| *** openstack has joined #openstack-ironic | 00:04 | |
| *** naohirot has joined #openstack-ironic | 00:08 | |
| *** boris-42 has quit IRC | 00:12 | |
| *** mtanino has quit IRC | 00:15 | |
| *** Sukhdev has quit IRC | 00:16 | |
| *** achanda has quit IRC | 00:17 | |
| *** romcheg has quit IRC | 00:20 | |
| *** ijw has quit IRC | 00:24 | |
| *** davideagnello has joined #openstack-ironic | 00:28 | |
| *** max_lobur has quit IRC | 00:34 | |
| *** achanda has joined #openstack-ironic | 00:36 | |
| *** barra204 has quit IRC | 00:39 | |
| *** athomas has joined #openstack-ironic | 00:42 | |
| *** cing has joined #openstack-ironic | 00:45 | |
| *** achanda has quit IRC | 00:56 | |
| *** Pradip has quit IRC | 00:59 | |
| *** achanda has joined #openstack-ironic | 01:07 | |
| *** hemna has joined #openstack-ironic | 01:14 | |
| devananda | jroll: re: provider network, I think we're talking about different things. I want BM node to be on *tenant* network (some random VLAN that only tenant X has access to), whereas you (RS OnMental) need it to be connect to a provider network (VLAN 123 which routes between OnMetal region and other regions of RS cloud) | 01:18 |
|---|---|---|
| jroll | devananda: I want both. | 01:18 |
| jroll | devananda: *today* we only have nodes on the provider network | 01:19 |
| jroll | devananda: but clearly we want to enable isolated tenant networks | 01:19 |
| devananda | right | 01:20 |
| jroll | devananda: and others likely want both, imagine cinder support, that likely won't be on a tenant network | 01:20 |
| devananda | indeed | 01:20 |
| jroll | ok cool, same page | 01:21 |
| *** hemna has quit IRC | 01:21 | |
| devananda | yah. in skimming scrollback, it looked like you were only talking about provider network | 01:21 |
| jroll | well, provider networks are where vlan info is defined today | 01:21 |
| jroll | tenant networks have no concept of vlans | 01:22 |
| jroll | AIUI | 01:22 |
| jroll | provider nets are also where we have the admin-only problem | 01:22 |
| jroll | so I was focusing on that :) | 01:22 |
| devananda | heh | 01:23 |
| devananda | so what, we need to add vlan id to tenant network? | 01:23 |
| *** ijw has joined #openstack-ironic | 01:23 | |
| jroll | something like that | 01:23 |
| jroll | or the port; I'm indifferent, honestly | 01:24 |
| devananda | *nod* | 01:24 |
| * jroll selfishly wants to let morgabra figure that out | 01:24 | |
| *** hemna has joined #openstack-ironic | 01:27 | |
| *** puranamr has joined #openstack-ironic | 01:41 | |
| *** ijw_ has joined #openstack-ironic | 01:45 | |
| *** bizarrochristy has joined #openstack-ironic | 01:46 | |
| *** ijw has quit IRC | 01:49 | |
| *** puranamr has joined #openstack-ironic | 01:50 | |
| *** bizarrochristy has quit IRC | 01:50 | |
| *** puranamr has quit IRC | 01:52 | |
| *** puranamr has joined #openstack-ironic | 01:58 | |
| *** puranamr has quit IRC | 01:59 | |
| *** ijw has joined #openstack-ironic | 02:04 | |
| *** ijw_ has quit IRC | 02:08 | |
| *** puranamr has joined #openstack-ironic | 02:14 | |
| *** achanda has quit IRC | 02:23 | |
| *** zhenguo has joined #openstack-ironic | 02:32 | |
| *** chlong-afk has quit IRC | 02:37 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add db migration for tags and node_tags table https://review.openstack.org/192576 | 02:40 |
| *** chlong has joined #openstack-ironic | 02:42 | |
| *** sinval has quit IRC | 02:43 | |
| *** puranamr has quit IRC | 02:47 | |
| *** ijw has quit IRC | 02:52 | |
| *** hakimo has joined #openstack-ironic | 02:53 | |
| *** hakimo_ has quit IRC | 02:55 | |
| *** Nisha has joined #openstack-ironic | 03:02 | |
| *** puranamr has joined #openstack-ironic | 03:07 | |
| *** naohirot has quit IRC | 03:07 | |
| *** hakimo_ has joined #openstack-ironic | 03:10 | |
| *** hakimo has quit IRC | 03:10 | |
| *** puranamr has quit IRC | 03:12 | |
| *** boris-42 has joined #openstack-ironic | 03:14 | |
| *** achanda has joined #openstack-ironic | 03:23 | |
| *** hakimo has joined #openstack-ironic | 03:27 | |
| *** hakimo_ has quit IRC | 03:27 | |
| *** achanda has quit IRC | 03:28 | |
| *** puranamr has joined #openstack-ironic | 03:29 | |
| *** puranamr has quit IRC | 03:33 | |
| *** hakimo_ has joined #openstack-ironic | 03:44 | |
| *** hakimo has quit IRC | 03:44 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add db migration for tags and node_tags table https://review.openstack.org/192576 | 03:44 |
| *** hblixt has joined #openstack-ironic | 03:45 | |
| *** pal has joined #openstack-ironic | 03:46 | |
| *** puranamr has joined #openstack-ironic | 03:48 | |
| *** puranamr has quit IRC | 03:51 | |
| *** Nisha has quit IRC | 03:51 | |
| *** puranamr has joined #openstack-ironic | 03:53 | |
| *** hblixt has quit IRC | 03:54 | |
| *** coolsvap|away is now known as coolsvap | 03:54 | |
| *** achanda has joined #openstack-ironic | 03:55 | |
| *** Nisha has joined #openstack-ironic | 03:56 | |
| *** mitchjameson has quit IRC | 03:56 | |
| *** zz_natorious is now known as natorious | 03:58 | |
| *** hemna has quit IRC | 03:58 | |
| *** achanda has quit IRC | 04:00 | |
| *** puranamr has quit IRC | 04:00 | |
| *** hakimo has joined #openstack-ironic | 04:01 | |
| *** hakimo_ has quit IRC | 04:01 | |
| *** puranamr has joined #openstack-ironic | 04:01 | |
| *** pal has quit IRC | 04:04 | |
| *** naohirot has joined #openstack-ironic | 04:06 | |
| *** hemna has joined #openstack-ironic | 04:08 | |
| *** puranamr has quit IRC | 04:09 | |
| *** hemna has quit IRC | 04:11 | |
| *** hemna has joined #openstack-ironic | 04:12 | |
| *** hemna has quit IRC | 04:12 | |
| *** hakimo_ has joined #openstack-ironic | 04:18 | |
| *** hakimo has quit IRC | 04:18 | |
| openstackgerrit | Shivanand Tendulker proposed stackforge/proliantutils: Add RIS support for updating boot device https://review.openstack.org/201420 | 04:33 |
| *** bizarrochristy has joined #openstack-ironic | 04:37 | |
| *** bizarrochristy has quit IRC | 04:41 | |
| *** bizarrochristy has joined #openstack-ironic | 04:41 | |
| *** zhenguo has quit IRC | 04:43 | |
| *** bizarrochristy has quit IRC | 04:45 | |
| *** hemna has joined #openstack-ironic | 04:48 | |
| openstackgerrit | Shivanand Tendulker proposed stackforge/proliantutils: Add RIS support for updating boot device https://review.openstack.org/201420 | 04:48 |
| *** hemna has quit IRC | 04:49 | |
| *** hemna has joined #openstack-ironic | 04:52 | |
| *** pal has joined #openstack-ironic | 05:08 | |
| *** rameshg87 has joined #openstack-ironic | 05:11 | |
| *** dguerri` is now known as dguerri | 05:19 | |
| *** boris-42_ has joined #openstack-ironic | 05:20 | |
| *** lekha_ has joined #openstack-ironic | 05:21 | |
| *** boris-42 has quit IRC | 05:23 | |
| *** lekha has quit IRC | 05:23 | |
| *** cing has quit IRC | 05:23 | |
| *** egon has quit IRC | 05:23 | |
| *** egon_ has joined #openstack-ironic | 05:23 | |
| *** cing has joined #openstack-ironic | 05:23 | |
| *** lekha_ is now known as lekha | 05:23 | |
| *** boris-42_ is now known as boris-42 | 05:24 | |
| *** lazy_prince has joined #openstack-ironic | 05:24 | |
| *** stendulker has joined #openstack-ironic | 05:28 | |
| *** Sukhdev has joined #openstack-ironic | 05:43 | |
| *** ramineni has joined #openstack-ironic | 06:03 | |
| *** pal has quit IRC | 06:07 | |
| *** ramineni has quit IRC | 06:08 | |
| *** ig0r_ has joined #openstack-ironic | 06:08 | |
| *** achanda has joined #openstack-ironic | 06:21 | |
| *** karimb has joined #openstack-ironic | 06:32 | |
| *** natorious is now known as zz_natorious | 06:33 | |
| *** hemna has quit IRC | 06:36 | |
| *** Sukhdev has quit IRC | 06:37 | |
| *** achanda has quit IRC | 06:39 | |
| *** pal has joined #openstack-ironic | 06:39 | |
| *** dguerri is now known as dguerri` | 06:40 | |
| *** pal has quit IRC | 06:41 | |
| *** ndipanov has quit IRC | 06:42 | |
| *** pal has joined #openstack-ironic | 06:43 | |
| *** ndipanov has joined #openstack-ironic | 06:43 | |
| *** achanda has joined #openstack-ironic | 06:45 | |
| *** ig0r_ has quit IRC | 06:45 | |
| openstackgerrit | Satoru Moriya proposed openstack/ironic-specs: Add attributes about volume conneciton into nodes table https://review.openstack.org/200496 | 06:47 |
| *** ig0r_ has joined #openstack-ironic | 06:50 | |
| *** bradjones has quit IRC | 07:08 | |
| *** lazy_prince has quit IRC | 07:10 | |
| *** bradjones has joined #openstack-ironic | 07:11 | |
| *** bradjones has quit IRC | 07:11 | |
| *** bradjones has joined #openstack-ironic | 07:11 | |
| *** boris-42 has quit IRC | 07:12 | |
| *** dtantsur|afk is now known as dtantsur | 07:15 | |
| dtantsur | Morning folks | 07:15 |
| Haomeng | dtantsur: good morning:) | 07:21 |
| *** achanda has quit IRC | 07:27 | |
| *** athomas has quit IRC | 07:31 | |
| *** ifarkas has joined #openstack-ironic | 07:33 | |
| *** athomas has joined #openstack-ironic | 07:35 | |
| *** yog_ has joined #openstack-ironic | 07:37 | |
| *** achanda has joined #openstack-ironic | 07:42 | |
| *** Nisha has quit IRC | 07:50 | |
| *** coolsvap is now known as coolsvap|afk | 07:51 | |
| *** pal has quit IRC | 07:52 | |
| *** romainh has joined #openstack-ironic | 08:01 | |
| *** Marga_ has quit IRC | 08:04 | |
| *** lucasagomes has joined #openstack-ironic | 08:05 | |
| *** Nisha has joined #openstack-ironic | 08:06 | |
| *** jistr has joined #openstack-ironic | 08:15 | |
| *** achanda has quit IRC | 08:16 | |
| *** tiagogomes has joined #openstack-ironic | 08:27 | |
| *** dtantsur is now known as dtantsur|brb | 08:27 | |
| *** pal has joined #openstack-ironic | 08:31 | |
| *** rcarrillocruz has quit IRC | 08:35 | |
| *** derekh has joined #openstack-ironic | 08:37 | |
| *** romcheg has joined #openstack-ironic | 08:40 | |
| *** yuanying has quit IRC | 08:43 | |
| *** coolsvap|afk is now known as coolsvap | 08:54 | |
| *** romcheg has quit IRC | 08:55 | |
| *** max_lobur has joined #openstack-ironic | 08:56 | |
| *** pal has quit IRC | 08:57 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic-specs: UEFI secure boot support for pxe_ilo driver https://review.openstack.org/174295 | 08:57 |
| *** e0ne has joined #openstack-ironic | 09:01 | |
| *** romcheg has joined #openstack-ironic | 09:07 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/python-ironicclient: DO NOT MERGE - TESTING TRIPLEO CI https://review.openstack.org/202935 | 09:10 |
| *** dlpartain has joined #openstack-ironic | 09:15 | |
| *** dlpartain has left #openstack-ironic | 09:16 | |
| sambetts | Morning all o/ TGIF :D | 09:20 |
| lucasagomes | sambetts, hi there | 09:24 |
| lucasagomes | good morning | 09:24 |
| sambetts | lucasagomes o/ | 09:25 |
| *** stendulker_ has joined #openstack-ironic | 09:31 | |
| *** stendulker has quit IRC | 09:34 | |
| *** romcheg has quit IRC | 09:35 | |
| *** marzif_ has joined #openstack-ironic | 09:39 | |
| *** pal has joined #openstack-ironic | 09:39 | |
| *** pal has quit IRC | 09:44 | |
| *** romcheg has joined #openstack-ironic | 09:44 | |
| *** dtantsur|brb is now known as dtantsur | 09:46 | |
| dtantsur | sambetts, morning! | 09:46 |
| dtantsur | oh really TGIF!! | 09:46 |
| * dtantsur is tired like hell | 09:46 | |
| openstackgerrit | Merged openstack/ironic: db: use new EngineFacade feature of oslo.db https://review.openstack.org/191801 | 09:46 |
| dtantsur | lucasagomes, hi there. I don't know what was decided about the release, we should ask devananda tonight | 09:48 |
| lucasagomes | dtantsur, ack, yeah | 09:48 |
| *** romainh has quit IRC | 09:49 | |
| dtantsur | lucasagomes, and good question about target_provision_state. I don't know the answer :) | 09:49 |
| dtantsur | lucasagomes, also any reason not to approve https://review.openstack.org/#/c/202699/ ? | 09:49 |
| lucasagomes | dtantsur, oh not really, forgot to push the final button | 09:50 |
| sambetts | hey dtantsur | 09:53 |
| *** saripurigopi has joined #openstack-ironic | 09:58 | |
| saripurigopi | lucasagomes: Hi | 09:58 |
| lucasagomes | saripurigopi, hi there | 09:59 |
| *** naohirot has quit IRC | 09:59 | |
| saripurigopi | lucasagomes: when I set node capabilities with boot_option=local, does this mean, Ironic will set the boot device to HDD after PXE deployment? | 10:00 |
| lucasagomes | saripurigopi, yes, but this capabilities has to match with the flavor | 10:02 |
| lucasagomes | saripurigopi, so the nova scheduler will pick that node | 10:02 |
| *** tiagogomes has quit IRC | 10:03 | |
| rameshg87 | lucasagomes: dtantsur: when you have some time, can you please have a look at https://review.openstack.org/#/c/174295/ . this from ilo driver has been lying around for quite some time :( | 10:13 |
| dtantsur | ack | 10:13 |
| lucasagomes | will do! | 10:14 |
| rameshg87 | thanks .. | 10:15 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix and enhance contributing documentation on writing plugins https://review.openstack.org/202963 | 10:17 |
| dtantsur | sambetts, related to your change ^^ | 10:17 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix and enhance contributing documentation on writing plugins https://review.openstack.org/202963 | 10:18 |
| *** amotoki has joined #openstack-ironic | 10:18 | |
| *** tiagogomes has joined #openstack-ironic | 10:19 | |
| *** athomas has quit IRC | 10:22 | |
| dtantsur | yui<TAB> :( | 10:26 |
| dtantsur | lucasagomes, could you review a small inspector patch https://review.openstack.org/#/c/202128/ please? I want to link to this documentation from ironic documentation | 10:26 |
| sambetts | dtantsur: Commented :D | 10:26 |
| dtantsur | thnx | 10:26 |
| *** Nisha has quit IRC | 10:26 | |
| *** amotoki_ has joined #openstack-ironic | 10:28 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix and enhance contributing documentation on writing plugins https://review.openstack.org/202963 | 10:29 |
| *** pal has joined #openstack-ironic | 10:30 | |
| *** amotoki has quit IRC | 10:31 | |
| *** stendulker_ has quit IRC | 10:31 | |
| *** openstackgerrit has quit IRC | 10:31 | |
| saripurigopi | lucasagomes: okay | 10:32 |
| *** openstackgerrit has joined #openstack-ironic | 10:32 | |
| *** athomas has joined #openstack-ironic | 10:33 | |
| openstackgerrit | Merged openstack/ironic-inspector: Clarify version support matrix https://review.openstack.org/202128 | 10:33 |
| sambetts | dtantsur: I left a comment on the matrix one too but I was too slow :( | 10:34 |
| dtantsur | sambetts, may I ask you to post a follow-up please? | 10:35 |
| lucasagomes | sambetts, I can abort the merge | 10:35 |
| lucasagomes | oh it's already merged | 10:35 |
| lucasagomes | damn inspector gate is too quick | 10:36 |
| dtantsur | hehe, we don't have dsvm voting yet :D | 10:36 |
| dtantsur | (which sucks) | 10:36 |
| sambetts | dtantsur: I'll chuck out a follow up :P | 10:36 |
| dtantsur | sambetts, awesome, thanks! | 10:36 |
| * dtantsur is messing with inspector milestones | 10:36 | |
| *** stendulker has joined #openstack-ironic | 10:37 | |
| sambetts | dtantsur, lucasagomes: I never know if ironic should be "Ironic" or "ironic" in things like this | 10:39 |
| dtantsur | I'm still using Ironic for consistency | 10:39 |
| dtantsur | if we change it, we should change everything at once | 10:39 |
| lucasagomes | I tend to use Ironic (capitalized) | 10:39 |
| sambetts | ok, I'll follow suit | 10:39 |
| dtantsur | (but I don't like it changed) | 10:39 |
| *** lsmola has joined #openstack-ironic | 10:40 | |
| stendulker | dtantsur: Hi | 10:44 |
| dtantsur | stendulker, o/ | 10:45 |
| stendulker | dtantsur: Wanted to understand your comment for review https://review.openstack.org/#/c/200398/ Fixes a testcase related to trusted boot in UEFI boot mode | 10:45 |
| dtantsur | stendulker, you have a bunch of similar tests using "properties". you're changing one of them. why? | 10:46 |
| dtantsur | at least the next one looks similar | 10:46 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Improve support matrix readablity https://review.openstack.org/202974 | 10:47 |
| stendulker | dtantsur: This testcase was about testing the uefi boot mode when trusted boot is selected | 10:47 |
| stendulker | dtantsur: And we were passing the boot_mode through capabilities/properties | 10:47 |
| dtantsur | stendulker, maybe my problem is that test_validate_fail_trusted_boot says strictly nothing about what this test is testing. | 10:48 |
| stendulker | dtantsur: The deploy_utils.py:get_boot_mode_for_deploy() checks if the trusted boot is present in instance_info before checking properties/capabilities | 10:48 |
| stendulker | dtantsur: And it returns boot mode as bios if the trsuted_boot is present in instance_info | 10:49 |
| dtantsur | stendulker, so the next test is wrong too, no? | 10:50 |
| stendulker | dtantsur: That is true. Testcase do not say anything abt it. But if you look at it, it seemed to me it wanted to test trusted_boot and uefi combo | 10:50 |
| dtantsur | stendulker, anyway, could you please rename test_validate_fail_trusted_boot to something understandable? maybe also leave comment what it is testing? | 10:51 |
| dtantsur | otherwise it looks like string shuffling to me :D | 10:51 |
| stendulker | dtantsur: sure. the functiona names used are not so intuitive here | 10:52 |
| stendulker | dtantsur: I accidentaly stumbled on this one... | 10:52 |
| dtantsur | stendulker, thanks! | 10:53 |
| stendulker | dtantsur: The next testcase about the invalid value being passed to trusted_boot parameter | 10:53 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add RPCAPIs for RAID configuration https://review.openstack.org/196006 | 10:55 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add APIs for RAID configuration https://review.openstack.org/196007 | 10:55 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add RAIDInterface for RAID configuration https://review.openstack.org/196003 | 10:55 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add support for inband raid configuration agent ramdisk https://review.openstack.org/198238 | 10:55 |
| stendulker | dtantsur: thanks :) | 10:56 |
| *** dlpartain has joined #openstack-ironic | 10:59 | |
| *** cing has quit IRC | 11:00 | |
| * rameshg87 goes home | 11:01 | |
| *** rameshg87 has quit IRC | 11:02 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Allow upgrading shared mock to an exclusive one https://review.openstack.org/202558 | 11:02 |
| *** pal has quit IRC | 11:04 | |
| TheJulia | good morning | 11:05 |
| dtantsur | TheJulia, morning | 11:06 |
| *** dlpartain has quit IRC | 11:06 | |
| dtantsur | sambetts, I'll drop the docstring from an example, it just duplicates the base class (and we don't have docstrings in ExampleProcessingHook for this reason) | 11:07 |
| sambetts | dtantsur: Ok :D good idea | 11:07 |
| sambetts | Morning TheJulia o/ | 11:07 |
| dtantsur | sambetts, I wonder if we could you have a functional test on node_not_found hook... | 11:08 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix and enhance contributing documentation on writing plugins https://review.openstack.org/202963 | 11:08 |
| sambetts | dtantsur: Hmmm, so my idea for a node_not_found hook, would be when triggered it would add a new node into ironic, update the node-cache and then return the NodeInfo object, would that work in a functional testing env? \ | 11:10 |
| dtantsur | sambetts, we can mock the "add a new node into ironic" thing, everything else should work IMO | 11:11 |
| *** kcaj has joined #openstack-ironic | 11:13 | |
| sambetts | dtantsur: putting a node in ironic into maintainence mode disables all ironic management of that node right? | 11:13 |
| dtantsur | sambetts, everything automatic | 11:14 |
| *** coolsvap is now known as coolsvap|away | 11:14 | |
| dtantsur | you can still issue various requests, e.g. to power on | 11:14 |
| sambetts | dtantsur: ah ok, awesome :) | 11:15 |
| dtantsur | sambetts, is it related to the node_not_found topic? :) | 11:15 |
| sambetts | dtantsur: sort of, basicly what I was thinking is that when the not_not_found hook adds a node into ironic it should be in maintainence mode so that ironic doesn't imdiantly start trying to manage it, then its an operators active choice to take it out of maintainence mode and have it managed by ironic | 11:16 |
| dtantsur | sambetts, aha, my guess was right ;) no, you need ENROLL state instead | 11:17 |
| dtantsur | node in ENROLL state is not scheduled on | 11:17 |
| dtantsur | sambetts, read for details: http://specs.openstack.org/openstack/ironic-specs/specs/liberty/enroll-node-state.html | 11:17 |
| sambetts | ah I forgot about that | 11:18 |
| *** saripurigopi has quit IRC | 11:18 | |
| lucasagomes | TheJulia, good ugt morning | 11:19 |
| sambetts | dtantsur: that got merged this cycle already? | 11:19 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Fixes a testcase related to trusted boot in UEFI boot mode https://review.openstack.org/200398 | 11:20 |
| *** lucasagomes is now known as lucas-hungry | 11:20 | |
| dtantsur | sambetts, yep. it *could* be temporary reverted soon to allow a release without, but then it's going to be reitroduced quickly (ask devananda if he really plans to do it) | 11:20 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/202699 | 11:20 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/202700 | 11:20 |
| sambetts | dtantsur: Ah ok, awesome, that makes things much easier ;) | 11:21 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Fixes a testcase related to trusted boot in UEFI boot mode https://review.openstack.org/200398 | 11:24 |
| *** pal has joined #openstack-ironic | 11:27 | |
| *** stendulker has quit IRC | 11:32 | |
| *** pal has quit IRC | 11:33 | |
| *** alex_xu is now known as alexus | 11:34 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Clarify inspection upgrade guide https://review.openstack.org/202988 | 11:44 |
| dtantsur | this ^^ is referring to the version support matrix we just approved | 11:47 |
| *** dprince has joined #openstack-ironic | 11:49 | |
| openstackgerrit | Merged openstack/ironic-inspector: Improve support matrix readablity https://review.openstack.org/202974 | 11:55 |
| *** amotoki_ has quit IRC | 12:02 | |
| *** cing has joined #openstack-ironic | 12:03 | |
| *** thrash has quit IRC | 12:03 | |
| *** trown|outttypeww is now known as trown | 12:14 | |
| *** dlpartain has joined #openstack-ironic | 12:15 | |
| *** romainh has joined #openstack-ironic | 12:15 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Change default OS from Ubuntu to Debian https://review.openstack.org/203000 | 12:20 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Make test timeout a variable https://review.openstack.org/203001 | 12:20 |
| *** lsmola has quit IRC | 12:21 | |
| *** lucas-hungry is now known as lucasagomes | 12:22 | |
| *** lsmola has joined #openstack-ironic | 12:23 | |
| *** dlpartain has left #openstack-ironic | 12:23 | |
| dtantsur | I suspect that our gate is broken >_< | 12:26 |
| dtantsur | http://logs.openstack.org/99/202699/3/check/gate-ironic-python27/56910c9/testr_results.html.gz | 12:26 |
| dtantsur | I see it the 3d time now | 12:27 |
| TheJulia | If thats where I think it is, thats not a pretty area... | 12:31 |
| *** pal has joined #openstack-ironic | 12:37 | |
| *** coolsvap|away is now known as coolsvap | 12:37 | |
| jroll | morning all :) | 12:40 |
| TheJulia | good morning | 12:40 |
| jroll | dtantsur: new mock this morning might be suspect | 12:40 |
| jroll | hm, code looks like this should fail, though | 12:42 |
| jroll | or I should say the error looks reasonable for this code | 12:43 |
| * jroll needs more coffee to be sure | 12:43 | |
| jroll | so much for proposing a release :( | 12:44 |
| *** tiagogomes has left #openstack-ironic | 12:44 | |
| dtantsur | *sigh* | 12:44 |
| jroll | this code to "make a temp file with a password" is ridiculously complex, holy cow | 12:46 |
| jroll | or more the code to test it | 12:46 |
| *** thrash has joined #openstack-ironic | 12:47 | |
| TheJulia | So a user submitted a change to the ipmi retry logic, and I'd like to get others thoughts on the subject. They are proposing to always retry failed ipmitool commands unless "core dumped" is detected in the output. https://review.openstack.org/#/c/202751/ I'm weary of doing so, but I could see the case for identifying more errors we can retry upon and adding those into the list as it were | 12:48 |
| NobodyCam | good morning Ironicers | 13:00 |
| dtantsur | NobodyCam, morning! | 13:00 |
| dtantsur | and I can't believe it, but seems like TGIF! | 13:00 |
| NobodyCam | morning dtantsur :) I was Just about to say | 13:00 |
| jroll | dtantsur: I think I've got this, it's silly | 13:01 |
| NobodyCam | TGIF! | 13:01 |
| TheJulia | indeed! | 13:01 |
| jroll | tgif indeeeeed. | 13:01 |
| jroll | morning NobodyCam :) | 13:01 |
| NobodyCam | morning jroll and wow your up early today | 13:01 |
| NobodyCam | morning TheJulia :) | 13:01 |
| jroll | ya, crashed early last night | 13:01 |
| jroll | at the nova meetup next week so trying to wrap things up | 13:01 |
| thiagop | mornign ya'll | 13:01 |
| jroll | heya thiagop :) | 13:02 |
| NobodyCam | morning thiagop | 13:02 |
| NobodyCam | jroll: ++ | 13:02 |
| trown | good ugt morning and TGIF NobodyCam dtantsur jroll thiagop thiagop | 13:02 |
| jroll | \o trown | 13:02 |
| NobodyCam | mornign trown :) | 13:02 |
| thiagop | NobodyCam: have you received an e-mail from liliars? | 13:02 |
| dtantsur | trown, thiagop, o/ | 13:02 |
| NobodyCam | I have not checked yet. | 13:03 |
| thiagop | TGIF too, but fridays are a little stressing for me... | 13:03 |
| thiagop | but, at least, I know I can get drunk at the end of it | 13:03 |
| thiagop | :) | 13:03 |
| NobodyCam | :) | 13:04 |
| NobodyCam | oh coffee ready brb | 13:05 |
| *** dlpartain has joined #openstack-ironic | 13:05 | |
| jroll | dtantsur: lol https://review.openstack.org/#/c/202558/3//COMMIT_MSG | 13:07 |
| dtantsur | omg, devananda already fixed this one for me once :D | 13:07 |
| jroll | lolll | 13:07 |
| * dtantsur is fixing | 13:08 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Allow upgrading shared lock to an exclusive one https://review.openstack.org/202558 | 13:08 |
| *** bnemec has joined #openstack-ironic | 13:11 | |
| *** dlpartain has left #openstack-ironic | 13:11 | |
| trown | lol, mock on the mind | 13:12 |
| *** bizarrochristy has joined #openstack-ironic | 13:16 | |
| *** bizarrochristy has quit IRC | 13:16 | |
| NobodyCam | thiagop: I did get the email... will start reading | 13:20 |
| thiagop | NobodyCam: thanks | 13:21 |
| jroll | I filed bug #1475653 for the gate breakage | 13:22 |
| openstack | bug 1475653 in Ironic "Gate broken with mock 1.1.4 release" [Undecided,New] https://launchpad.net/bugs/1475653 | 13:22 |
| jroll | working on it now but it's weird | 13:22 |
| NobodyCam | awesome TY jroll :) | 13:22 |
| jroll | the break is https://github.com/testing-cabal/mock/commit/e9db0161fc11eceba189a0cc161deefce57529a8 | 13:23 |
| jroll | we were using that mock.return_value | 13:23 |
| jroll | directly, not by calling it | 13:23 |
| *** cing has quit IRC | 13:28 | |
| *** cing has joined #openstack-ironic | 13:29 | |
| *** dguerri` has quit IRC | 13:37 | |
| *** anteaya has quit IRC | 13:38 | |
| *** dguerri` has joined #openstack-ironic | 13:40 | |
| *** dguerri` is now known as dguerri | 13:41 | |
| *** dguerri has joined #openstack-ironic | 13:41 | |
| *** Shrews has quit IRC | 13:41 | |
| *** anteaya has joined #openstack-ironic | 13:42 | |
| *** hakimo has joined #openstack-ironic | 13:43 | |
| *** hakimo_ has quit IRC | 13:43 | |
| *** Shrews has joined #openstack-ironic | 13:43 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Insert artificial delay between sending virtual nodes on introspection https://review.openstack.org/203040 | 14:00 |
| dtantsur | what a nasty patch I just did... ^^ | 14:00 |
| *** rook has joined #openstack-ironic | 14:01 | |
| rook | morning | 14:01 |
| dtantsur | o/ | 14:01 |
| NobodyCam | morning rook | 14:02 |
| ifarkas | morning all | 14:02 |
| ifarkas | dtantsur, so regarding adding a driver specific check for introspection, where would you put it? | 14:02 |
| dtantsur | we're moving here our internal chat about one DRAC driver problem :) | 14:02 |
| NobodyCam | morning ifarkas :) | 14:02 |
| dtantsur | ifarkas, there should not be "driver specific check for introspection" | 14:02 |
| dtantsur | ifarkas, we either do it for all set_boot_device calls or just not :) | 14:02 |
| rook | I still like my idea, at the creation of the drac client/driver we check the queue and return it in the object. | 14:03 |
| rook | if the queue > 0 : we are pretty much hosed. | 14:03 |
| dtantsur | rook, you mean essentially at ironic start up? | 14:03 |
| dtantsur | won't if fail e.g. if we restart conductor while a node is rebooting? | 14:04 |
| ifarkas | rook, that's not necessarily true, after bios/raid configuration we already have a config job which means queue > 0 | 14:04 |
| rook | A little bit of background for NobodyCam or anyone listening: I have a R610 Dell machine that is giving me grief during introspection. it has a job that is "stuck" and cannot be cleared. | 14:04 |
| rook | which is making introspection puke. | 14:04 |
| rook | ifarkas well, for jobs that never "clear" | 14:04 |
| ifarkas | dtantsur, then I don't really understand. that was the proposal of rook which you suggested to put in to the drac driver | 14:04 |
| rook | ifarkas: which have 3 machines in my env that has this problem... so I am sure that it is going to happen elsewhere. | 14:04 |
| dtantsur | ifarkas, it's not introspection dependent, that's all :) so we should probably retry in set_boot_device (and other calls) if we get such error | 14:05 |
| dtantsur | but if a job is just stuck... I'm not sure what to do | 14:05 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix mock 1.1.4 breakage https://review.openstack.org/203042 | 14:06 |
| rook | and my comments are without knowing the code ifarkas dtantsur so, i really am depending on you guys for ideas :/ | 14:06 |
| jroll | dtantsur: ^ that should fix the gate | 14:06 |
| dtantsur | jroll, awesome! | 14:06 |
| ifarkas | dtantsur, we already have that check in set_boot_device :-) https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/drac/management.py#L382 | 14:06 |
| jroll | dtantsur: it's so silly | 14:07 |
| dtantsur | ifarkas, I don't see it retrying, I see it just failing | 14:07 |
| dtantsur | jroll, ETOOMUCHMOCKING Oo | 14:07 |
| ifarkas | dtantsur, we should retry for how long? is it ok for set_boot_device waits for an hour for example? | 14:07 |
| dtantsur | nope | 14:07 |
| jroll | dtantsur: mmmmmhm. | 14:08 |
| ifarkas | dtantsur, that's the problem, some task might take more than an hour | 14:08 |
| dtantsur | ifarkas, I'm not sure we should retry, I'm translating rook's idea (as I understood) | 14:08 |
| dtantsur | ifarkas, so, here's the thing: if ironic does something with a node, it should hold a lock on it | 14:08 |
| ifarkas | dtantsur, yeah, I am also just translating to code implication :-) | 14:09 |
| dtantsur | ifarkas, if it's NOT ironic who does something with a node, that's a bad condition | 14:09 |
| lucasagomes | jroll, https://bugs.launchpad.net/ironic/+bug/1475672 | 14:09 |
| openstack | Launchpad bug 1475672 in Ironic "agent_* drivers could timed out mid-deployment" [Medium,New] - Assigned to Lucas Alvares Gomes (lucasagomes) | 14:09 |
| lucasagomes | I'm working on a fix for it | 14:09 |
| lucasagomes | jroll, and morning | 14:09 |
| ifarkas | dtantsur, then we should probably delete most drivers :-) | 14:09 |
| *** cinerama has quit IRC | 14:09 | |
| ifarkas | dtantsur, as a part of them use bmc and async config jobs | 14:10 |
| dtantsur | ifarkas, good idea :) but async jobs still have lock (or at least are supposed to have). Only DEPLOYWAIT state does not have lock | 14:10 |
| rook | so, what happens with that code ifarkas we are not doing anything with that exception. | 14:10 |
| dtantsur | (now CLEANWAIT as well) | 14:10 |
| rook | ^ in the case there are existing jobs. | 14:11 |
| jroll | dtantsur: I WIP'd that mock fix, see https://review.openstack.org/#/c/203043/ | 14:11 |
| *** hakimo has quit IRC | 14:11 | |
| *** hakimo_ has joined #openstack-ironic | 14:11 | |
| ifarkas | rook, yep, we are just throwing that exception because setting the boot device while there's another config job present is not possible | 14:12 |
| ifarkas | dtantsur, rook, alternatively we could add something like /introspect/validate call which could check the status | 14:13 |
| ifarkas | dtantsur, rook, that would be generic enough | 14:14 |
| dtantsur | ifarkas, call to which service? and what should it do? | 14:14 |
| ifarkas | dtantsur, I am not sure if drac async job have any locks. those are not exposed via any api | 14:14 |
| ifarkas | dtantsur, discoverd -> ironic | 14:15 |
| dtantsur | ifarkas, ok man it's Friday evening and I'm probably being stupid :) could you please explain in more details? | 14:15 |
| ifarkas | dtantsur, as rook suggested we need a driver specific check before introspection starts | 14:16 |
| ifarkas | dtantsur, one solution could be adding an /introspection/validate api call which would make sure the node is in a state which is acceptable for introspection | 14:17 |
| dtantsur | ifarkas, why do you think it's specific for introspection? any other operation would fail just as well, e.g. deployment | 14:17 |
| trown | ifarkas: couldnt that happen in Ironic ... ie during transition from enroll to manageable | 14:18 |
| trown | or other transitions too maybe | 14:18 |
| ifarkas | dtantsur, set_power_state wouldn't fail | 14:18 |
| dtantsur | you mean in deployment? why? it's the same operation in the same order | 14:18 |
| lucasagomes | dtantsur, are you working on that fix for the gate? | 14:19 |
| dtantsur | lucasagomes, jroll is | 14:20 |
| lucasagomes | oh jroll is | 14:20 |
| lucasagomes | nice ok | 14:20 |
| ifarkas | trown, that might work as well if there's a specific transition tied to introspection | 14:20 |
| dtantsur | sigh. it's not tied to introspection. introspection does the same things as deploy | 14:21 |
| ifarkas | dtantsur, i am not sure I understand but set_power_state call wouldn't fail during deployment as well | 14:21 |
| trown | ifarkas: well enroll to managable should confirm that the node is indeed manageable, i.e. validate we can set boot device | 14:21 |
| dtantsur | trown, interesting idea. not covered by the current enroll code, but should be implementable. | 14:22 |
| dtantsur | trown, we can generalize it as check_sanity call on the power interface :) | 14:23 |
| *** pal has quit IRC | 14:23 | |
| ifarkas | dtantsur, that's point it's not tied to the power interface | 14:24 |
| ifarkas | dtantsur, admin should be able to restart the node anytime | 14:24 |
| dtantsur | ifarkas, well, to management interface even. As the problem is: we can't set boot device. Nobody can: introspection cannot, deployment cannot, cleaning cannot | 14:25 |
| dtantsur | now we have to "fix" it somehow. probably we should have done the root cause analysis better | 14:25 |
| dtantsur | before jumping into solutions | 14:25 |
| *** pal has joined #openstack-ironic | 14:26 | |
| ifarkas | dtantsur, yep, that's correct, it's tied to management interface | 14:26 |
| dtantsur | ifarkas, I see we have 2 problems: 1. one can't set boot device twice - easy to fix | 14:26 |
| ifarkas | dtantsur, yep | 14:27 |
| dtantsur | 2. if there stale jobs on DRAC, everything might fail to work | 14:27 |
| dtantsur | I'm starting to think that such node should be put in maintenance mode | 14:27 |
| ifarkas | dtantsur, it might or might not | 14:27 |
| trown | I think the tough part is determining "stale" | 14:28 |
| ifarkas | dtantsur, there might be jobs from bios or raid configuration | 14:28 |
| dtantsur | ifarkas, they should hold a lock on a node | 14:28 |
| rook | trown: i think that should be easy | 14:28 |
| rook | time0 check for jobs, timen if job id is the same - stale | 14:29 |
| rook | stale and being the % complete never changs | 14:29 |
| rook | ^ maybe too simple? | 14:29 |
| dtantsur | rook, I have even simpler idea with my ironic-core hat on | 14:29 |
| dtantsur | if something is going on on a node that ironic is not aware of (aka lock is not held) - node goes to maintenance | 14:29 |
| dtantsur | amen | 14:30 |
| ifarkas | dtantsur, adding lock around every operation that creates a remote jobs might work... | 14:30 |
| dtantsur | ifarkas, we have to do it. you don't want your BIOS update to be interrupted.. | 14:30 |
| ifarkas | dtantsur, right | 14:30 |
| dtantsur | that was my initial objection against DRAC passthru operations: it ignored our locking and state machine considerations | 14:31 |
| dtantsur | I think it still does, which is a problem | 14:31 |
| rook | lol - dtantsur drops mic | 14:32 |
| *** romcheg has quit IRC | 14:37 | |
| jroll | lucasagomes: dtantsur https://review.openstack.org/#/c/203043/ | 14:38 |
| *** dlpartain has joined #openstack-ironic | 14:39 | |
| jroll | the pin landed | 14:39 |
| lucasagomes | jroll, oh... hmm we should update our requirements then | 14:39 |
| dtantsur | I guess the bot will do it soon | 14:39 |
| lucasagomes | tho it would be nice to fix that test anyway | 14:39 |
| jroll | lucasagomes: it hasn't merged yet... requirmenets bump will come out after it merges | 14:39 |
| jroll | yeah | 14:39 |
| jroll | but hard to fix in a compatible way | 14:39 |
| lucasagomes | I see... ok so should we wait the bot and then merge the pin ? | 14:39 |
| jroll | yeah | 14:40 |
| lucasagomes | ack | 14:40 |
| jroll | also, I still want to do a server release | 14:40 |
| jroll | releasing the client is what breaks default behavior | 14:40 |
| jroll | so I think we should release the server and then figure out the client | 14:40 |
| lucasagomes | you mean for the enroll state thing right? | 14:41 |
| jroll | yeah | 14:41 |
| jroll | it will still default to old behavior | 14:41 |
| jroll | (when using the client) | 14:42 |
| jroll | until we release a new client | 14:42 |
| lucasagomes | true... the only problem is if we want to include the 'abort' clenwait/deploywait | 14:42 |
| lucasagomes | that will need to bump the microversion again | 14:42 |
| dtantsur | mmm, versioning rocks :) | 14:42 |
| lucasagomes | and the client would probably need to point to the new one which will also introduce the enroll | 14:42 |
| lucasagomes | but, I'm ok just releasing without it | 14:42 |
| jroll | we can release again in a few weeks with it if we want :) | 14:43 |
| jroll | I'm going to repeat myself again | 14:43 |
| jroll | we should require a version argument. | 14:43 |
| dtantsur | what we're doing with our client is working around the whole versioning | 14:43 |
| jroll | right, the whole version negotiation thing is just causing problems | 14:43 |
| *** mgarza_ has joined #openstack-ironic | 14:43 | |
| jroll | because clients don't think about what they're doing | 14:43 |
| jroll | it's trying to pretend we don't do versioning | 14:43 |
| dtantsur | exactly! | 14:44 |
| jroll | 99% of the time I don't use the client, fwiw | 14:44 |
| dtantsur | then let us either 1. not do it at all, 2. do it completely | 14:44 |
| jroll | partially for this reason | 14:44 |
| jroll | so anyway let's figure out the client thing later | 14:44 |
| *** krtaylor has quit IRC | 14:45 | |
| jroll | so now the only thing to figure out is... | 14:45 |
| lucasagomes | jroll, make --ironic-api-version mandatory? | 14:46 |
| jroll | https://github.com/openstack/ironic/blob/master/setup.cfg#L3 | 14:46 |
| jroll | what version should we be? :) | 14:46 |
| *** hemna has joined #openstack-ironic | 14:46 | |
| jroll | lucasagomes: yes | 14:46 |
| lucasagomes | hmm /me thinks | 14:46 |
| lucasagomes | jroll, dtantsur also https://review.openstack.org/#/c/202565/ | 14:47 |
| lucasagomes | re version negotiation | 14:47 |
| jroll | dhellmann: hey, I don't see a follow up to https://review.openstack.org/#/c/192404/ | 14:47 |
| jroll | dhellmann: looking at doing a release and trying to figure out version number things | 14:47 |
| *** romcheg has joined #openstack-ironic | 14:48 | |
| *** dlpartain has quit IRC | 14:49 | |
| *** mtanino has joined #openstack-ironic | 14:49 | |
| jroll | lucasagomes: that test isn't very obvious | 14:50 |
| jroll | the end result is 1.6 | 14:50 |
| * jroll tests locally I guess | 14:50 | |
| *** hemna has quit IRC | 14:52 | |
| *** achanda has joined #openstack-ironic | 14:53 | |
| jroll | ok yeah, it works | 14:53 |
| jroll | or I should say it fails without the code change | 14:53 |
| jroll | +A | 14:54 |
| lucasagomes | yeah, I know it's not straight forward | 14:56 |
| lucasagomes | and currently the client is at 1.8 the test is adding 1.10 | 14:56 |
| lucasagomes | it's wierd, but it's just to exercise the change | 14:56 |
| NobodyCam | i dont think it weird :-p | 14:56 |
| jroll | nah I don't mind it exercising versions we don't have | 14:57 |
| *** krtaylor has joined #openstack-ironic | 14:57 | |
| jroll | just that it isn't immediately obvious why it chooses 1.6 | 14:57 |
| jroll | but it's fine | 14:57 |
| *** mgoddard1 has joined #openstack-ironic | 15:00 | |
| *** jaypipes has joined #openstack-ironic | 15:01 | |
| *** mgoddard has quit IRC | 15:01 | |
| *** dprince has quit IRC | 15:03 | |
| *** ibravo has joined #openstack-ironic | 15:04 | |
| *** pal has quit IRC | 15:04 | |
| *** achanda has quit IRC | 15:05 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Insert artificial delay between sending virtual nodes on introspection https://review.openstack.org/203040 | 15:11 |
| dtantsur | will probably call it a day, see you next week | 15:13 |
| *** dtantsur is now known as dtantsur|afk | 15:14 | |
| NobodyCam | have a great weekend dtantsur|afk | 15:14 |
| *** dprince has joined #openstack-ironic | 15:16 | |
| *** coolsvap is now known as coolsvap|away | 15:23 | |
| JoshNang | looks like i picked the wrong day to be sick yesterday :/ | 15:27 |
| *** thiagop has quit IRC | 15:27 | |
| *** cinerama has joined #openstack-ironic | 15:27 | |
| *** jaypipes is now known as blockedpipes | 15:27 | |
| *** ibravo has quit IRC | 15:29 | |
| lucasagomes | JoshNang, feeling better today? | 15:34 |
| *** ricky1 has joined #openstack-ironic | 15:37 | |
| JoshNang | lucasagomes: definitely :) | 15:40 |
| JoshNang | looks like the discussion for vlans in metadata is on the mailing list now. | 15:40 |
| *** chlong has quit IRC | 15:49 | |
| *** romainh has left #openstack-ironic | 15:52 | |
| *** yog_ has quit IRC | 15:58 | |
| *** mgoddard1 has quit IRC | 16:02 | |
| *** mgoddard has joined #openstack-ironic | 16:03 | |
| openstackgerrit | Merged openstack/python-ironicclient: Fix version negotiation https://review.openstack.org/202565 | 16:03 |
| *** amotoki has joined #openstack-ironic | 16:03 | |
| *** egon_ is now known as egon | 16:04 | |
| *** eglute has quit IRC | 16:04 | |
| *** eglute has joined #openstack-ironic | 16:05 | |
| *** romcheg has quit IRC | 16:10 | |
| *** zz_natorious is now known as natorious | 16:12 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/202699 | 16:13 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/202700 | 16:13 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/203115 | 16:13 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/201608 | 16:17 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/203136 | 16:17 |
| lucasagomes | jroll, ^ | 16:22 |
| devananda | mmm, versionning for breakfast | 16:22 |
| devananda | (g'morning, all) | 16:22 |
| lucasagomes | devananda, morning | 16:22 |
| lucasagomes | devananda, yeah it's actually fixing our gate, because mock 1.1.4 broke it | 16:23 |
| NobodyCam | good mornign devananda | 16:23 |
| devananda | yay a nother broken mock release? :( | 16:23 |
| jroll | woo | 16:23 |
| lucasagomes | devananda, unfortunately... well the first one was actually good. We were using mock incorrectly and it just showed us that | 16:23 |
| *** eglute has quit IRC | 16:23 | |
| jroll | do we need 2 +2 on these? | 16:24 |
| devananda | lucasagomes: oh | 16:24 |
| lucasagomes | jroll, usually yes | 16:24 |
| lucasagomes | devananda, now this one with 1.1.4 is actually a bug | 16:24 |
| jroll | can we change that? :P | 16:24 |
| *** eglute has joined #openstack-ironic | 16:24 | |
| lucasagomes | jroll, we just do it to have a simple rule that all patches require 2+2 | 16:24 |
| lucasagomes | but if it's something trivial I don't care if it's 1+2 | 16:24 |
| jroll | aw, I don't have +2 on python-ironic-inspector-client :( | 16:25 |
| jroll | nor inspector | 16:25 |
| jroll | anyway, some of those updates are now +A, others are the max vote I have | 16:25 |
| lucasagomes | devananda, for context https://github.com/testing-cabal/mock/issues/288 | 16:26 |
| lucasagomes | jroll, cool thank you | 16:26 |
| devananda | jroll: what's your take on the default ENROLL state change? | 16:26 |
| jroll | devananda: so | 16:26 |
| jroll | I think we should release the server now | 16:26 |
| jroll | and then I think we should talk about the client | 16:27 |
| jroll | devananda: and my opinion on the client is that we should always require a version and not do negotiation | 16:27 |
| jroll | 14:43:41 jroll | right, the whole version negotiation thing is just causing problems | 16:27 |
| jroll | 14:43:48 jroll | because clients don't think about what they're doing | 16:27 |
| jroll | 14:43:55 jroll | it's trying to pretend we don't do versioning | 16:27 |
| devananda | a) would you have the server reject connections from older clients, or clients that don't pass the header? | 16:28 |
| devananda | b) what version header should the next release of the client default to sending? | 16:29 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Fix the DEPLOYWAIT check for agent_* drivers https://review.openstack.org/203157 | 16:29 |
| lucasagomes | jroll, ^ you may be interested on that | 16:29 |
| devananda | c) what if that client is pointed at a Kilo released server? | 16:29 |
| jroll | devananda: I think I would have the client either error before hitting the server, or pass the minimum version | 16:30 |
| devananda | d) how do we signal to users ofthe new client "hey, your scripts are going to break when you create a Node" ?? | 16:30 |
| jroll | lucasagomes: oh good call | 16:30 |
| jroll | thanks, will look | 16:30 |
| lucasagomes | jroll, cool, currently our default callback_timeout is big enough. But if one tries to deploy something like a windows image which is pretty big it may fail due that timeout | 16:31 |
| *** jistr has quit IRC | 16:31 | |
| jroll | devananda: those are good questions I need to think about more | 16:31 |
| jroll | lucasagomes: yep, makes sense | 16:31 |
| devananda | we collectively, myself included, didn't actually think through the impact that changing the default state is going to have. every single user who has any automation around "node-create" is going to be broken by this becaues they can not choose not to opt in to the change | 16:31 |
| jroll | devananda: what | 16:31 |
| jroll | devananda: they certainly can choose to not opt in | 16:32 |
| devananda | nope | 16:32 |
| devananda | they can choose not to upgrade | 16:32 |
| jroll | no | 16:32 |
| jroll | they can use the previous api version | 16:32 |
| devananda | right | 16:32 |
| devananda | they can choose "do not give me any new features" | 16:32 |
| jroll | "until I fix my shit" | 16:32 |
| devananda | they CAN NOT choose "give me new features but let me keep my old work flow" | 16:32 |
| *** ashuk has joined #openstack-ironic | 16:32 | |
| *** saripurigopi has joined #openstack-ironic | 16:33 | |
| jroll | they actually can | 16:33 |
| devananda | oh? | 16:33 |
| jroll | they can use the old version only for node-create | 16:33 |
| *** pal has joined #openstack-ironic | 16:33 | |
| jroll | and newer versions otherwise | 16:33 |
| devananda | :( | 16:33 |
| jroll | or whatever | 16:33 |
| devananda | that is a HORRIBLE interface | 16:33 |
| jroll | until they fix their code | 16:33 |
| jroll | so the things is | 16:33 |
| *** eglute has quit IRC | 16:33 | |
| lucasagomes | yeah the mix and match is not ideal... but that's how it is actually | 16:34 |
| jroll | we've caused this by letting the client choose a version for them | 16:34 |
| devananda | how are they supposed to discover that? | 16:34 |
| jroll | release notes? | 16:34 |
| *** eglute has joined #openstack-ironic | 16:34 | |
| lucasagomes | unless we have a config option to pin the ironic-api server in a specific version | 16:34 |
| jroll | this whole thing is caused by us defaulting clients to whatever version *we* think is best | 16:34 |
| jroll | and so users don't think about the impact of using a given version | 16:34 |
| jroll | our client pretends we don't do versioning at all | 16:35 |
| jroll | if we forced users to think about what they're doing, and choose a version explicitly, this wouldn't be a problem | 16:36 |
| jroll | "oh, I need this new feature, let me upgrade to version X, let me also check if that will break me" | 16:36 |
| *** ricky1 is now known as rcarrillocruz | 16:36 | |
| jroll | instead they're like "HEY YOUR NEW CLIENT BROKE ME WTF" | 16:36 |
| jroll | agree/disagree? | 16:38 |
| *** max_lobur has quit IRC | 16:38 | |
| devananda | jroll: right. so you're saying, client needs to default to 1.1 (juno-ish) or 1.6 (kilo) -- which it does | 16:39 |
| jroll | um, does it? | 16:39 |
| jroll | AIUI it finds the max version that the client and server both support. | 16:39 |
| devananda | hm, it got bumped to 1.8 | 16:39 |
| jroll | also, I'm saying the client needs to default to "tell me the version" | 16:39 |
| devananda | ironicclient/common/http.py:DEFAULT_VER = '1.8' | 16:40 |
| devananda | jroll: "default to tell me the version" sounds like "error if user does not "--version xx.yy" | 16:41 |
| devananda | jroll: which is also a breaking change | 16:41 |
| jroll | devananda: yes. | 16:42 |
| jroll | devananda: right, I haven't figured out the mechanics of that deprecation or whatever | 16:42 |
| jroll | at any rate, if the client actually defaults to 1.8, then I don't see how releasing ENROLL breaks people | 16:42 |
| devananda | because a minor version change should not break a user | 16:43 |
| lucasagomes | jroll, it won't tho usually when we add stuff to the client to work with the new things we also bump that version | 16:43 |
| devananda | the moment anyone starts sending 1.10, they will wonder "omg why did my system stop doing what it used to do" | 16:43 |
| jroll | ... that's the whole point of "micro" versions | 16:43 |
| lucasagomes | I believe that requiring the user to always say which version he wants in the long run can actually be a good thing | 16:43 |
| lucasagomes | it's explicit | 16:43 |
| jroll | exactly. make the user think about what they're doing. | 16:44 |
| lucasagomes | tho as devananda it's a breaking change, we would probably need to communicate it very loud | 16:44 |
| jroll | instead of just YOLO HOPE THIS ALWAYS WORKS | 16:44 |
| devananda | having the server require it -- I'm reasonably OK with | 16:44 |
| jroll | we already require something like 5 arguments in our client, why not one more? | 16:45 |
| jroll | maybe just 4. | 16:45 |
| jroll | username, password, tenant, auth_url | 16:45 |
| lucasagomes | what we could do is to release a version of the client now pinned in 1.8 just as-is | 16:45 |
| lucasagomes | but when the version is not specified print a big warning message | 16:46 |
| jroll | mhmm | 16:46 |
| lucasagomes | saying that in the future the version will be mandatory | 16:46 |
| lucasagomes | at least it bring visibility (if we go in the path of making it mandatory) | 16:46 |
| jroll | let me point out a recent nova client change: | 16:46 |
| jroll | WARNING: Command volume-list is deprecated and will be removed after Nova 2016.1 is released. Use python-cinderclient or openstackclient instead. | 16:46 |
| devananda | jroll: it requires 2 ENV vars: token and endpoint | 16:46 |
| jroll | devananda: ok sure | 16:47 |
| jroll | require 3 instead | 16:47 |
| jroll | I don't see why it's a problem | 16:47 |
| jroll | most people won't be using token though, they'll be using username/password | 16:47 |
| jroll | and auth url | 16:47 |
| jroll | just another thing to stick in the file you source when you run ironic cli | 16:48 |
| devananda | in any case, ^ exactly | 16:48 |
| jroll | I'm happy to bring this to the mailing list if we want others' input | 16:48 |
| jroll | dtantsur|afk seemed to agree earlier, though, too | 16:49 |
| devananda | yes pls | 16:49 |
| jroll | ok, I'll do that at some point today | 16:49 |
| devananda | cause I think every angle on this bad for our users - -and the same sort of things are big concerns for nova and soon for other projects looking at versions | 16:49 |
| devananda | nova has a much longer history and historically much higher aversion to change ... | 16:50 |
| jroll | before everyone disappears for oscon/nova midcycle >.> | 16:50 |
| devananda | heh | 16:50 |
| * devananda disappears to work on his talk for OSCON | 16:51 | |
| jroll | mhmm | 16:51 |
| jroll | devananda: do you know how nova's client behaves with this right now? | 16:51 |
| *** natorious has quit IRC | 16:52 | |
| *** natorious has joined #openstack-ironic | 16:54 | |
| lucasagomes | would be good to investigate | 16:55 |
| lucasagomes | actually I think it worth to add [nova] to the ML | 16:55 |
| jroll | yeah, I will | 16:57 |
| *** ijw_ has joined #openstack-ironic | 17:01 | |
| *** derekh has quit IRC | 17:01 | |
| *** ijw__ has joined #openstack-ironic | 17:04 | |
| mgagne | jroll: regarding Nova midcycle meetup, I just thought about something Nova should fix or handle better but it's not related to Ironic :-/ | 17:05 |
| mgagne | jroll: it's about how Nova supports (or not) multiple nics and security groups :D | 17:06 |
| *** ijw_ has quit IRC | 17:07 | |
| lucasagomes | folks I'm going to call it a day | 17:10 |
| lucasagomes | have a great night everyone, enjoy the weekend! | 17:11 |
| *** lucasagomes is now known as lucas-beer | 17:11 | |
| NobodyCam | have great weekend lucas-beer | 17:12 |
| *** trown is now known as trown|lunch | 17:12 | |
| *** Pradip has joined #openstack-ironic | 17:18 | |
| *** Nisha has joined #openstack-ironic | 17:19 | |
| * NobodyCam eats a bagel | 17:21 | |
| *** pal has quit IRC | 17:21 | |
| *** mitchjameson has joined #openstack-ironic | 17:24 | |
| *** mitchjameson has quit IRC | 17:24 | |
| *** bizarrochristy has joined #openstack-ironic | 17:24 | |
| *** harlowja has quit IRC | 17:26 | |
| *** harlowja has joined #openstack-ironic | 17:26 | |
| *** ashuk has quit IRC | 17:26 | |
| *** praneshp has joined #openstack-ironic | 17:31 | |
| *** praneshp has left #openstack-ironic | 17:31 | |
| *** sambetts has quit IRC | 17:32 | |
| *** amotoki has quit IRC | 17:33 | |
| *** sambetts has joined #openstack-ironic | 17:35 | |
| *** achanda has joined #openstack-ironic | 17:35 | |
| *** eglute has quit IRC | 17:35 | |
| *** eglute has joined #openstack-ironic | 17:36 | |
| jroll | mgagne: interesting, I know nothing about that :P | 17:36 |
| *** rook has quit IRC | 17:42 | |
| *** Sukhdev has joined #openstack-ironic | 17:43 | |
| *** e0ne has quit IRC | 17:44 | |
| *** boris-42 has joined #openstack-ironic | 17:44 | |
| *** karimb has quit IRC | 17:44 | |
| mgagne | jroll: could you link to the driver supporting teaming + trunked vlans? | 17:48 |
| *** max_lobur has joined #openstack-ironic | 17:51 | |
| *** romcheg has joined #openstack-ironic | 17:52 | |
| jroll | mgagne: damnit, I'm bad at emails | 17:53 |
| jroll | mgagne: send another reply with it, but it's https://github.com/rackerlabs/ironic-neutron-plugin | 17:54 |
| mgagne | jroll: haha "some drivers support it, obviously ours :D" | 17:55 |
| jroll | :) | 17:56 |
| mgagne | jroll: I think our use case is similar to this blueprint: https://blueprints.launchpad.net/neutron/+spec/vlan-aware-vms | 17:56 |
| jroll | mgagne: yep, that was highlighted as a thing | 17:57 |
| mgagne | jroll: so I guess it's a matter of implementing the specs | 17:57 |
| *** lucas-beer has quit IRC | 17:57 | |
| openstackgerrit | Christopher Dearborn proposed openstack/ironic: This change is to test the corporate firewall https://review.openstack.org/203189 | 17:58 |
| devananda | ^ lol ? | 17:59 |
| jroll | hahaha | 17:59 |
| TheJulia | lol | 17:59 |
| devananda | LOL | 18:00 |
| devananda | This change is to test the corporate firewall. Please do not review or merge. | 18:00 |
| devananda | I may have been watching a lot of silly cat videos recently | 18:00 |
| devananda | so I find this very amusing | 18:00 |
| devananda | TheJulia: it's your fault, btw :) | 18:01 |
| TheJulia | \o/ | 18:02 |
| TheJulia | at least it is death by cute | 18:02 |
| *** vdrok has quit IRC | 18:02 | |
| *** athomas has quit IRC | 18:02 | |
| *** puranamr has joined #openstack-ironic | 18:03 | |
| *** praneshp has joined #openstack-ironic | 18:05 | |
| *** JayF has quit IRC | 18:09 | |
| *** JayF has joined #openstack-ironic | 18:09 | |
| *** athomas has joined #openstack-ironic | 18:11 | |
| *** puranamr has quit IRC | 18:16 | |
| *** Nisha_away has joined #openstack-ironic | 18:20 | |
| *** Nisha has quit IRC | 18:20 | |
| *** puranamr has joined #openstack-ironic | 18:25 | |
| *** praneshp has quit IRC | 18:25 | |
| *** praneshp has joined #openstack-ironic | 18:28 | |
| *** e0ne has joined #openstack-ironic | 18:29 | |
| *** puranamr has quit IRC | 18:29 | |
| *** puranamr has joined #openstack-ironic | 18:30 | |
| *** puranamr has quit IRC | 18:30 | |
| *** puranamr has joined #openstack-ironic | 18:32 | |
| *** puranamr has quit IRC | 18:35 | |
| devananda | ok - something is bothering me in our API again (no surprise, right?) | 18:36 |
| devananda | I can't walk the REST API in any way that reveals the node vendor_passthru end point | 18:36 |
| devananda | I mean - i'm all for discouraging its use | 18:37 |
| devananda | but it's not discoverable at all | 18:37 |
| *** praneshp_ has joined #openstack-ironic | 18:38 | |
| *** praneshp has quit IRC | 18:41 | |
| *** praneshp_ is now known as praneshp | 18:41 | |
| *** puranamr has joined #openstack-ironic | 18:43 | |
| *** puranamr has quit IRC | 18:44 | |
| *** puranamr has joined #openstack-ironic | 18:50 | |
| *** praneshp has left #openstack-ironic | 18:53 | |
| *** thiagop has joined #openstack-ironic | 18:58 | |
| *** saripurigopi has quit IRC | 19:00 | |
| *** natorious is now known as zz_natorious | 19:07 | |
| *** ndipanov has quit IRC | 19:09 | |
| *** trown|lunch is now known as trown | 19:20 | |
| openstackgerrit | Merged openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/201608 | 19:26 |
| lifeless | jroll: around? | 19:38 |
| lifeless | jroll: actually -> nova | 19:38 |
| *** Sukhdev has quit IRC | 19:43 | |
| *** ig0r_ has quit IRC | 19:47 | |
| *** Nisha has joined #openstack-ironic | 20:02 | |
| *** Nisha_away has quit IRC | 20:02 | |
| *** jjohnson2 has joined #openstack-ironic | 20:05 | |
| *** mjturek1 has quit IRC | 20:06 | |
| *** afaranha has quit IRC | 20:28 | |
| *** lsmola has quit IRC | 20:35 | |
| *** e0ne has quit IRC | 20:43 | |
| *** jjohnson2_ has joined #openstack-ironic | 20:44 | |
| *** Pradip has quit IRC | 20:44 | |
| *** jjohnson2 has quit IRC | 20:46 | |
| *** Pradip has joined #openstack-ironic | 20:48 | |
| *** dprince has quit IRC | 20:53 | |
| *** bnemec has quit IRC | 20:54 | |
| *** ijw__ has quit IRC | 20:57 | |
| *** trown is now known as trown|outttypeww | 21:03 | |
| *** bizarrochristy has quit IRC | 21:10 | |
| *** dguerri is now known as dguerri` | 21:12 | |
| *** ijw_ has joined #openstack-ironic | 21:18 | |
| *** max_lobur has quit IRC | 21:22 | |
| *** Nisha has quit IRC | 21:23 | |
| *** blockedpipes has quit IRC | 21:29 | |
| *** jjohnson2_ has quit IRC | 21:30 | |
| devananda | if anyone is randomly curious (like me) what vendor passthru methods are being exposed by drivers today -- http://paste.openstack.org/show/Vms7qxYrK7mbrFBe7Ytn/ | 21:49 |
| *** Pradip has quit IRC | 22:14 | |
| *** davideagnello has quit IRC | 22:21 | |
| jroll | devananda: no agent driver? | 22:22 |
| jroll | I'm also curious if there's a designation between node- and driver-vendor-passthru in those responses | 22:22 |
| jroll | but I think only agent_* has driver- | 22:22 |
| devananda | jroll: there is not | 22:24 |
| devananda | this is coming from the REST API, which can not distinguish | 22:24 |
| jroll | :/ | 22:25 |
| devananda | also it is not a discoverable endpoint | 22:25 |
| jroll | right, you mentioned that | 22:25 |
| devananda | there are actually a few non-discoverable end points | 22:26 |
| devananda | /v1/nodes/fake/vendor_passthru/methods generated that list ^ | 22:26 |
| devananda | s/fake/every driver/ | 22:26 |
| jroll | s/every driver/pxe_*/ | 22:26 |
| jroll | :P | 22:26 |
| devananda | nope | 22:26 |
| jroll | wait what | 22:26 |
| devananda | oh wait, did i? | 22:26 |
| jroll | idk | 22:26 |
| devananda | oh lol | 22:27 |
| jroll | there's no agent in there | 22:27 |
| devananda | grep snuck in | 22:27 |
| devananda | http://paste.openstack.org/raw/Upd7hbpi2K7zzazh9Lxu/ | 22:28 |
| devananda | there we go | 22:28 |
| jroll | so really that doesn't help discover anything | 22:28 |
| jroll | because they don't tell you it's at /v1/nodes/xxx/vendor... or /v1/drivers/xxx/vendor... | 22:29 |
| jroll | oh, driver-vendor-passthru isn't in that list at all | 22:29 |
| jroll | good! | 22:29 |
| jroll | :| | 22:29 |
| devananda | no - that's a whole different thing! | 22:29 |
| jroll | oh you has /v1/nodes/blahblahblah | 22:30 |
| jroll | had* | 22:30 |
| *** thiagop has quit IRC | 22:30 | |
| devananda | so v1/drivers/xxx/vendor_passthru/methods always returns this | 22:34 |
| devananda | http://paste.openstack.org/show/384740/ | 22:35 |
| devananda | for every driver | 22:35 |
| devananda | I think it's a bug :( | 22:35 |
| devananda | like - it returns that for pxe_ssh too | 22:36 |
| devananda | anyway - i gotta run | 22:36 |
| NobodyCam | have a great weekend devananda | 22:36 |
| jroll | hrm | 22:36 |
| NobodyCam | :) | 22:36 |
| jroll | have a good one | 22:36 |
| NobodyCam | ya kinda looks like what a bug would do | 22:37 |
| *** ijw_ has quit IRC | 22:48 | |
| *** mgarza_ has quit IRC | 22:49 | |
| *** achanda has quit IRC | 23:04 | |
| *** ijw_ has joined #openstack-ironic | 23:06 | |
| *** ijw_ has quit IRC | 23:12 | |
| *** ijw_ has joined #openstack-ironic | 23:13 | |
| *** davideagnello has joined #openstack-ironic | 23:17 | |
| *** davideagnello has quit IRC | 23:20 | |
| *** bizarrochristy has joined #openstack-ironic | 23:22 | |
| *** ijw_ has quit IRC | 23:26 | |
| *** cing has quit IRC | 23:26 | |
| *** athomas has quit IRC | 23:26 | |
| *** bizarrochristy has quit IRC | 23:27 | |
| *** puranamr has quit IRC | 23:32 | |
| *** puranamr has joined #openstack-ironic | 23:33 | |
| *** puranamr has quit IRC | 23:42 | |
| *** puranamr has joined #openstack-ironic | 23:43 | |
| *** Pradip has joined #openstack-ironic | 23:43 | |
| Pradip | hi, I am trying to use ironic inspector, it is giving me and error as follows " Driver pxe_ipmitool does not support inspect (disabled or not implemented)." , any ideas? | 23:46 |
| *** trown|outttypeww is now known as trown | 23:46 | |
| jroll | Pradip: you need to enable inspector in config | 23:48 |
| jroll | CONF.inspector.enabled | 23:48 |
| trown | Pradip: there are two ways to use ironic inspector, one is via an inspect driver, but there is also a cli client | 23:48 |
| jroll | you should see a log with the same in the conductor log | 23:48 |
| Pradip | btw I am using it with devstack | 23:49 |
| trown | ah so pxe_ipmitool has an inspect driver...wasnt sure about that | 23:49 |
| jroll | Pradip: yes, set [inspector] enabled=true | 23:49 |
| jroll | in the ironic config | 23:49 |
| Pradip | it is set there | 23:49 |
| * jroll checks localrc | 23:49 | |
| jroll | did you restart the conductor after setting it? | 23:50 |
| Pradip | yes i did | 23:50 |
| Pradip | just to be sure, pxe-ipmi supports the inspector, right? | 23:51 |
| jroll | hmmm | 23:51 |
| jroll | it looks like it, yes | 23:51 |
| Pradip | then i am not supposed to see this error . huh | 23:52 |
| trown | Pradip: where is that error coming from? | 23:53 |
| Pradip | trown: conductor/manager.py | 23:54 |
| Pradip | http://paste.openstack.org/show/384952/ | 23:56 |
| trown | Pradip: hmm, the line number there does not match up with trunk, is this a new devstack env? | 23:58 |
| trown | Pradip: or could you provide the git hash of HEAD in /opt/stack/ironic | 23:59 |
| Pradip | trown: I used the stable/kilo version | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!