lintan__ has quit IRC | 19:55 | |
| afaranha | devananda: Yes, I'm running devstack with all services needed | 19:56 |
| *** davideagnello has quit IRC | 19:57 | |
| *** dkingshott has quit IRC | 19:58 | |
| devananda | afaranha: then I suspect something is misconfigured with regard to networking | 19:58 |
| devananda | afaranha: are you usnig the dhcp service that devstack set up, or something else? | 19:59 |
| afaranha | devananda, Maybe I need to do some configuration on neutron, or do you mean the lab network configuration? | 19:59 |
| *** anderbubble has joined #openstack-ironic | 19:59 | |
| afaranha | devananda, something else, I'm using an already setup dhcp, I just set the "next-server" variable to refers to the ironic server | 19:59 |
| devananda | afaranha: "next-server" should point to ironic-conductor, so that tftpboot works. | 20:01 |
| devananda | afaranha: cloud-init uses the magic IP 169.254.169.254 -- that should, IIRC, route to the nova-metadata service | 20:01 |
| afaranha | devananda, yes, it tries to get the metadata from this IP but then it starts looking in the DHCP server | 20:02 |
| devananda | hmm | 20:02 |
| devananda | off hand, I'm not sure then | 20:03 |
| devananda | lucas might know more (he has done a bunch with external DHCP) | 20:03 |
| devananda | two other options: | 20:04 |
| afaranha | Do you use a DHCP in the same server as Ironic? | 20:04 |
| devananda | - bake the credentials into the image (not good for multitenant, but sufficient for single tenant or just testing) | 20:04 |
| devananda | - use configdrive and "agent" family of drivers (though I dont think all of these patches have landed yet, so this may not be a good option just yet) | 20:06 |
| *** dkingshott has joined #openstack-ironic | 20:06 | |
| afaranha | devananda, I tried to run the dhcp in the same server as ironic, but now I got in 3 different machines error in the tftp, I'm still workinf on that though | 20:07 |
| devananda | afaranha: you do not need to put dhcp server on the same host as ironic-conductor. that said, I use neutron's dhcp, which is configured by nova and ironic during deploy ... | 20:07 |
| afaranha | So, for test purposes I don't need to set up a DHCP, Can I just run devstack and the neutron dhcp will works without any other configuration? | 20:08 |
| devananda | that is what i would recommend | 20:09 |
| devananda | you may need to modify the network config slightly from devstack's defaults so as to provision external physical machines on your networ | 20:10 |
| devananda | but that should be it | 20:10 |
| devananda | should not need any other services | 20:10 |
| afaranha | devananda, sure, I'm gonna test that, thanks | 20:11 |
| openstackgerrit | Sirushti Murugesan proposed openstack/ironic: Adds support for deploying whole disk images https://review.openstack.org/150142 | 20:15 |
| NobodyCam | brb | 20:17 |
| *** achanda has quit IRC | 20:20 | |
| *** achanda has joined #openstack-ironic | 20:21 | |
| jlvillal | So I'm looking at this patch https://review.openstack.org/#/c/149823/7/ironic/tests/api/v1/test_nodes.py (and nothing against this patch). But I was wondering if it was considered bad form to have a for loop in the testing? | 20:23 |
| jlvillal | for key in (bunch, of, keys, to, look, for)... assertIn(key, data) | 20:24 |
| rloo | jroll: I know you +2, just want to check again, you good with approving https://review.openstack.org/#/c/152055/ | 20:27 |
| jroll | rloo: I don't +2 things I don't want to approve :) | 20:28 |
| rloo | jroll: it was the 'I would love it to be written in full...' that made me wonder ;) | 20:29 |
| jroll | rloo: "but a follow up patch is fine" | 20:29 |
| jroll | :) | 20:29 |
| rloo | jroll: ah, missed that part. good. | 20:29 |
| *** anderbubble has quit IRC | 20:31 | |
| *** achanda has quit IRC | 20:32 | |
| rloo | jroll: agent_ssh is voting now, right? gate-tempest-dsvm-ironic-agent_ssh | 20:32 |
| rloo | jroll: but just in gate? is the check different? check-tempest-dsvm-ironic-agent_ssh-nv | 20:33 |
| jroll | rloo: correct | 20:33 |
| jroll | also in check, but that check went before it was voting :) | 20:33 |
| *** harlowja_away is now known as harlowja | 20:33 | |
| rloo | so the check will change to voting too? | 20:33 |
| jroll | yes | 20:34 |
| jroll | I verified already | 20:34 |
| rloo | thx for clarifying. that confused me. | 20:34 |
| *** davideagnello has joined #openstack-ironic | 20:34 | |
| devananda | jroll: periodic task parallel=False blocks *all* tasks, including parallel=True ones? | 20:34 |
| jroll | devananda: blocks them from starting the next iteration | 20:35 |
| *** penick has joined #openstack-ironic | 20:35 | |
| devananda | oh | 20:35 |
| devananda | that's very poor | 20:35 |
| devananda | why would that even be an option? | 20:38 |
| mrda-away | Morning Ironic | 20:39 |
| *** mrda-away is now known as mrda | 20:39 | |
| *** achanda has joined #openstack-ironic | 20:40 | |
| devananda | never mind .. it's late and I should be eating dinner | 20:40 |
| NobodyCam | oh: https://www.surveymonkey.com/r/openstack-l-naming | 20:41 |
| NobodyCam | devananda: before you go off to eat got a second to look at a stable patch: https://review.openstack.org/#/c/150096 | 20:41 |
| rloo | NobodyCam: me sad; I thought they'd have 'lions' for Vancouver's Lions gate bridge. | 20:43 |
| NobodyCam | oh :( I wounder if it was on the larger list | 20:44 |
| trown | lizard and love, but no lizardlove | 20:44 |
| NobodyCam | lol | 20:44 |
| jroll | devananda: I don't remember why that's an option | 20:44 |
| rloo | morning mrda | 20:45 |
| mrda | hey rloo - thnx for the reviews | 20:47 |
| rloo | mrda: yw, it is very close... | 20:47 |
| mrda | hopefully :) | 20:47 |
| *** ifarkas has quit IRC | 20:47 | |
| rloo | devananda: if you're still around, wrt MANAGEABLE https://review.openstack.org/#/c/150073/, should I +2 and you can clarify it later? | 20:48 |
| *** afaranha has quit IRC | 20:52 | |
| rloo | devananda or anyone else, wrt the microversioning, it seems like we should get https://review.openstack.org/#/c/152527/ done before approving anything changing the API? | 21:01 |
| rloo | jroll: do you know? ^^ | 21:01 |
| jroll | rloo: theoretically, but it's fine if we don't | 21:02 |
| harmw | any way of making a new node drop into debug mode or something when it hits a HTTP/400 right before rebooting? | 21:02 |
| devananda | I can address both of lucas' coments in a follow on | 21:02 |
| harmw | since pressing 't' is not working | 21:02 |
| devananda | just haven't gotten to it yet. because time. | 21:02 |
| rloo | devananda: ok thx. | 21:03 |
| devananda | hmm, and for manageable patch, it looks like nisha pushed another rev. I wonder if that was intentional? | 21:04 |
| jroll | no, it wasn't >.> | 21:05 |
| rloo | devananda: my theory is that his patch depends on the manageable patch, and when he updates his, the manageable one gets pushed. seems like the 2nd time it happened but i don't think there were any changes. | 21:05 |
| jroll | yes, because nisha doesn't pull the latest version | 21:06 |
| devananda | that means nisha is not rebasing on latest | 21:06 |
| devananda | which means it is overwriting other people's changes | 21:06 |
| jroll | right | 21:06 |
| jroll | that was discovered when it happened | 21:06 |
| jroll | and apparently nisha didn't fix it. | 21:06 |
| mrda | there were 43 patch revisisions to rebase on yesterday, so easy to miss | 21:07 |
| devananda | https://review.openstack.org/#/c/150073/12..13/ironic/api/controllers/v1/node.py,unified | 21:09 |
| devananda | https://review.openstack.org/#/c/150073/12..15/ironic/api/controllers/v1/node.py,unified | 21:09 |
| devananda | yes, nisha clobbered Shrews' fix :( | 21:09 |
| Shrews | devananda: speaking of which, sorry for the back-and-forth on that. didn't realize it was needed until it wasn't there. | 21:10 |
| devananda | Shrews: np | 21:11 |
| devananda | Shrews: want to "fix" it again? | 21:11 |
| Shrews | devananda: is no a valid answer? | 21:12 |
| devananda | rloo: your comment about the verbs changing is not what I think I meant to do | 21:12 |
| devananda | Shrews: totally | 21:12 |
| rloo | I must be missing something; Shrew's changes look like they are still there? | 21:12 |
| rloo | devananda: well, let me know what you meant ;) I think you should just delete those two entries and leave the ones related to manageable. | 21:12 |
| devananda | rloo: oh wait. is it? | 21:13 |
| devananda | oooh | 21:13 |
| Shrews | my changes are there | 21:13 |
| rloo | they aren't related to manageable anyway. | 21:13 |
| devananda | r13 removed some lines, r14 added them back | 21:13 |
| Shrews | devananda: yup. the back-and-forth | 21:13 |
| devananda | I diffed 13..15 and it looked like nisha dropped something. bah | 21:13 |
| devananda | it's fine | 21:13 |
| devananda | :) | 21:13 |
| digimonde | guys, a dummy question; reviewers told me that I should run a unit test for a change I committed. I did so (tox -- test_xxxx.py) and it succeeded. How can I update gerrit so that they now I did it? | 21:13 |
| rloo | digimonde: gerrit doesn't need to know that you ran the test. jenkins will run the test. | 21:14 |
| rloo | digimonde: as long as the test is in the patch ;) | 21:14 |
| devananda | digimonde: you should run unit tests before you "git review". | 21:15 |
| Shrews | rloo: i can take care of those nits in the comment on 150073 in a separate patch for you | 21:15 |
| digimonde | rloo: and nothing else? just git review? | 21:15 |
| devananda | digimonde: but if you want to see the rsults of the test run, you can click on the name of the job on the gerrit status page, eg, "check-ironic-python27" | 21:15 |
| *** ChuckC_ is now known as ChuckC | 21:16 | |
| devananda | digimonde: there will be a page like this one: http://logs.openstack.org/73/150073/15/check/gate-ironic-python27/5b85ab8/testr_results.html.gz | 21:16 |
| rloo | Shrews: thx. I think the only thing i object to is the changes wrt delete/deploy in the VERBS cuz they didn't make sense to me. | 21:16 |
| digimonde | ok. thanks rloo and devenanda! | 21:16 |
| devananda | rloo: there shouldn't be any change required in the verbs | 21:16 |
| rloo | digimonde: yw | 21:16 |
| devananda | rloo: eg, a client can still PUT /v1/nodes/NNN/states/provision {'target': 'active'} | 21:17 |
| * devananda goes back to bed | 21:17 | |
| rloo | devananda: right, but the description associated with the verbs was to also allow 'deploy' but the code doesn't allow that. | 21:17 |
| NobodyCam | oh food ready... brb | 21:18 |
| rloo | devananda, Shrews (I guess), PUT /v1/nodes/NNN/states/provision {'target': 'deploy'} doesn't work | 21:19 |
| *** anderbubble has joined #openstack-ironic | 21:19 | |
| rloo | devananda, Shrews: nor does PUT /v1/nodes/NNN/states/provision {'target': 'deleted'} | 21:19 |
| rloo | devananda, Shrews, yet the comment in states.py sez '...and it now also accepts the "delete" verb...' | 21:20 |
| rloo | devananda, Shrews: sigh. I meant 'target': 'delete' above. | 21:21 |
| rloo | Shrews (don't want to wake up devananda), what am I missing? | 21:22 |
| openstackgerrit | John Villalovos proposed openstack/ironic: Documentation: Describe the 'spacings' argument https://review.openstack.org/153056 | 21:25 |
| openstackgerrit | John Villalovos proposed openstack/ironic: Documentation: Describe the 'spacing' argument https://review.openstack.org/153056 | 21:27 |
| Shrews | rloo: sorry, i've context switched out of that. what do you mean by "doesn't work"? | 21:27 |
| Shrews | deploys should still work (otherwise tempest tests wouldn't pass) | 21:28 |
| rloo | the comment in the states.py file sez that the 'delete' verb should work. | 21:28 |
| rloo | Shrews: but if you do PUT /v1/nodes/NNN/states/provision {'target': 'delete'}, it won't work | 21:28 |
| rloo | Shrews: unless 'work' means getting an error | 21:28 |
| Shrews | rloo: more specifically, what is the error message you get? | 21:28 |
| rloo | Shrews: or maybe I don't understand the comment in states.py file. | 21:29 |
| rloo | Shrews: I don't recall now, but I added a comment in the code. sec. | 21:29 |
| rloo | Shrews: in https://review.openstack.org/#/c/150073/14/ironic/api/controllers/v1/node.py | 21:30 |
| rloo | Shrews: InvalidStateRequested exception is raised | 21:30 |
| Shrews | rloo: oh, maybe this line in that file needs changed: elif target == ir_states.DELETED | 21:31 |
| Shrews | or, wait... | 21:31 |
| rloo | Shrews: yeah. We can fix it. I just couldn't figure out what the intent was with the code changes there. | 21:31 |
| Shrews | that's not it | 21:31 |
| rloo | Shrews: wrt 'active' and 'deleted' provision targets. what do you think this patch intended/does? | 21:33 |
| *** dprince has quit IRC | 21:33 | |
| Shrews | rloo: i'd check that task.process_event(action) is handling that correctly in manager.py | 21:33 |
| *** achanda has quit IRC | 21:33 | |
| jroll | rloo: the 'delete verb' is likely DELETE v1/nodes/uuid, no? | 21:33 |
| Shrews | rloo: sorry, caught up in several other things, atm :( | 21:33 |
| *** Marga_ has quit IRC | 21:33 | |
| jroll | rloo: to be fair idk what I'm looking at | 21:34 |
| *** Marga_ has joined #openstack-ironic | 21:34 | |
| rloo | jroll: https://review.openstack.org/#/c/150073/14/ironic/api/controllers/v1/node.py | 21:34 |
| jroll | rloo: for tear down, I think you want PUT /v1/nodes/NNN/states/provision {'target': 'deleted'} | 21:34 |
| jroll | with a d | 21:34 |
| rloo | jroll: right, that's how it works in v1.0 or v1 or whatever | 21:35 |
| jroll | mmm | 21:35 |
| rloo | jroll: with this 150073 patch though... i don't know if I don't understand the comments or if the code doesn't reflect what the comments say it should be doing | 21:35 |
| jroll | ok, I think you're right, carry on | 21:35 |
| jroll | (at a glance) | 21:35 |
| rloo | jroll: in the new VERBS https://review.openstack.org/#/c/150073/15/ironic/common/states.py | 21:36 |
| rloo | jroll: it sez 'it also provides mapping for ...' | 21:36 |
| jroll | rloo: right, I think you're right about api versions | 21:36 |
| rloo | jroll: I interpret that as the code will support 'delete' as well as 'deleted', and 'deploy' as well as 'active' | 21:37 |
| rloo | jroll: as verbs to the provisioning API | 21:37 |
| rloo | jroll: but it ('delete' and 'deploy') don't work | 21:37 |
| jroll | rloo: right, but the older API should not support delete/deploy IMO | 21:37 |
| jroll | oh? | 21:37 |
| jroll | yeah, you're right | 21:38 |
| *** achanda has joined #openstack-ironic | 21:38 | |
| rloo | jroll: so you interpret that as this patch will support 'delete' and 'deploy' in v1.3 (or whatever version this is in) but not before that. | 21:39 |
| jroll | rloo: yes | 21:40 |
| rloo | jroll: anyway, that's why i think it is easiest/fastest to delete the active/deleted entries in VERBS, and do that in a separate patch. they have nothing to do with manageable. | 21:40 |
| jroll | well. | 21:40 |
| jroll | rloo: this is why we have versioning, it sohuldn't be supported before 1.3 | 21:40 |
| jroll | ++ that makes sense to me | 21:40 |
| rloo | jroll: yeah, I don't think they should be supported before 1.3 either. Just wasn't clear either way. | 21:41 |
| rloo | and the first time I reviewed this, I ignored those verbs cuz I didn't quite understand but figured whatever they were there for, worked :-( | 21:41 |
| *** jlvillal has quit IRC | 21:42 | |
| jroll | rloo: yeah, I see, I agree they shouldn't be there | 21:42 |
| rloo | Shrews: I am ready to +2 150073 if all mention of those two VERBS 'active' & 'deleted' are removed from the patch. | 21:43 |
| rloo | thx for listening jroll ;) | 21:45 |
| jroll | rloo: anytime :P | 21:45 |
| *** EmilienM|afk is now known as EmilienM | 21:45 | |
| Shrews | rloo: it's deva's patch | 21:47 |
| rloo | Shrews: you're co-author. I'm fine waiting for deva too. Just trying to help out here. | 21:48 |
| openstackgerrit | Erhan Ekici proposed openstack/ironic: BugFix - write() expects string but gets something else https://review.openstack.org/152702 | 21:49 |
| Shrews | rloo: ah, i see. i'm more comfortable with him making that change though. | 21:49 |
| rloo | Shrews: that's fine. Or he can fix it but after discussing with jroll, I think it should be a separate patch. too much stuff going on in there. | 21:50 |
| *** jlvillal has joined #openstack-ironic | 21:51 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: Add logical name support to Ironic https://review.openstack.org/141737 | 21:56 |
| mrda | This time for sure :P | 21:57 |
| *** bradjones has quit IRC | 21:57 | |
| *** bradjones has joined #openstack-ironic | 22:00 | |
| *** ijw_ has joined #openstack-ironic | 22:01 | |
| *** ijw has quit IRC | 22:04 | |
| *** achanda has quit IRC | 22:08 | |
| *** andreykurilin_ has joined #openstack-ironic | 22:08 | |
| rloo | mrda: do you know what the story is with microversioning and your logical names patch? | 22:12 |
| mrda | No, deva did mention they were landing about the same time | 22:14 |
| mrda | but that was it | 22:14 |
| rloo | mrda: I'm confused about that. I +1 cuz I wasn't sure. | 22:14 |
| mrda | Maybe we have to wait until morning when deva's back to answer that one? | 22:14 |
| *** kevinbenton_ is now known as kevinbenton | 22:15 | |
| mrda | devananda: if you're back tonight, I'd like to discuss what needs to happen with logical names (141737) for microversioning. If you could let me know what I need to do, I think logical names *might* be ready :) Thanks in advance... | 22:16 |
| rloo | mrda: thinking... jroll and I had a discussion about another patch earlier today, and decided they needed to bump the version. so I think that applies here too. | 22:16 |
| mrda | do you have a ref? | 22:16 |
| rloo | mrda: sec | 22:17 |
| mrda | (to the patch) | 22:17 |
| jroll | I mean | 22:18 |
| jroll | the logical name patch changes the api response | 22:18 |
| jroll | and request, even | 22:18 |
| jroll | the version should be bumped | 22:18 |
| openstackgerrit | Kyle Stevenson proposed openstack/ironic-python-agent: Add additional logging and return results for IPA extensions https://review.openstack.org/152755 | 22:18 |
| mrda | (backwards compat on the request jroll :P | 22:18 |
| jroll | and names should be allowed post-bump only (imho) | 22:18 |
| rloo | mrda: here's the earlier patch: https://review.openstack.org/#/c/150688/ | 22:18 |
| kylestev | jroll: clif_h: JayF: ^^^^ if you'd like to take another look | 22:18 |
| rloo | mrda: there's also a patch in flight that might affect things but I'm not sure (https://review.openstack.org/#/c/152527/) | 22:19 |
| rloo | mrda: but I think the versioning needs to be bumped cuz I see deva commented on that in your patch. | 22:20 |
| mrda | so this just means bumping MAX_VER ? | 22:21 |
| rloo | mrda: yeah, I believe so. | 22:21 |
| mrda | (and commenting to that fact) | 22:21 |
| *** jjohnson2 has quit IRC | 22:21 | |
| rloo | mrda: oh yeah. i wonder if the commit msg needs a DocImpact and/or an API Impact (or whatever that other tag was) | 22:22 |
| mrda | easy peesy | 22:22 |
| mrda | Please hold, your call is important to us... | 22:22 |
| rloo | mrda: just no cruddy music please | 22:22 |
| mrda | oh, the HEAD version of __init__.py doesn't have any comments. Just the raw MIN/MAX_VER idents | 22:24 |
| jlvillal | So on this patch: https://review.openstack.org/#/c/153056/ It seems to me that the failure is a false positive. But how do I prove that? :) | 22:24 |
| mrda | unlike https://review.openstack.org/#/c/152527/2/ironic/api/controllers/v1/__init__.py | 22:24 |
| rloo | mrda: yeah, that's why I wasn't sure if we should not approve anything until 152527 lands. but no one mentioned to me that we shouldn't, and we know that your change needs to bump the version... | 22:26 |
| mrda | I'll start playing elevator music if you're not careful :) | 22:26 |
| rloo | jlvillal: I think that's one of the tests we can sort of ignore. but there is a way to have the tests run again, via recheck. | 22:29 |
| rloo | jlvillal: this documentation might help: https://wiki.openstack.org/wiki/GerritJenkinsGit#Test_Failures | 22:29 |
| rloo | mrda: I'll hang up | 22:29 |
| jlvillal | rloo: Thanks! | 22:29 |
| openstackgerrit | Erhan Ekici proposed openstack/ironic: BugFix - write() expects string but gets something else https://review.openstack.org/152702 | 22:29 |
| mrda | lol | 22:29 |
| mrda | ok, so I've put it out there, let's see what happens | 22:30 |
| openstackgerrit | Michael Davies proposed openstack/ironic: Add logical name support to Ironic https://review.openstack.org/141737 | 22:30 |
| rloo | jlvillal: I reviewed your patch; that should trigger the tests to run again. | 22:31 |
| jlvillal | rloo: cool. And thanks for the review. | 22:32 |
| rloo | jlvillal: yw | 22:32 |
| jlvillal | rloo: I keep saying on a lot of patches for that test... | 22:32 |
| jlvillal | rloo: I keep seeing failures on a lot of patches for that test... | 22:33 |
| rloo | jlvillal: yeah, I think that one isn't that 'reliable' | 22:33 |
| jlvillal | I agree :) | 22:33 |
| rloo | jlvillal: I think it is a nonvoting test. Not sure though although I should know. | 22:34 |
| jlvillal | It didn't vote, from what I saw. | 22:34 |
| *** mlsoule has quit IRC | 22:35 | |
| *** imtiaz has joined #openstack-ironic | 22:38 | |
| openstackgerrit | wei du proposed openstack/ironic: Fix argument value for work_on_disk() in unit test https://review.openstack.org/153072 | 22:42 |
| *** davideagnello has quit IRC | 22:44 | |
| *** andreykurilin_ has quit IRC | 22:46 | |
| *** digimonde has quit IRC | 22:46 | |
| *** mjturek has quit IRC | 22:48 | |
| *** anderbubble has quit IRC | 22:51 | |
| *** andreykurilin_ has joined #openstack-ironic | 22:55 | |
| openstackgerrit | Merged openstack/ironic: Update agent driver with new field driver_internal_info https://review.openstack.org/152055 | 22:58 |
| *** davideagnello has joined #openstack-ironic | 23:04 | |
| *** davideagnello has quit IRC | 23:04 | |
| *** davideagnello has joined #openstack-ironic | 23:04 | |
| *** achanda has joined #openstack-ironic | 23:08 | |
| *** imtiaz_ has joined #openstack-ironic | 23:11 | |
| *** penick has quit IRC | 23:13 | |
| *** achanda has quit IRC | 23:15 | |
| *** imtiaz has quit IRC | 23:15 | |
| *** penick has joined #openstack-ironic | 23:17 | |
| NobodyCam | tripleO check broken? | 23:28 |
| *** absubram has quit IRC | 23:28 | |
| NobodyCam | error: Domain not found: no domain with matching name 'baremetalbrbm2_14' | 23:28 |
| *** korekhov has joined #openstack-ironic | 23:34 | |
| *** achanda has joined #openstack-ironic | 23:37 | |
| *** jjohnson2 has joined #openstack-ironic | 23:43 | |
| *** stendulker has joined #openstack-ironic | 23:47 | |
| jlvillal | NobodyCam: I think it is broken. I have seen failures on almost all patches. | 23:48 |
| NobodyCam | ack :) | 23:48 |
| *** jjohnson2 has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!