| dtrainor | Anyone have an example of using driver_info in a node definition file? | 00:01 |
|---|---|---|
| dtrainor | i want to use pxe_snmp but i understand that I need to create some properties of that driver under driver_info | 00:02 |
| rloo | dtrainor: this might help https://docs.openstack.org/ironic/latest/admin/drivers/snmp.html | 00:04 |
| rloo | dtrainor: i gotta run. sorry. | 00:04 |
| dtrainor | yea, that's where i saw the need for driver_info :) | 00:05 |
| dtrainor | thanks | 00:05 |
| dtrainor | i get: ValueError: Expecting , delimiter: line 7 column 18 (char 134) | 00:05 |
| *** zshi_laptop has joined #openstack-ironic | 00:12 | |
| *** efried has quit IRC | 00:15 | |
| *** efried has joined #openstack-ironic | 00:25 | |
| *** makowals has quit IRC | 00:30 | |
| *** charcol has quit IRC | 00:34 | |
| *** charcol has joined #openstack-ironic | 00:35 | |
| *** makowals has joined #openstack-ironic | 00:42 | |
| *** zhangfei has joined #openstack-ironic | 00:43 | |
| *** zhangfei has quit IRC | 00:47 | |
| *** makowals has quit IRC | 00:53 | |
| *** rbudden has quit IRC | 00:54 | |
| *** zhangfei has joined #openstack-ironic | 01:00 | |
| *** rloo has quit IRC | 01:01 | |
| *** priteau has joined #openstack-ironic | 01:03 | |
| *** efried has quit IRC | 01:03 | |
| *** makowals has joined #openstack-ironic | 01:07 | |
| *** priteau has quit IRC | 01:07 | |
| openstackgerrit | Merged openstack/ironic master: Documentation for 'snmp' hardware type https://review.openstack.org/493468 | 01:07 |
| *** zshi_laptop has quit IRC | 01:11 | |
| *** efried has joined #openstack-ironic | 01:13 | |
| *** wong has joined #openstack-ironic | 01:14 | |
| wong | morning | 01:14 |
| *** yangyapeng has joined #openstack-ironic | 01:15 | |
| *** dyasny has joined #openstack-ironic | 01:19 | |
| *** ijw has joined #openstack-ironic | 01:23 | |
| *** ijw has quit IRC | 01:23 | |
| *** litao__ has joined #openstack-ironic | 01:29 | |
| *** makowals has quit IRC | 01:30 | |
| *** shweta has quit IRC | 01:34 | |
| *** makowals has joined #openstack-ironic | 01:54 | |
| *** makowals has quit IRC | 02:03 | |
| *** yangyapeng has quit IRC | 02:10 | |
| *** yangyapeng has joined #openstack-ironic | 02:10 | |
| *** cdearborn_ has quit IRC | 02:19 | |
| *** makowals has joined #openstack-ironic | 02:28 | |
| openstackgerrit | Dao Cong Tien proposed openstack/ironic master: Documentation for irmc hardware type https://review.openstack.org/490483 | 02:33 |
| *** weshay is now known as weshay_PTO_AM | 02:37 | |
| *** mtanino has quit IRC | 02:40 | |
| TheJulia | dtrainor: it might help if you could provide an direct example of what your sending the api to get such an error | 02:49 |
| TheJulia | wong: good morning, and goodnight :) | 02:49 |
| wong | :) | 02:50 |
| *** priteau has joined #openstack-ironic | 03:03 | |
| *** priteau has quit IRC | 03:08 | |
| jobcespedes | TheJulia: Thanks for your reply, it confirms the options available. I'd rather have automatation, but some colaborators, with a better secutiry backgroud, have those concerns. | 03:12 |
| *** aparnav has joined #openstack-ironic | 03:15 | |
| TheJulia | jobcespedes: understood, and they are natural security concerns to have, which ultimately centralizes on the risk/reward trade-off of automation, coupled with security posture, and who is actually interacting with the baremeral hardware fleet | 03:17 |
| TheJulia | I'm suddenly surprised we've not pondered better controls around boot security, in terms of executed bytecode, but really that would be a novaish discussion or a discussion of controls and processes if directly managing a fleet without nova. | 03:18 |
| TheJulia | jobcespedes: you could do uefi with the system configured to allow particular signers. | 03:20 |
| TheJulia | That might be acceptable, from a wholistic standpoint | 03:21 |
| *** yangyape_ has joined #openstack-ironic | 03:23 | |
| *** yangyapeng has quit IRC | 03:23 | |
| jobcespedes | TheJulia: Sorry, your recomendation goes beyond my expertise/knowledge :P I'm going to share it with those collaborators I mentioned, they are more prepare than me in the infrastructure and security scopes . The problem was, as I understood and as they described it, access to boot config means access to everything (if credentials were compromised) | 03:35 |
| TheJulia | jobcespedes: I'm not too familiar with the internals of ilo permissions, but it might be worth also discussing with some of the ilo devs. Nisha_Agarwal is a good person to ask when she is around. | 03:39 |
| jobcespedes | TheJulia: I'm going to try. Thank you :) | 03:41 |
| TheJulia | jobcespedes: No problem! Let us know if we can be of any further assistance. :) | 03:43 |
| *** udesale has joined #openstack-ironic | 03:43 | |
| TheJulia | And with that, I'm going to go get some sleep | 03:43 |
| TheJulia | Have a wonderful day! | 03:43 |
| wong | bye,good night | 03:54 |
| *** Nisha_Agarwal has joined #openstack-ironic | 03:54 | |
| jobcespedes | Good night! | 03:56 |
| *** jobcespedes has quit IRC | 03:57 | |
| *** links has joined #openstack-ironic | 03:57 | |
| *** ijw has joined #openstack-ironic | 04:00 | |
| *** ijw has quit IRC | 04:05 | |
| *** vmud213 has joined #openstack-ironic | 04:11 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic master: Update documentation for ``ilo`` hardware type https://review.openstack.org/496480 | 04:23 |
| openstackgerrit | zenghui.shi proposed openstack/ironic-specs master: New driver interface for BIOS configuration https://review.openstack.org/496481 | 04:41 |
| *** makowals has quit IRC | 04:42 | |
| *** MattMan_1 has quit IRC | 04:46 | |
| *** MattMan_1 has joined #openstack-ironic | 04:46 | |
| *** makowals has joined #openstack-ironic | 04:47 | |
| *** makowals has quit IRC | 04:57 | |
| *** jtomasek_ has joined #openstack-ironic | 05:01 | |
| *** aignatov has quit IRC | 05:02 | |
| *** jtomasek has quit IRC | 05:02 | |
| *** namnh has quit IRC | 05:02 | |
| *** zhenguo has quit IRC | 05:02 | |
| *** dims has quit IRC | 05:02 | |
| *** sdake has quit IRC | 05:02 | |
| *** fpxie has quit IRC | 05:02 | |
| *** mordred has quit IRC | 05:02 | |
| *** afazekas has quit IRC | 05:02 | |
| *** sergek_ has quit IRC | 05:02 | |
| *** kiseok7 has quit IRC | 05:02 | |
| *** andyjones2001 has quit IRC | 05:02 | |
| *** eglute has quit IRC | 05:02 | |
| *** jtanner has quit IRC | 05:02 | |
| *** hogepodge has quit IRC | 05:02 | |
| *** ildikov has quit IRC | 05:02 | |
| *** pas-ha has quit IRC | 05:02 | |
| *** kencjohnston_ has quit IRC | 05:02 | |
| *** zul has quit IRC | 05:02 | |
| *** namnh has joined #openstack-ironic | 05:02 | |
| *** kiseok7 has joined #openstack-ironic | 05:02 | |
| *** zhenguo has joined #openstack-ironic | 05:02 | |
| *** hogepodge has joined #openstack-ironic | 05:02 | |
| *** afazekas has joined #openstack-ironic | 05:02 | |
| *** ildikov has joined #openstack-ironic | 05:02 | |
| *** makowals has joined #openstack-ironic | 05:03 | |
| *** sdake has joined #openstack-ironic | 05:03 | |
| *** sdake has quit IRC | 05:03 | |
| *** sdake has joined #openstack-ironic | 05:03 | |
| *** sergek_ has joined #openstack-ironic | 05:03 | |
| *** pas-ha has joined #openstack-ironic | 05:03 | |
| *** jcook has quit IRC | 05:04 | |
| *** mordred has joined #openstack-ironic | 05:04 | |
| *** eglute has joined #openstack-ironic | 05:04 | |
| *** priteau has joined #openstack-ironic | 05:04 | |
| *** kencjohnston has joined #openstack-ironic | 05:04 | |
| *** dims has joined #openstack-ironic | 05:04 | |
| *** aignatov has joined #openstack-ironic | 05:05 | |
| *** radez has quit IRC | 05:06 | |
| *** trown|outtypewww has quit IRC | 05:06 | |
| *** rook has quit IRC | 05:06 | |
| *** trown has joined #openstack-ironic | 05:06 | |
| *** jcook has joined #openstack-ironic | 05:07 | |
| *** priteau has quit IRC | 05:09 | |
| *** sasha2 has quit IRC | 05:09 | |
| *** kzaitsev_pi has joined #openstack-ironic | 05:11 | |
| *** radez has joined #openstack-ironic | 05:11 | |
| *** rook has joined #openstack-ironic | 05:12 | |
| *** rook is now known as Guest91464 | 05:12 | |
| *** priteau has joined #openstack-ironic | 05:21 | |
| *** zshi_laptop has joined #openstack-ironic | 05:27 | |
| *** jtanner has joined #openstack-ironic | 05:31 | |
| *** mjura has joined #openstack-ironic | 05:32 | |
| *** masber has quit IRC | 05:34 | |
| *** amotoki has joined #openstack-ironic | 05:34 | |
| *** baoli has joined #openstack-ironic | 05:36 | |
| *** oomichi has quit IRC | 05:37 | |
| *** oomichi has joined #openstack-ironic | 05:39 | |
| *** baoli_ has joined #openstack-ironic | 05:49 | |
| *** ChubYann has quit IRC | 05:51 | |
| *** baoli has quit IRC | 05:52 | |
| *** hoonetorg has quit IRC | 05:54 | |
| *** baoli has joined #openstack-ironic | 05:55 | |
| *** rbartal has joined #openstack-ironic | 05:56 | |
| *** charcol has quit IRC | 05:57 | |
| *** baoli_ has quit IRC | 05:58 | |
| *** priteau has quit IRC | 05:59 | |
| *** baoli has quit IRC | 06:02 | |
| *** baoli has joined #openstack-ironic | 06:03 | |
| *** hoonetorg has joined #openstack-ironic | 06:08 | |
| *** vishwana_ has joined #openstack-ironic | 06:15 | |
| *** wong_ has joined #openstack-ironic | 06:15 | |
| *** wong has quit IRC | 06:16 | |
| *** rcernin has joined #openstack-ironic | 06:16 | |
| *** vishwanathj has quit IRC | 06:18 | |
| *** jlvillal has quit IRC | 06:21 | |
| *** jlvillal has joined #openstack-ironic | 06:21 | |
| *** kevinbenton has quit IRC | 06:22 | |
| *** kevinbenton has joined #openstack-ironic | 06:24 | |
| *** baoli has quit IRC | 06:28 | |
| *** masber has joined #openstack-ironic | 06:32 | |
| *** aparnav has quit IRC | 06:39 | |
| *** aparnav has joined #openstack-ironic | 06:39 | |
| Tengu | hello ironic :) | 06:43 |
| Tengu | small question: how may I ensure ironic activated the GPT thing for volumes >2To ? I'm not 100% it's the case, apparently 200G are missing from my 2.2To volume… | 06:43 |
| *** JayF has quit IRC | 06:44 | |
| Tengu | i.e. is there any flag in the "capabilities" as shown with "ironic node show <uid>"? | 06:44 |
| *** baoli has joined #openstack-ironic | 06:45 | |
| *** yangyape_ has quit IRC | 06:47 | |
| *** yangyapeng has joined #openstack-ironic | 06:48 | |
| *** baoli has quit IRC | 06:49 | |
| *** baoli has joined #openstack-ironic | 06:51 | |
| *** baoli has quit IRC | 06:55 | |
| *** fpxie has joined #openstack-ironic | 06:59 | |
| *** andyjones2001 has joined #openstack-ironic | 07:00 | |
| *** zul has joined #openstack-ironic | 07:00 | |
| *** charcol has joined #openstack-ironic | 07:07 | |
| *** dtrainor has quit IRC | 07:09 | |
| *** pcaruana has joined #openstack-ironic | 07:11 | |
| *** tesseract-RH has joined #openstack-ironic | 07:17 | |
| *** charcol has quit IRC | 07:27 | |
| *** baoli has joined #openstack-ironic | 07:28 | |
| *** hjensas has joined #openstack-ironic | 07:31 | |
| *** hjensas has quit IRC | 07:31 | |
| *** hjensas has joined #openstack-ironic | 07:31 | |
| *** baoli has quit IRC | 07:32 | |
| *** dtrainor has joined #openstack-ironic | 07:32 | |
| *** racedo has joined #openstack-ironic | 07:38 | |
| *** hjensas has quit IRC | 07:39 | |
| wong_ | hi, anybody saw this before?http://paste.openstack.org/show/619133/ | 07:43 |
| wong_ | I type clean-step to config raid,after that,the node status was in manageable,then I made status to provide,the error appeared | 07:45 |
| *** hjensas has joined #openstack-ironic | 07:46 | |
| *** hjensas has quit IRC | 07:46 | |
| *** hjensas has joined #openstack-ironic | 07:46 | |
| *** athomas has joined #openstack-ironic | 07:49 | |
| *** vmud213 has quit IRC | 07:49 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic master: Set device_owner/vnic_type when update mac_address https://review.openstack.org/496249 | 07:53 |
| *** etingof has quit IRC | 07:55 | |
| *** livelace has joined #openstack-ironic | 07:59 | |
| *** openstackgerrit has quit IRC | 08:03 | |
| *** Nisha_Agarwal has quit IRC | 08:06 | |
| *** derekh has joined #openstack-ironic | 08:07 | |
| *** amarao has joined #openstack-ironic | 08:09 | |
| *** hjensas has quit IRC | 08:11 | |
| *** etingof has joined #openstack-ironic | 08:14 | |
| *** hjensas has joined #openstack-ironic | 08:14 | |
| *** hjensas has quit IRC | 08:14 | |
| *** hjensas has joined #openstack-ironic | 08:14 | |
| *** openstackgerrit has joined #openstack-ironic | 08:15 | |
| openstackgerrit | Merged openstack/bifrost master: Updated from global requirements https://review.openstack.org/495175 | 08:15 |
| *** karimb has joined #openstack-ironic | 08:16 | |
| *** lucas-afk is now known as lucasagomes | 08:18 | |
| *** amarao has quit IRC | 08:19 | |
| *** kristaps_ has quit IRC | 08:25 | |
| *** amarao has joined #openstack-ironic | 08:34 | |
| pas-ha | morning ironic :) | 08:35 |
| *** mgoddard has joined #openstack-ironic | 08:39 | |
| *** aparnav has quit IRC | 08:44 | |
| vdrok | morning all! | 08:45 |
| *** aparnav has joined #openstack-ironic | 08:45 | |
| *** hjensas has quit IRC | 08:47 | |
| * Tengu checks his watch | 08:50 | |
| Tengu | hmm. timezone magic ;). | 08:50 |
| *** litao__ has quit IRC | 08:51 | |
| *** fdegir has quit IRC | 08:51 | |
| Tengu | Anyone can tell me how to ensure ironic activated the GPT thing? I'm not really sure it's the case on my computes, because there's missing 200G on the 2.2To. | 08:51 |
| *** litao__ has joined #openstack-ironic | 08:51 | |
| *** rwsu has joined #openstack-ironic | 08:51 | |
| *** sergek_ has quit IRC | 08:52 | |
| *** zhenguo has quit IRC | 08:52 | |
| *** fdegir has joined #openstack-ironic | 08:52 | |
| *** sergek_ has joined #openstack-ironic | 08:52 | |
| *** zhenguo has joined #openstack-ironic | 08:53 | |
| *** hjensas has joined #openstack-ironic | 09:00 | |
| *** hjensas has joined #openstack-ironic | 09:00 | |
| makowals | Tengu: How about this one ? https://docs.openstack.org/project-install-guide/baremetal/draft/advanced.html#when-used-with-compute-service | 09:01 |
| makowals | Tengu: How about just checking "fdisk -l device" from inside the box to check what happened ? | 09:01 |
| *** hjensas has quit IRC | 09:01 | |
| *** karimb has quit IRC | 09:02 | |
| *** tuanluong has joined #openstack-ironic | 09:04 | |
| Tengu | makowals: we see the 200G at the end of the partition | 09:05 |
| Tengu | what's weird: I think it should have added the gpt flag while introspecting, shouldn't it? | 09:06 |
| makowals | Tengu: How about posting output of fdisk ? | 09:06 |
| Tengu | 2s | 09:06 |
| Tengu | makowals: http://paste.openstack.org/show/619146/ | 09:07 |
| *** e0ne has joined #openstack-ironic | 09:07 | |
| *** etingof has quit IRC | 09:08 | |
| Tengu | aha, I have to add the capability to the flavor as well. That's most probably the error I have. | 09:08 |
| makowals | Tengu: Okay, so it's obvious it's missing this gpt flag and created dos partition label; good question, should inspector even detect it ? | 09:08 |
| Tengu | makowals: I think so, yes. but according the doc you pointed, the flavor must have that flag as well. | 09:09 |
| Tengu | so detecting while introspecting would partially solve the issue | 09:09 |
| Tengu | I know now what to do, thanks :) | 09:09 |
| *** karimb has joined #openstack-ironic | 09:10 | |
| makowals | No problem, cheers ! | 09:10 |
| *** sdake has quit IRC | 09:15 | |
| *** fdegir has quit IRC | 09:16 | |
| *** sdake has joined #openstack-ironic | 09:16 | |
| *** sdake is now known as Guest28943 | 09:16 | |
| *** fdegir has joined #openstack-ironic | 09:17 | |
| *** karimb has quit IRC | 09:20 | |
| *** sambetts|afk is now known as sambetts | 09:20 | |
| sambetts | Morning all | 09:21 |
| *** etingof has joined #openstack-ironic | 09:21 | |
| lucasagomes | sambetts, morning | 09:23 |
| *** karimb has joined #openstack-ironic | 09:24 | |
| openstackgerrit | liumengke proposed openstack/ironic master: fix a typo in agent.py: s/doman/domain/ https://review.openstack.org/496574 | 09:25 |
| sambetts | o/ lucasagomes | 09:25 |
| *** Nisha_Agarwal has joined #openstack-ironic | 09:27 | |
| *** amarao1 has joined #openstack-ironic | 09:30 | |
| *** vmud213 has joined #openstack-ironic | 09:32 | |
| *** zshi has joined #openstack-ironic | 09:33 | |
| *** Nisha_Agarwal has quit IRC | 09:34 | |
| *** mtanino has joined #openstack-ironic | 09:41 | |
| *** tuanluong has quit IRC | 09:48 | |
| *** vmud213 has quit IRC | 10:13 | |
| *** yangyapeng has quit IRC | 10:18 | |
| *** hjensas has joined #openstack-ironic | 10:19 | |
| *** hjensas has quit IRC | 10:19 | |
| *** hjensas has joined #openstack-ironic | 10:19 | |
| *** amarao1 has quit IRC | 10:23 | |
| *** milan has joined #openstack-ironic | 10:25 | |
| milan | morning Ironic! :) | 10:26 |
| milan | #pixiesay -mdead alive? | 10:27 |
| milan | hmmmmm | 10:27 |
| *** dtantsur|afk is now known as dtantsur | 10:30 | |
| dtantsur | late morning folks! | 10:31 |
| sambetts | o/ dtantsur | 10:31 |
| dtantsur | oh, so my Pixie container is not so HA, I see :D | 10:31 |
| *** udesale has quit IRC | 10:31 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Finalize migration to keystoneauth adapters https://review.openstack.org/478825 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Use adapters for cinderclient https://review.openstack.org/476171 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Use adapters for neutronclient https://review.openstack.org/476170 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Rework keystone auth for glance https://review.openstack.org/467728 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Fix swiftclient creation https://review.openstack.org/476173 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Use adapters for inspectorclient https://review.openstack.org/476172 | 10:31 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic master: Introduce keystoneauth adapters for clients https://review.openstack.org/469627 | 10:31 |
| *** mtanino has quit IRC | 10:31 | |
| milan | morning dtantsur pas-ha sambetts! :) | 10:32 |
| pas-ha | morning milan | 10:32 |
| *** PixieBoots has joined #openstack-ironic | 10:34 | |
| dtantsur | #pixiesay -mmeh Containers are overrated | 10:35 |
| PixieBoots | ¯\_(ツ)_/¯: Containers are overrated | 10:35 |
| *** yuriyz has joined #openstack-ironic | 10:37 | |
| *** karimb has quit IRC | 10:38 | |
| *** Guest91464 is now known as rook | 10:39 | |
| *** jkilpatr has quit IRC | 10:42 | |
| *** hoonetorg has quit IRC | 10:43 | |
| *** hoonetorg has joined #openstack-ironic | 10:43 | |
| openstackgerrit | Merged openstack/ironic master: Set explicit default to enabled driver interfaces https://review.openstack.org/492877 | 10:50 |
| *** karimb has joined #openstack-ironic | 10:52 | |
| vdrok | morning milan dtantsur and sambetts lucasagomes ! | 10:52 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-ui stable/pike: Imported Translations from Zanata https://review.openstack.org/496200 | 10:54 |
| *** vmud213 has joined #openstack-ironic | 10:56 | |
| *** amarao1 has joined #openstack-ironic | 11:00 | |
| lucasagomes | o/ | 11:02 |
| nicodemos | morning | 11:11 |
| *** fellypefca has joined #openstack-ironic | 11:14 | |
| dtantsur | lucasagomes: Pixie died in the container, just imagine :) I had to resurrect it with my dark PTL powers :) | 11:16 |
| lucasagomes | dtantsur, oh noes | 11:18 |
| lucasagomes | dtantsur, network failure or ? | 11:18 |
| dtantsur | lucasagomes: I could not figure out, and could not afford spending too much time on it | 11:18 |
| lucasagomes | fair enough, yeah the code is very PoC-ish | 11:19 |
| lucasagomes | not sure it would handle reconnecting for example | 11:19 |
| *** jkilpatr has joined #openstack-ironic | 11:20 | |
| dtantsur | I suspect it did not reconnect indeed | 11:22 |
| dtantsur | our lab does go offline from time to time | 11:22 |
| dtantsur | anyway, docker restart works quite well :) | 11:22 |
| *** aparnav has quit IRC | 11:22 | |
| lucasagomes | cool :-) | 11:22 |
| *** baoli has joined #openstack-ironic | 11:29 | |
| *** bfournie has quit IRC | 11:30 | |
| *** yangyapeng has joined #openstack-ironic | 11:31 | |
| sambetts | dtantsur: I have a concern that zeroing out the flavor properties might break https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/iscsi_deploy.py#L88 | 11:34 |
| *** baoli has quit IRC | 11:34 | |
| *** yangyapeng has quit IRC | 11:35 | |
| dtantsur | sambetts: we don't zero flavor properties, we use special resource markers for that | 11:37 |
| dtantsur | zeroing flavor properties won't work with nova either | 11:37 |
| *** mat128 has joined #openstack-ironic | 11:37 | |
| sambetts | dtantsur: so you have to specify the CPU/DISK/RAM for a flavor anyway? ... I don't think we specify anywhere what those should be in a resource class world, we specfically talk about setting them to Zero using the custom resource classes | 11:39 |
| sambetts | dtantsur: which in "new flavor" terms is setting them to zero | 11:39 |
| openstackgerrit | Merged openstack/ironic master: Correct URLs in install docs https://review.openstack.org/496411 | 11:39 |
| sambetts | dtantsur: we should say something like, you need to set the flavor disk_gb to be the lowest amount offered by the hardware pool assigned the resource class you've given to the falvor | 11:40 |
| * sambetts hates resource classes so much... and wished he'd been involved earlier to prevent this... and go down the sensible path of just setting min and max amounts on the resource classes in the resource provider | 11:46 | |
| sambetts | literally we could have solved all of ironic's scheduling pains by just having the ironic virt driver report vcpus = 10 min = 10 max = 10 and the same for disk and ram to the placement API for each node and taken advantage of existing features of placement instead of inventing the worlds most hacky solution | 11:52 |
| *** led_ has quit IRC | 11:54 | |
| *** lets has quit IRC | 11:54 | |
| *** litao__ has quit IRC | 11:57 | |
| *** yangyapeng has joined #openstack-ironic | 11:57 | |
| *** lucasagomes is now known as lucas-hungry | 12:01 | |
| *** bfournie has joined #openstack-ironic | 12:02 | |
| jroll | tripleo doesn't do exact matching, fwiw | 12:04 |
| *** zhangfei has quit IRC | 12:04 | |
| *** dprince has joined #openstack-ironic | 12:04 | |
| jroll | this is the best we could do that people wouldn't rage about and block | 12:04 |
| jroll | sorry ¯\_(ツ)_/¯ | 12:04 |
| * Tengu enforce profile directly when importing nodes in openstack… | 12:05 | |
| Tengu | easier, faster, probably cleaner. I should also automatically add the right kernel stuff so that introspection would be useless. | 12:05 |
| sambetts | jroll: solution for that is that you put bounds on what ironic thinks a node has in-terms of resources, e.g. only tell ironic a node has 500gb of disk | 12:10 |
| sambetts | when it really has 1tb | 12:11 |
| jroll | from what I remember, that wouldn't work for the people not doing exact match | 12:11 |
| jroll | but I'm not really here to debate solutions | 12:11 |
| jroll | just to tell you we couldn't find any good alternatives in two years of debate | 12:12 |
| *** kencjohnston has quit IRC | 12:12 | |
| *** ildikov has quit IRC | 12:12 | |
| sambetts | jroll: yeah :/ its too late now... I managed to convince jaypipes at the summit that it was probably a neter solution just to use the mechanisums that exist today in the placement API but it was too late at that point | 12:13 |
| jroll | yeah | 12:13 |
| jroll | the hope was to work back towards that | 12:13 |
| * sambetts hadn't been involved enough to realise what was going on, or what the capabilites of the placement API was until jay gave an in-depth run down at the summit | 12:13 | |
| jroll | I also think those capabilities grew since we started down this path | 12:14 |
| sambetts | +10000 to working back towards using real world values that make sense to the people selecting the flavor | 12:14 |
| jroll | i.e. what we had when we started this didn't have enough to do the right thing | 12:14 |
| sambetts | yeah | 12:14 |
| *** kencjohnston has joined #openstack-ironic | 12:14 | |
| sambetts | got to love that bleeding edge | 12:14 |
| *** ildikov has joined #openstack-ironic | 12:14 | |
| jroll | bleeding edge? in openstack? :P | 12:15 |
| sambetts | ;) | 12:15 |
| *** karimb has quit IRC | 12:16 | |
| *** zhangfei has joined #openstack-ironic | 12:16 | |
| *** mat128 has quit IRC | 12:19 | |
| *** jcoufal has joined #openstack-ironic | 12:19 | |
| *** francoblanco has quit IRC | 12:22 | |
| *** francoblanco has joined #openstack-ironic | 12:22 | |
| *** jaypipes has joined #openstack-ironic | 12:23 | |
| *** pas-ha has quit IRC | 12:23 | |
| *** dutsmoc has quit IRC | 12:23 | |
| *** pas-ha has joined #openstack-ironic | 12:24 | |
| *** comstud has joined #openstack-ironic | 12:28 | |
| dtantsur | exact matching does not really work for installer case like tripleo.. | 12:36 |
| dtantsur | anyway, to your question sambetts: the CPU/DISK/RAM set in flavor is not going anywhere, until flavors themselves are going (which I think is the plan) | 12:36 |
| sambetts | dtantsur: we need to make it obvious in the docs that even though we're setting in all caps those fields to zero, that they actually do still play a part | 12:37 |
| sambetts | and should be set accordingly | 12:37 |
| dtantsur | also the DISK is a really severe problem, because bare metals have many disks | 12:37 |
| sambetts | +1 to the many disks thing | 12:37 |
| dtantsur | sambetts: ++ to make it clear in the docs (feel free to comment on my patch) | 12:38 |
| sambetts | dtantsur: already done :-P | 12:38 |
| milan | morning vdrok! :) | 12:38 |
| dtantsur | nice! /me was having coffee | 12:38 |
| sambetts | dtantsur: it was while I was reviewing that patch that caused me to think oh dear whats are we supposed to do now | 12:39 |
| dtantsur | no options are great, when you try to fit all varieties of hardware into just a handful of properties to schedule on | 12:39 |
| dtantsur | this is why we invented introspection rules initially (at least that's why I needed them): to be able to set capabilities automatically and schedule on them | 12:39 |
| sambetts | oh totally, thats why I love the placement API and resource class thing, I just don't think that having a single resource class for baremetal makes sense, | 12:40 |
| sambetts | 12:40 | |
| dtantsur | I can see why it's a problem.. an alternative would be to make scheduler aware of things like ironic that are always all-or-nothing | 12:41 |
| sambetts | dtantsur: it has it | 12:42 |
| dtantsur | really? I think nova folks told me otherwise, but I may have misunderstood | 12:42 |
| sambetts | dtantsur: in the placement API it has the concept of minimum_divisible_amounts and maximium_divisible_amount for every resource class | 12:42 |
| sambetts | dtantsur: so for a VM hypervisor you can have 128gb of ram and say mimomum_divisible_amount: 1gb | 12:43 |
| dtantsur | oh, sweet. then I wonder why we don't use them indeed | 12:43 |
| sambetts | dtantsur: which means that flavors asking for less than 1gb can't be put on the hypervisor | 12:43 |
| *** hw_wutianwei has quit IRC | 12:43 | |
| sambetts | and so for baremetal we just set minimum to the same as total | 12:43 |
| *** jlabarre has joined #openstack-ironic | 12:44 | |
| dtantsur | well, we don't need requests to be rejected, we need them rounded up | 12:44 |
| dtantsur | so e.g. I can ask for 512 MiB, but I'll still get all 128GiB | 12:44 |
| *** karimb has joined #openstack-ironic | 12:44 | |
| dtantsur | refusing to schedule in this case is essentially forcing exact filters, which are just as an ugly hack | 12:45 |
| sambetts | that could be an operator choice though | 12:45 |
| sambetts | we could have the minimum/max fed into placement for each node | 12:45 |
| sambetts | so for some nodes you might only want them matched by exact falvors | 12:45 |
| dtantsur | I'm not sure how it would look.. would it be something new on flavors to say that? on nodes? | 12:46 |
| sambetts | dtantsur: flavors would stay the same as whereever they are going | 12:46 |
| sambetts | this is all to do with populating the information in the placement API to make it aware of what a node is capable of | 12:47 |
| dtantsur | so how would you populate it? if you report min=max=properties.memory_mb, how will it work with smaller flavors? | 12:47 |
| *** udesale has joined #openstack-ironic | 12:48 | |
| *** rbudden has joined #openstack-ironic | 12:48 | |
| dtantsur | also keep in mind that reporting any local_gb is meaningless is many real cases | 12:48 |
| sambetts | dtantsur: initally I would have a flag, which is "allow_smaller_flavors_on_bigger_nodes=True/False" which enables or disable setting minimum to total | 12:49 |
| dtantsur | and if minimum is not set to total, will it be divisible then? | 12:49 |
| dtantsur | so e.g. if I take 4 GiB out of 16, will placement think that 12 are still left? | 12:50 |
| sambetts | ah yeah... that wouldn't workl | 12:50 |
| sambetts | but tbh my thoughts on reporting to placement is that if you want nodes to be treated the same, you just tell Ironic that they have the same amount of resources | 12:50 |
| sambetts | and then use exact matching all the time | 12:51 |
| dtantsur | this is not going to fly, really.. | 12:51 |
| dtantsur | e.g. it throws away inspection | 12:51 |
| sambetts | nope it doesn'#t | 12:52 |
| sambetts | I already have a bunch of introspection rules that equal out all my nodes | 12:52 |
| sambetts | because they are all very slightly different | 12:52 |
| dtantsur | well, this is a gross hack :) | 12:52 |
| dtantsur | and keep in mind OOB inspection too | 12:52 |
| sambetts | so I have rules that say if disk_gb > 500gb then disk_gb == 500gb | 12:52 |
| dtantsur | we already give it some hard time with resource classes | 12:52 |
| dtantsur | with your rule, the nodes is no longer deployable by flavor requiring 600gb | 12:53 |
| sambetts | I don't create a flavor with 600gb | 12:53 |
| dtantsur | well, you don't, but people may :) | 12:53 |
| sambetts | the introspection rules that I create are designed to match my flavors | 12:53 |
| sambetts | so if I wanted a flavor that had 600gb then I would create another rule which is if disk_gb > 600gb then disk_gb == 600gb | 12:54 |
| sambetts | so then nodes with between 600 and 500 are equaled out to 500gb, and I can have a flavor that asks for 500 | 12:54 |
| sambetts | and then nodes greater than 600 and equaled out to 600 and match the flavor that asks for 600 | 12:55 |
| dtantsur | now, why is it better than having 2 resource classes? :) you're essentially reinventing them | 12:55 |
| sambetts | -1 resource classes have zero ties to physical capabilites, I'm creating bounds to group my nodes based on real values that they are capable of | 12:56 |
| dtantsur | yep, you're creating resource classes implicitly based on disk size (of which disk, by the way?) | 12:56 |
| sambetts | you might have some nodes that have 600gb of disk but also no ram | 12:56 |
| sambetts | almost* | 12:57 |
| sambetts | it all evens itself out and there are no hacks needed from a operators's perspective with custom flavors and setting arbitrty values on nodes | 12:58 |
| sambetts | I like having flavors that can map to a VM or Baremetal too | 12:58 |
| dtantsur | yeah, this is a big downside that we cannot do it | 12:58 |
| dtantsur | I'm not sure how many people require such flavors, to be honest, but it's sad it's no longer possible | 12:59 |
| *** cdearborn has joined #openstack-ironic | 12:59 | |
| *** sasha2 has joined #openstack-ironic | 13:01 | |
| *** rloo has joined #openstack-ironic | 13:02 | |
| *** lucas-hungry is now known as lucasagomes | 13:03 | |
| TheJulia | Good morning everyone | 13:03 |
| nicodemos | morning TheJulia | 13:04 |
| dtantsur | TheJulia: morning! https://review.openstack.org/#/c/495316/ is the biggest blocker for the release, mind checking it? | 13:05 |
| patchbot | patch 495316 - ironic (stable/pike) - [reno] Add prelude for Pike (9.1.0) release | 13:05 |
| TheJulia | Wow, lots of discussion today | 13:05 |
| sambetts | o/ TheJulia | 13:05 |
| sambetts | mostly me ranking | 13:05 |
| sambetts | ranting* | 13:05 |
| dtantsur | resource classes flame war, not the first one and certainly not the last... | 13:05 |
| sambetts | ;) | 13:05 |
| * dtantsur is not looking forward to more people figuring it out | 13:05 | |
| sambetts | just got it in me to pick holes in everything today... /me feels slightly sorry for the patches I'm reviewing | 13:06 |
| *** rpioso has joined #openstack-ironic | 13:06 | |
| rloo | hi and good morning sambetts, dtantsur, TheJulia. It is worth reading about sambetts's ranting? | 13:07 |
| rpioso | Good morning! | 13:08 |
| rloo | sambetts: fwiw, it is good to pick holes, if that helps things get better!' | 13:08 |
| rloo | morning rpioso | 13:08 |
| rpioso | rloo: :) | 13:08 |
| sambetts | rloo: perhaps if you aren't aware of some of the capabilites of the placement API | 13:08 |
| dtantsur | morning rloo | 13:08 |
| TheJulia | dtantsur: +2+A? | 13:08 |
| TheJulia | err | 13:09 |
| TheJulia | no ?, I just +A'ed it | 13:09 |
| dtantsur | good ;) | 13:09 |
| * TheJulia needs coffee | 13:09 | |
| dtantsur | rloo: would you like to wait for all the docs patches to land in stable/pike? I'd like to cut the release once the reno merges | 13:09 |
| * TheJulia resumes brain on BFV thoughts while making some coffee | 13:09 | |
| dtantsur | the docs are always backportable, we may just get a point release in the beginning of september | 13:09 |
| milan | morning nicodemos rloo rpioso TheJulia! :) | 13:09 |
| *** karimb has quit IRC | 13:10 | |
| milan | sambetts, https://review.openstack.org/#/c/490008/5/api-ref/source/baremetal-api-v1-drivers.inc I wanted to check out how it renders | 13:10 |
| patchbot | patch 490008 - ironic - Switch API ref to use versionadded syntax throughout | 13:10 |
| *** fellypefca has quit IRC | 13:10 | |
| *** fellypefca has joined #openstack-ironic | 13:10 | |
| milan | sambetts, but the links led to the current/latest docs | 13:10 |
| * milan might not know what he's doing though | 13:10 | |
| sambetts | milan: you need to click on the api-ref job | 13:11 |
| sambetts | milan: they are rendered separately from the main docs | 13:11 |
| milan | aaaah | 13:11 |
| milan | sambetts, thx | 13:11 |
| *** karimb has joined #openstack-ironic | 13:11 | |
| * milan wonders why the green text | 13:12 | |
| rpioso | milan: Morning :) | 13:13 |
| rloo | dtantsur: i'd like the patches that fix urls, to land if possible. | 13:13 |
| rloo | dtantsur: what is the status of those? let me see where we're at, just catching up ... | 13:13 |
| dtantsur | mm, okay. this probably pushes our release to tomorrow | 13:13 |
| fellypefca | Good morning, everyone. | 13:13 |
| rloo | dtantsur: oh? what's the 'deadline' for today? | 13:14 |
| rloo | dtantsur: i mean, time. or do you think cuz of how long it'll take to approve etc? | 13:14 |
| dtantsur | rloo: I just cannot imagine merging all this today :) | 13:14 |
| dtantsur | yeah, and there are only a couple of work hours left for me | 13:14 |
| rloo | dtantsur: doc patches should be fast... | 13:14 |
| rloo | dtantsur: so i'd like the configuration doc patch too but that might take more time, i see you had comments but i haven't looked into it | 13:15 |
| *** vmud213 has quit IRC | 13:15 | |
| rloo | dtantsur: could you or TheJulia +A this: https://review.openstack.org/#/c/496354/ | 13:16 |
| patchbot | patch 496354 - ironic (stable/pike) - Fix incorrect documentation urls | 13:16 |
| dtantsur | rloo: out of curiosity, why do you want them in pike right now? because of people packaging docs? | 13:16 |
| rloo | dtantsur: cuz i have spent over a day dedicated to doc changes, which is what i said i'd do before. i have other things i'd like to do so don't plan on continuing after this. | 13:16 |
| rloo | dtantsur: and cuz, if we are going to do it, we might as well do it now | 13:17 |
| TheJulia | dtantsur: do people even package the docs? | 13:17 |
| rloo | dtantsur: cuz i am familiar with the patches. and i'm not going to be later. | 13:17 |
| dtantsur | TheJulia: RDO does | 13:17 |
| rloo | dtantsur: that's all. | 13:17 |
| *** osdev has joined #openstack-ironic | 13:17 | |
| TheJulia | dtantsur: might be it as far as I'm aware. | 13:17 |
| dtantsur | I don't suggest to stop working on docs, I only suggest to cut the release at some point today | 13:18 |
| dtantsur | because it's a non-standard release, and I don't want to propose it hours before the final and utter deadline | 13:18 |
| rloo | dtantsur: right, and I'm only answering your questions :) | 13:18 |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: Correct URLs in install docs https://review.openstack.org/496705 | 13:19 |
| dtantsur | yeah, but my question was mostly about wanting them in pike *right now, before 9.1.0*, not about working on them at all or whether to backport them at all | 13:19 |
| rloo | dtantsur: ok, so let me answer it this way, there isn't anything urgent/blocking landing 9.1.0 as long as the prelude is in. | 13:20 |
| rloo | dtantsur: the rest is to make it better. | 13:20 |
| rloo | dtantsur: and i don't plan on spending much more time on the doc patches. so personally, i just want to get them out of the way. | 13:21 |
| dtantsur | rloo: got it! let me request the release today, and let's finish the doc patches today-tomorrow | 13:21 |
| rloo | dtantsur: and if we say 'we can still review and update docs', are we really going to do it soon as opposed to ... whenever... | 13:21 |
| *** mjturek has joined #openstack-ironic | 13:22 | |
| dtantsur | at least let's finish what we already have proposed (link updates mostly) | 13:22 |
| dtantsur | I don't suggest to keep your patches up for months ;) | 13:22 |
| rloo | dtantsur: no one suggests that but we do have patches that have been around for months :-(. | 13:22 |
| rloo | dtantsur: i actually started to look at patches, saw some more than 2 years old. sigh. | 13:23 |
| *** makowals_ has joined #openstack-ironic | 13:23 | |
| dtantsur | yeah, this happens | 13:23 |
| rloo | dtantsur: anyway, that's another story :) | 13:23 |
| * dtantsur remember his nova patch merging after more than a year of inactivity | 13:23 | |
| *** yangyapeng has quit IRC | 13:24 | |
| rloo | so these two doc (on master) patches have a +2 already: https://review.openstack.org/#/c/496320/ & https://review.openstack.org/#/c/496361/ | 13:24 |
| patchbot | patch 496320 - ironic - correct URLs in admin docs | 13:24 |
| patchbot | patch 496361 - ironic - correct URLs in contributor docs & main index | 13:24 |
| rloo | they should be easy to review/land and backport to pike | 13:24 |
| *** makowals has quit IRC | 13:25 | |
| rloo | then almost all of the links will be fixed. (the remainder are in the other doc patches that will land later I hope) | 13:25 |
| dtantsur | okay, let's finish there two as well | 13:25 |
| rloo | the doc migration broke more links than people may realize. I am concerned about that for docs out there (people's notes, whatever). | 13:26 |
| *** bnemec has joined #openstack-ironic | 13:26 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: Change perms for the VM logs, fix log publish https://review.openstack.org/494549 | 13:27 |
| dtantsur | rloo: yeah, I"m still fixing tripleo-docs | 13:29 |
| *** igordc has joined #openstack-ironic | 13:29 | |
| rloo | dtantsur: :-( | 13:29 |
| dtantsur | one approved, 2nd to go | 13:29 |
| TheJulia | rloo: you could have left the // ;) | 13:32 |
| rloo | TheJulia: Nope, that is wrong! | 13:33 |
| rloo | TheJulia: i can't help it, I like things to be juuuuust right :) | 13:33 |
| dtantsur | folks, I completely forgot about a meetup on the PTG.. does anyone know the Denver area enough to find some place for a dinner/beers? | 13:35 |
| TheJulia | sadly no | 13:35 |
| TheJulia | I will, however, be there a week before the PTG | 13:35 |
| TheJulia | so I can also go do some exploring | 13:35 |
| TheJulia | but seriously, if some knows of anything, it my gas tank would appreciate not driving and searching around downtown | 13:37 |
| dtantsur | TheJulia: yes please! it's somewhat harder to figure out things from over the ocean :) | 13:37 |
| dtantsur | rloo: you can already propose backports for your patches, just mind the cherry-pick line | 13:38 |
| *** links has quit IRC | 13:38 | |
| TheJulia | Is the linux-iscsi.org web page loading for people ? | 13:39 |
| rloo | dtantsur: right. and cherry-pick line is working now (with gerrit 'cherry-pick' button). i noticed it working yesterday. | 13:39 |
| dtantsur | TheJulia: "waiting.." for me | 13:39 |
| sambetts | TheJulia, dtantsur: if we managed to organise anything can we try for it to be a bit closer to the PTG than the meet up in Atlanta? | 13:39 |
| openstackgerrit | Ilya Etingof proposed openstack/ironic master: Configuration documentation migrated https://review.openstack.org/491861 | 13:39 |
| rloo | dtantsur: wrt meetup, we should come up with an evening too | 13:39 |
| dtantsur | sambetts: you mean, to the venue? | 13:40 |
| dtantsur | rloo: you mean, the time slot? I was about to create a doodle | 13:40 |
| rloo | dtantsur: yeah, that, and thx! | 13:40 |
| sambetts | dtantsur: yeah, in Atlanta the meet up was like 20min drive away or more if I remember correctly | 13:40 |
| dtantsur | I'll throw it on the ML, maybe somebody will tell us a good place | 13:40 |
| rloo | sambetts: you mean you didn't like walking or taking a cab to the place? :) | 13:40 |
| rloo | TheJulia: not working for me. that link i mean. | 13:40 |
| rloo | TheJulia: still spinning... | 13:40 |
| *** sauloaislan has joined #openstack-ironic | 13:41 | |
| sambetts | I don't mind walking, but in Atlanta it was a little too far to walk when I looked it up on the map | 13:41 |
| TheJulia | sambetts: I'll try for walking distance. It will be good for me to sample many burgers and beers | 13:41 |
| sauloaislan | Morning! | 13:41 |
| sambetts | TheJulia: thanks | 13:41 |
| sambetts | :) | 13:41 |
| *** jaypipes has quit IRC | 13:42 | |
| dtantsur | the poll is https://doodle.com/poll/nvavg9ab9ebq2e4v | 13:42 |
| sauloaislan | dtantsur: Hi. Can you take a look at this backport? https://review.openstack.org/#/c/489598/ | 13:43 |
| patchbot | patch 489598 - ironic (stable/ocata) - Removes agent mixin from oneview drivers | 13:43 |
| dtantsur | will do, not sure about today (release time!) | 13:43 |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: Documentation for 'snmp' hardware type https://review.openstack.org/496710 | 13:44 |
| dtantsur | also FYI team photos https://docs.google.com/spreadsheets/d/18DmI9ydq6dKr4hAcoAN83vYXWn8TV1NEvy09Ll02_Y0/edit#gid=0 | 13:47 |
| rloo | dtantsur: is the schedule out for PTG, we should do the team photo during one of our breaks, like last time | 13:48 |
| rloo | dtantsur: oh, it is just tues and wed. so wed for sure | 13:49 |
| sauloaislan | dtantsur: Ok thank you ! | 13:49 |
| dtantsur | rloo: the last time we just had a break when we needed | 13:50 |
| dtantsur | it's like 5 mins | 13:50 |
| rloo | dtantsur: i was trying to remember how it worked. i think we start at around 9am? | 13:51 |
| openstackgerrit | Merged openstack/ironic master: fix a typo in agent.py: s/doman/domain/ https://review.openstack.org/496574 | 13:51 |
| dtantsur | yep. we had a slot booked, so we just paused, went out for photo, and then went back to discussing stuff | 13:51 |
| openstackgerrit | Merged openstack/ironic stable/pike: [reno] Add prelude for Pike (9.1.0) release https://review.openstack.org/495316 | 13:51 |
| openstackgerrit | Merged openstack/ironic stable/pike: Fix incorrect documentation urls https://review.openstack.org/496354 | 13:51 |
| openstackgerrit | Merged openstack/ironic master: correct URLs in admin docs https://review.openstack.org/496320 | 13:51 |
| openstackgerrit | Merged openstack/ironic master: correct URLs in contributor docs & main index https://review.openstack.org/496361 | 13:52 |
| rloo | dtantsur: ok, i think any time is good then. even 10 would be a good time for a break. i just wouldn't pick 10 though cuz if they are late setting up or there are issues, it might take longer if we're the first team. | 13:52 |
| *** mat128 has joined #openstack-ironic | 13:52 | |
| dtantsur | I'd pick 11 or something like that | 13:52 |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: correct URLs in admin docs https://review.openstack.org/496712 | 13:53 |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: correct URLs in contributor docs & main index https://review.openstack.org/496713 | 13:53 |
| *** mat128 has quit IRC | 13:53 | |
| *** zhangfei has quit IRC | 13:54 | |
| rloo | dtantsur: that would work, thx. | 13:55 |
| *** ijw has joined #openstack-ironic | 13:55 | |
| dtantsur | anybody else has an opinion on photo time? I'd go ahead and ask for 11 otherwise | 13:56 |
| *** Goneri has joined #openstack-ironic | 13:57 | |
| openstackgerrit | Merged openstack/ironic master: Documentation for irmc hardware type https://review.openstack.org/490483 | 13:57 |
| *** mat128 has joined #openstack-ironic | 13:57 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: Documentation for irmc hardware type https://review.openstack.org/496716 | 13:59 |
| milan | sambetts, sry, the green text is killing my eyes :P | 14:01 |
| rloo | dtantsur: all the doc patches that i wanted for pike, have been approved in master branch. except yours: https://review.openstack.org/#/c/496309/ | 14:01 |
| patchbot | patch 496309 - ironic - Minor improvements to the resource classes documen... | 14:01 |
| *** mjura has quit IRC | 14:01 | |
| *** mtanino has joined #openstack-ironic | 14:02 | |
| rloo | dtantsur: i didn't look to see the comments, is that something you want to try to get in? | 14:02 |
| dtantsur | rloo: it's not THAT urgent IMO. mostly clarifications. | 14:03 |
| openstackgerrit | Merged openstack/ironic stable/pike: Correct URLs in install docs https://review.openstack.org/496705 | 14:03 |
| rloo | dtantsur: ok then. i'll ping you & TheJulia when the stable/pike patches are ready to approve. | 14:04 |
| dtantsur | rloo: do we have backports proposed from everythin we want in Pike right now? | 14:04 |
| dtantsur | aha, good :) | 14:04 |
| rloo | dtantsur: yup | 14:04 |
| rloo | dtantsur, TheJulia: after those pike patches land, TheJulia will need to update https://review.openstack.org/#/c/495315/ | 14:04 |
| patchbot | patch 495315 - ironic (stable/pike) - Change doc links for pike branch | 14:04 |
| rloo | dtantsur, TheJulia: this is ready to land in pike: https://review.openstack.org/#/c/496712/ | 14:05 |
| patchbot | patch 496712 - ironic (stable/pike) - correct URLs in admin docs | 14:05 |
| openstackgerrit | Sam Betts proposed openstack/ironic master: Add documentation describing each Ironic state https://review.openstack.org/496719 | 14:05 |
| sambetts | rloo: ^ | 14:05 |
| rloo | sambetts: what? trying to get something in under the wire? | 14:05 |
| *** hamzy has quit IRC | 14:05 | |
| sambetts | rloo: had a "get it done" moment when I saw you discussion with dtantsur about the link to the spec | 14:06 |
| dtantsur | MOAR DOCS MOAAARRRR | 14:06 |
| dtantsur | this is a good thing, but not blocking the release | 14:06 |
| sambetts | dtantsur: nope, I don't expect it to be backported | 14:06 |
| *** zhangfei has joined #openstack-ironic | 14:07 | |
| rloo | sambetts: thx! but yeah. I'll look at it today though :) | 14:07 |
| sambetts | just wanted to get it up there and done so that we don't forget about it again | 14:07 |
| masber | good afternoon, which interface is Ironic using for PXE boot? | 14:07 |
| sambetts | milan: thats part of the openstack doc's theme unfortunately | 14:07 |
| rloo | dtantsur, TheJulia: this is also ready to approve in pike: https://review.openstack.org/#/c/496713/ | 14:07 |
| patchbot | patch 496713 - ironic (stable/pike) - correct URLs in contributor docs & main index | 14:07 |
| *** udesale has quit IRC | 14:08 | |
| dtantsur | masber: you mean, which NIC? whatever is configured in BIOS, we don't change it | 14:08 |
| sambetts | masber: it'll attempt to use all NICs you've told Ironic about that don't have pxe_enabled=False | 14:08 |
| milan | sambetts, maybe the "note" shouldn't bee that long at least, or use the optional second arg <a brief description> | 14:08 |
| milan | sambetts, or wdyt about making it a paragraph? | 14:09 |
| dtantsur | rloo: done. now only TheJulia's patch left? | 14:09 |
| * TheJulia has too many windows open :( | 14:09 | |
| rloo | dtantsur: yup. although i wanted to ask you, did you want to also get this one in: https://review.openstack.org/#/c/491861/ | 14:09 |
| patchbot | patch 491861 - ironic - Configuration documentation migrated | 14:09 |
| TheJulia | and a kitten driving me insane | 14:10 |
| sambetts | milan: yeah thats probably a good idea... | 14:10 |
| rloo | TheJulia: don't worry about it, things are in control. I'll ping you when it is time to update your patch. Things are working their way to being merged. | 14:10 |
| *** yangyapeng has joined #openstack-ironic | 14:10 | |
| openstackgerrit | Merged openstack/ironic master: Configuration documentation migrated https://review.openstack.org/491861 | 14:11 |
| sambetts | milan: tbh that one might actually be worth changing to a .. versionchanged:: | 14:11 |
| * TheJulia goes back to hunting something down in linux-iscsi docs | 14:11 | |
| milan | sambetts, yup | 14:11 |
| openstackgerrit | Ruby Loo proposed openstack/ironic stable/pike: Configuration documentation migrated https://review.openstack.org/496720 | 14:12 |
| *** zshi_laptop has quit IRC | 14:12 | |
| milan | sambetts, hopefully that doesn't render violet ;) | 14:12 |
| sambetts | milan: orange I think | 14:12 |
| *** shoaib has joined #openstack-ironic | 14:12 | |
| milan | omg | 14:12 |
| milan | :D | 14:12 |
| openstackgerrit | Merged openstack/ironic stable/pike: correct URLs in admin docs https://review.openstack.org/496712 | 14:13 |
| rloo | dtantsur: sorry, i forgot, another one ready in pike: https://review.openstack.org/#/c/496716/ | 14:14 |
| patchbot | patch 496716 - ironic (stable/pike) - Documentation for irmc hardware type | 14:14 |
| masber | dtantsur, sorry I didn't explain myself properly, I meant which Openstack interface? external, management or guest? or any of them? | 14:14 |
| sambetts | neutron port? | 14:14 |
| masber | yes | 14:15 |
| *** udesale has joined #openstack-ironic | 14:15 | |
| dtantsur | rloo: do you want https://review.openstack.org/#/c/491861/ there too? | 14:16 |
| patchbot | patch 491861 - ironic - Configuration documentation migrated (MERGED) | 14:16 |
| openstackgerrit | Merged openstack/ironic stable/pike: correct URLs in contributor docs & main index https://review.openstack.org/496713 | 14:16 |
| dtantsur | this is the last chance to get something in :) | 14:16 |
| sambetts | that depends how you've deployed Ironic, if you are using ironic with dynamic networking then it'll pxe boot for deployment in the provisioning network defined in your ironic.conf, if you are using ironic with flat networking, then it'll try to use whatever the network interface that you pass into nova | 14:17 |
| rloo | dtantsur: already cherry-picked, waiting for jenkins | 14:17 |
| rloo | dtantsur: https://review.openstack.org/#/c/496720/ :) | 14:17 |
| patchbot | patch 496720 - ironic (stable/pike) - Configuration documentation migrated | 14:17 |
| shoaib | Hi guys, I am very new to ironic, and desperately need some assistance. I installed openstack using packstack all-in-one RDO on a Hardware node. Enrolled an Ironic node, but wwhen the node comes up, It doesn't get any DHCP response from the Controller node running Ironic conductor, Neutron, Nova-compute etc https://thepasteb.in/p/j2hlQ3m5nxOcP | 14:17 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: DNM: Change perms for the VM logs, fix log publish https://review.openstack.org/494549 | 14:18 |
| dtantsur | sambetts: maybe you know ^^^ I was not able to quickly figure out yesterday | 14:18 |
| shoaib | Ironic port-create with MAC ID, does not get any ""vif_port_id" automatically assigned to it, and that port is also not in the Neutron port-list... Any help would be greatly appreciated :) https://thepasteb.in/p/j2hlQ3m5nxOcP | 14:19 |
| rloo | dtantsur: ugh, i just saw this, haven't looked at it yet: https://review.openstack.org/#/c/496480/ | 14:19 |
| patchbot | patch 496480 - ironic - Update documentation for ``ilo`` hardware type | 14:19 |
| rloo | dtantsur: forget it, that patch updates a release note and i think it is too late for that right? cuz we cut Pike? (We have to document this, i keep forgetting...) | 14:20 |
| *** rbartal has quit IRC | 14:21 | |
| sambetts | shoaib: are you using nova to deploy the node? | 14:21 |
| dtantsur | rloo: yeah, we probably should not update release notes, they don't replace documentation after all | 14:22 |
| dtantsur | I mean, after the release | 14:22 |
| rloo | dtantsur: if i recall, if you update after the branch is cut, it will put that reno in the master/current releases instead. | 14:22 |
| shoaib | sambetts: I tried using Nova boot but that failed saying No valid host found, Doesn't deployment comes after a Node PXE boots ? | 14:22 |
| shoaib | sambetts: Does ironic-node create also fill that port in the neutron list ? | 14:23 |
| openstackgerrit | Merged openstack/ironic stable/pike: Documentation for irmc hardware type https://review.openstack.org/496716 | 14:23 |
| sambetts | shoaib: no, when you issue nova boot --nic net-id=<neutron network>, nova creates a neutron port for you, then schedules the instance onto the ironic node, then attaching the neutron port to the ironic node, then triggers Ironic to deploy the node | 14:24 |
| dtantsur | rloo: yeah, likely | 14:25 |
| sambetts | shoaib: when using ironic with nova, you shouldn't intract with ironic directly to deploy nodes | 14:25 |
| sambetts | shoaib: if you are getting No Valid Host Found that indicates to me that there are bigger issues with your environment than neutron ports not being created and attached | 14:25 |
| sambetts | shoaib: you should check your node.properties are set up correctly and your nova flavors are created correctly to match your node's properties | 14:26 |
| *** makowals_ has quit IRC | 14:26 | |
| sambetts | shoaib: and make sure your ironic node is in avaiable state and not in maintenance | 14:26 |
| *** Nisha_Agarwal has joined #openstack-ironic | 14:27 | |
| sambetts | if all those things are correct, you should see your resources appear in nova hypervisors-stats | 14:27 |
| sambetts | and when you do nova boot it should work | 14:27 |
| shoaib | sambetts: My node shows Available and Power on, https://thepasteb.in/p/P1hvEoX1Py4il | 14:28 |
| shoaib | sambetts: Nova should be used when its not a Standalone Bare Metal node right ? I'm trying to get the concept, because openstack Wiki shows several ways of deploying it | 14:29 |
| shoaib | sambetts: Also I am able to power on/off/ reboot my Node using IPMI with ironic commands | 14:31 |
| *** udesale has quit IRC | 14:31 | |
| *** udesale__ has joined #openstack-ironic | 14:31 | |
| *** baoli has joined #openstack-ironic | 14:32 | |
| *** JayF has joined #openstack-ironic | 14:32 | |
| *** JayF has joined #openstack-ironic | 14:32 | |
| *** baoli has quit IRC | 14:33 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: DNM: Change perms for the VM logs, fix log publish https://review.openstack.org/494549 | 14:34 |
| * jlvillal hopes to get his development workstation back up and running today after an SSD drive failure yesterday | 14:35 | |
| sambetts | shoaib: yes, nova should be used when Ironic is in non-standalone mode | 14:36 |
| jlvillal | sambetts, shoaib: Or possibly using Mogan. https://wiki.openstack.org/wiki/Mogan Though that is not used very much, but it is another option out there. | 14:38 |
| openstackgerrit | Merged openstack/ironic stable/pike: Configuration documentation migrated https://review.openstack.org/496720 | 14:39 |
| *** trown is now known as trown|brb | 14:41 | |
| rloo | dtantsur: all the stablel/pike patches have merged, except: https://review.openstack.org/#/c/496710/. | 14:41 |
| patchbot | patch 496710 - ironic (stable/pike) - Documentation for 'snmp' hardware type | 14:41 |
| rloo | dtantsur: i just looked and it makes a change to drivers/modules/snmp.py so it won't be fast | 14:41 |
| rloo | dtantsur: so i think we shouldn't wait for it | 14:42 |
| rloo | dtantsur: unless you want to? | 14:42 |
| shoaib | sambetts: Btw I'm not using ironic-inspector, thats not necessary right ? And I also am now so confused about the sequence in which to boot up the system .. I did it in this order, configured neutron first, then created ironic node, updated it with deploy images, followed by ironic port creation, and then Nova boot ? | 14:42 |
| openstackgerrit | Sam Betts proposed openstack/ironic master: Switch API ref to use versionadded syntax throughout https://review.openstack.org/490008 | 14:42 |
| dtantsur | rloo: let's release without it, I guess | 14:43 |
| rloo | TheJulia: we're ready to get an updated https://review.openstack.org/#/c/495315/ | 14:43 |
| patchbot | patch 495315 - ironic (stable/pike) - Change doc links for pike branch | 14:43 |
| *** hamzy has joined #openstack-ironic | 14:43 | |
| shoaib | Should a Node be Booted with Nova before or after Ironic node-create ? Thanks..I've been using these instructions here https://docs.openstack.org/ironic/latest/install/index.html , but they are not so easy to follow | 14:44 |
| sambetts | after | 14:44 |
| sambetts | shoaib: inspector is optional, when you created the ironic node did you specify all its properties, cpu, local_gb and memory_mb ? | 14:44 |
| *** Nisha_Agarwal has quit IRC | 14:44 | |
| sambetts | shoaib: also which version of OpenStack are you using? | 14:44 |
| rloo | dtantsur: yup. we can always backport later :) | 14:45 |
| TheJulia | rloo: okay, finishing thoughts into an ether pad, and I'll go update it | 14:45 |
| rloo | TheJulia: thx. I suspect most of your changes in that patch will be removed cuz I changed a lot of the links to be relative/internal ones. | 14:45 |
| * dtantsur wastes a lot of corporate paper and toner on his relocation paperwork in the meantime :) | 14:49 | |
| *** trown|brb is now known as trown | 14:50 | |
| masber | sambets, I am using kolla-ansible for deployment, does it means that the neutron port is neutron_external_interface? | 14:50 |
| sambetts | neutron_external_interface is the neutron public API IP interface | 14:51 |
| shoaib | sambetts: I'm using Ocata | 14:51 |
| sambetts | shoaib: ok, did you set the properties, as specified in enrollment? | 14:52 |
| *** lihi has quit IRC | 14:52 | |
| *** oanson has quit IRC | 14:53 | |
| shoaib | sambetts: Yes i specified the properties, here is the output of my ironic node-show https://thepasteb.in/p/k5hYzrBJpJJfE | 14:53 |
| *** leyal has quit IRC | 14:53 | |
| *** lihi has joined #openstack-ironic | 14:54 | |
| *** oanson has joined #openstack-ironic | 14:54 | |
| shoaib | sambetts: Maybe im using thr wrong Nova boot command, I used this "nova boot --flavor my-baremetal-flavor --image my-image --key-name admin metal" | 14:54 |
| sambetts | shoaib: does your baremetal flavor exactly match your node's properties?? | 14:54 |
| *** Nisha_Agarwal has joined #openstack-ironic | 14:54 | |
| *** leyal has joined #openstack-ironic | 14:55 | |
| TheJulia | wheeee lots of merge conflicts | 14:56 |
| shoaib | sambetts: No actually, my flavor has "memory_mb=1024, Disk=100, " different from node properties | 14:56 |
| shoaib | sambetts: My flavor VCPU=2, whereas for the node, its set to 48, Do all these 3 parameters need to match ? | 14:57 |
| sambetts | shoaib: your flavor needs to exactly match your nodes properties to schedulae correctly | 14:58 |
| *** livelace has quit IRC | 14:58 | |
| rloo | TheJulia: sorry. I was wondering if it might be easier if you started from scratch, using your script again | 14:59 |
| TheJulia | rloo: I'd have to write it again since I did it as a one-liner | 14:59 |
| * TheJulia just fixes | 14:59 | |
| dtantsur | find doc/source -name \*.rst -exec sed -i s/latest/pike/g \{\} \; :) | 15:00 |
| * dtantsur did not check it | 15:00 | |
| TheJulia | hehe | 15:01 |
| TheJulia | Catches some other links though | 15:01 |
| *** crushil has joined #openstack-ironic | 15:01 | |
| TheJulia | so the search/replace needs to be much more specific. | 15:01 |
| shoaib | sambetts: Ok, I'll modify the node/flavor properties ... Also with "nova boot --nic net-id=<neutron network>", Is the neutron network the output of my "neutron net-list" or "neutron subnet-list " ? | 15:01 |
| sambetts | net-list | 15:02 |
| *** Nisha_Agarwal has quit IRC | 15:05 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic stable/pike: Change doc links for pike branch https://review.openstack.org/495315 | 15:05 |
| *** zhangfei has quit IRC | 15:05 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: DNM: Change perms for the VM logs, fix log publish https://review.openstack.org/494549 | 15:09 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: Change perms for the VM logs https://review.openstack.org/494549 | 15:09 |
| *** makowals has joined #openstack-ironic | 15:12 | |
| *** srobert has joined #openstack-ironic | 15:13 | |
| dtantsur | TheJulia, rloo is right, DIB does not branch | 15:14 |
| TheJulia | Yeah, they don't | 15:14 |
| TheJulia | did I not fix that? | 15:14 |
| rloo | TheJulia: do you think it is possible to get whoever/infra? to put a link from https://docs.openstack.org/diskimage-builder/pike/ to ..latest | 15:14 |
| rloo | TheJulia: you did, but I think I missed some from previous patch set. not sure. | 15:15 |
| TheJulia | funky.... | 15:15 |
| * dtantsur is on a meeting | 15:15 | |
| TheJulia | rloo: I'll just edit the text since they don't branch | 15:15 |
| *** udesale__ has quit IRC | 15:15 | |
| rloo | TheJulia: my other thought is for a link from https://docs.openstack.org/diskimage-builder to go to ../latest. then we can delete 'latest' from the urls. | 15:16 |
| rloo | TheJulia: yeah, edit the patch works for me! | 15:16 |
| openstackgerrit | Julia Kreger proposed openstack/ironic stable/pike: Change doc links for pike branch https://review.openstack.org/495315 | 15:19 |
| TheJulia | they might branch one day, who knows | 15:20 |
| *** makowals has quit IRC | 15:20 | |
| NobodyCam | Good Morning Ironic'ers | 15:21 |
| shoaib | sambetts: I have another question, how does Ironic or Nova gets to know about the tftp Boot .. Is it through the TFTP boot server port 69, which is opened by the server ? Thanks | 15:23 |
| rloo | TheJulia: true... | 15:24 |
| dtantsur | morning NobodyCam | 15:24 |
| NobodyCam | hey hey Morning dtantsur :) | 15:24 |
| rloo | morning NobodyCam! | 15:24 |
| NobodyCam | morning rloo :) | 15:24 |
| shoaib | When using Ironic with Nova, do i need to set "dhcp_provider = neutron" in the ironic.conf file ? Thanks | 15:36 |
| *** yufei has joined #openstack-ironic | 15:38 | |
| dtantsur | TheJulia, rloo, do we want to fix https://review.openstack.org/#/c/495315/4/doc/source/admin/boot-from-volume.rst@19 ? this should be an internal link... | 15:40 |
| patchbot | patch 495315 - ironic (stable/pike) - Change doc links for pike branch | 15:40 |
| rloo | dtantsur: i don't think it is urgent for pike. i will fix on master though. | 15:40 |
| dtantsur | okie | 15:40 |
| rloo | dtantsur: the link works, just suboptimal :) | 15:40 |
| dtantsur | okay, now it travels through the gate, thanks rloo and TheJulia | 15:42 |
| rloo | dtantsur: yay! | 15:43 |
| *** yufei has quit IRC | 15:43 | |
| openstackgerrit | Rushil Chugh proposed openstack/ironic-specs master: Add spec for XClarity Driver https://review.openstack.org/476922 | 15:44 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: DNM: Change perms for the VM logs https://review.openstack.org/494549 | 15:45 |
| * dtantsur likes how https://docs.openstack.org/ironic/latest/ looks now! | 15:46 | |
| *** bnemec has quit IRC | 15:46 | |
| rloo | dtantsur: are we ready to bike shed? I think user guide should be sooner. | 15:46 |
| *** bnemec has joined #openstack-ironic | 15:47 | |
| rloo | dtantsur: but maybe we shouldn't discuss now. i really would like some openstack guidelines/consistency wrt that landing page. | 15:47 |
| dtantsur | yeah.. I'm fine as long as it can be found. | 15:47 |
| shoaib | Hey guys, can anyone kindly answer this for me ? When using Ironic with Nova, do i need to set "dhcp_provider = neutron" in the ironic.conf file ? Thank You very much | 15:47 |
| dtantsur | shoaib: isn't it the default? I think this is the correct value, yes | 15:47 |
| rloo | dtantsur: guidelines AND change the highlighting from red to something not so bad. | 15:48 |
| dtantsur | rloo: have you raised this color issue to someone? | 15:48 |
| rloo | dtantsur: no, i should, i know. do you know who? or should i just do mail list thing. | 15:49 |
| dtantsur | rloo: this feels like something to get an ML thread | 15:49 |
| dtantsur | .. about | 15:49 |
| rloo | dtantsur: ok, will do that today... | 15:49 |
| dtantsur | thanks! | 15:50 |
| dtantsur | -2 POST_FAILURE \o/ | 15:50 |
| shoaib | dntantsur: This isn't the default, This is disabled by default. I used RDO packstack to install openstack, and by default its set to None, and I see no DHCP response coming back to the Bare metal node, maybe this is the cause | 15:51 |
| *** yufei has joined #openstack-ironic | 15:53 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic master: [doc] change absolute to relative URL https://review.openstack.org/496804 | 15:54 |
| rloo | ugh, what happened here? https://review.openstack.org/#/c/495315/ | 15:54 |
| patchbot | patch 495315 - ironic (stable/pike) - Change doc links for pike branch | 15:54 |
| rloo | oh, i see dtantsur has already rechecked. very odd... | 15:55 |
| dtantsur | POST_FAILURE is an infra issue | 15:55 |
| dtantsur | usually when a worker cannot upload the logs back | 15:55 |
| dtantsur | rloo: ^^^ | 15:55 |
| rloo | dtantsur: ahh. thx for explaining. | 15:56 |
| *** rama_y has joined #openstack-ironic | 15:56 | |
| dtantsur | np | 15:56 |
| shoaib | When using nova boot to boot ironic baremetal instance, Do i have to provide the UUID of the Ironic node that was created ? I'm trying to figure out the command.. Can anyone kindly assist or give me the syntax of the full command | 15:59 |
| *** Guest28943 is now known as sdake | 16:00 | |
| *** sdake has joined #openstack-ironic | 16:00 | |
| shoaib | nova boot --image BAREMETAL_USER_IMAGE --flavor BAREMETAL_FLAVOR --nic net-id=IRONIC_NETWORK_ID --key default <IRONIC-NODE_UUID> ? Does this look right ? | 16:00 |
| shoaib | sambetts: nova boot --image BAREMETAL_USER_IMAGE --flavor BAREMETAL_FLAVOR --nic net-id=IRONIC_NETWORK_ID --key default <IRONIC-NODE_UUID> ? Does this command look right ? | 16:01 |
| dtantsur | shoaib: nova picks a node itself, based on the flavor you request. you can't provide it | 16:01 |
| shoaib | dtantsur: So the node nova boots, would have a different uuid from the one that ironic created ? Thanks | 16:02 |
| dtrainor | Hey there TheJulia, you're right, I could have provided some more information about my instackenv.json formatting and using pxe_snmp :) I think it's just a formatting error with driver_info, python-openstackclient never makes an API call. Here's what I'm working with http://paste.openstack.org/show/619198/ | 16:02 |
| dtrainor | Per https://docs.openstack.org/ironic/latest/admin/drivers/snmp.html , and that's why I was looking for an example using driver_info to make sure I wasn't doing something not smart | 16:02 |
| dtantsur | shoaib: nova instance UUID is not related to ironic node UUID. when nova picks a node, it will put the instance UUID into node's instance_uuid field | 16:03 |
| dtantsur | dtrainor: I wonder if tripleo even supports the snmp driver.. | 16:03 |
| TheJulia | dtrainor: also, driver_info is not a list | 16:04 |
| dtantsur | dtrainor: tripleo does not support snmp out of box. you have to use native ironic commands instead of instackenv.json and tripleo commands.. | 16:04 |
| dtantsur | dtrainor: actually, your instackenv.json is a correct file (minus driver_info, which is a dict) to be loaded into ironic directly. this is not an instackenv.json as understood by tripleo | 16:05 |
| rloo | dtantsur: jenkins just gave its nod to this. i still think it may not be worth waiting for but thought i'd let you know: https://review.openstack.org/#/c/496710/ | 16:05 |
| patchbot | patch 496710 - ironic (stable/pike) - Documentation for 'snmp' hardware type | 16:05 |
| * dtantsur saves some good rant | 16:05 | |
| dtrainor | haha | 16:06 |
| dtantsur | rloo: I'm going soon, so I'd prefer to request the release already | 16:06 |
| rloo | dtantsur: ++ | 16:06 |
| TheJulia | dtrainor: doooo it! | 16:06 |
| *** bnemec has quit IRC | 16:06 | |
| * dtantsur watches https://review.openstack.org/#/c/495315/ | 16:07 | |
| patchbot | patch 495315 - ironic (stable/pike) - Change doc links for pike branch | 16:07 |
| dtantsur | I wonder why we even run the coverage test on docs-only patches.. | 16:07 |
| dtrainor | hrm ok then, i'll look in to a) using driver_info as a dict b) look in to using native ironic commands instead of instackenv.json and tripleo commands | 16:07 |
| dtrainor | thank you | 16:07 |
| *** jkilpatr has quit IRC | 16:08 | |
| * milan relocate | 16:08 | |
| *** milan has quit IRC | 16:08 | |
| dtantsur | TheJulia: tomorrow I think of releasing ironic-ui again to include the translations | 16:10 |
| TheJulia | Likely a good idea, I think it was korean that was added | 16:10 |
| dtantsur | I think I've approved German as well | 16:10 |
| dtantsur | yeah, and the proposed on is korean | 16:11 |
| openstackgerrit | Merged openstack/ironic stable/pike: Change doc links for pike branch https://review.openstack.org/495315 | 16:11 |
| dtantsur | w0000000t | 16:11 |
| TheJulia | I'll double check later today | 16:11 |
| dtantsur | and https://review.openstack.org/#/c/496820/ is our release request | 16:13 |
| patchbot | patch 496820 - releases - Release ironic 9.1.0 (final) for Pike | 16:13 |
| *** karimb has quit IRC | 16:20 | |
| *** racedo has quit IRC | 16:20 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost master: DNM: Change perms for the VM logs https://review.openstack.org/494549 | 16:20 |
| *** weshay_PTO_AM is now known as weshay | 16:25 | |
| dtantsur | okay, our release request has got 1x +2, so I think I'm good to call it a day :) | 16:27 |
| dtantsur | see you all tomorrow | 16:27 |
| sambetts | \o/ | 16:27 |
| sambetts | thanks dtantsur, have a good evening | 16:27 |
| *** dtantsur is now known as dtantsur|afk | 16:29 | |
| *** rcernin has quit IRC | 16:29 | |
| *** yufei has quit IRC | 16:29 | |
| *** ijw has quit IRC | 16:30 | |
| rloo | thx and bye dtantsur|afk! | 16:31 |
| *** srobert has quit IRC | 16:31 | |
| shoaib | sambetts: Hi Sam, sorry to bug you again, but i fixed the flavor and Ironic node properties, but when i do Nova compute, i get this error "VirtualInterfacePlugException: Cannot attach VIF 1904bf62-a5cc-4d18-919a-933aa98ad297 to the node 786303b7-4d25-4dac-ad6a-ef5cf05ee99e due to error: UUID or name of cleaning network is not set in configuration ( | 16:32 |
| shoaib | HTTP 400) | 16:32 |
| shoaib | sambetts: I have the properties of flavor and instance matching, but now with nova boot i get this error "error: UUID or name of cleaning network is not set in configuration (HTTP 400 | 16:32 |
| *** e0ne has quit IRC | 16:32 | |
| shoaib | What am i missing now. Whats a cleaning network ? | 16:33 |
| sambetts | that is a configuration issue in your ironic.conf, in the [neutron] section cleaning_network needs to be configured to either the name or the uuid of the network in neutron that should be used as the cleaning network | 16:33 |
| *** mat128 has quit IRC | 16:35 | |
| rama_y | Hi sambetts: when I boot an instance on the baremetal node, cannot ping/ssh into the instance; neutron-port shows status DOWN. Any suggestions? | 16:35 |
| *** yufei has joined #openstack-ironic | 16:38 | |
| sambetts | rama_y: the neutron port will say DOWN because of a bug in ironic + neutron integration but it should actually be working, the ability to ping/ssh into the instance might be two things, 1. is your network accessible from the outside world, if not then you need to use a neutron router and floating ip to provide external access to the node, and 2. have you configured your neutron security groups | 16:38 |
| *** yufei has left #openstack-ironic | 16:38 | |
| sambetts | to allow ping/ssh to the instance | 16:38 |
| shoaib | sambetts: To resolve the cleaning network issue, I must add the flat_network name in this paramter "cleaning_network = <None>" in ironic.conf ? | 16:39 |
| sambetts | shoaib: yes | 16:39 |
| * sambetts --> home | 16:42 | |
| *** sambetts is now known as sambetts|afk | 16:42 | |
| sambetts|afk | Night all! | 16:42 |
| jlvillal | sambetts|afk: Ciao! | 16:42 |
| rloo | good night sambetts|afk | 16:43 |
| rama_y | sambetts|afk, thanks & good night | 16:43 |
| *** jkilpatr has joined #openstack-ironic | 16:53 | |
| *** yangyapeng has quit IRC | 16:53 | |
| *** mgoddard has quit IRC | 16:55 | |
| *** jcoufal has quit IRC | 16:56 | |
| *** trown is now known as trown|lunch | 16:56 | |
| *** amarao1 has quit IRC | 17:00 | |
| *** amarao has quit IRC | 17:00 | |
| *** mtanino_ has joined #openstack-ironic | 17:03 | |
| *** mtanino_ has quit IRC | 17:03 | |
| *** karimb has joined #openstack-ironic | 17:03 | |
| *** mtanino has quit IRC | 17:04 | |
| shoaib | sambetts: I still run into "No Valid Hosts Found" when i do nova-boot according to the nova-conductor.log... The CPU/Memory settings are set to much less then what my HW node can handle | 17:05 |
| shoaib | Anybody still here to help out ? I'm facing issues trying to Boot a bare metal instance with Nova . Keep on getting "No Valid Hosts found. there are not enough hosts available" error.. Thanks | 17:07 |
| *** lucasagomes is now known as lucas-afk | 17:07 | |
| *** pcaruana has quit IRC | 17:08 | |
| *** ijw has joined #openstack-ironic | 17:08 | |
| *** derekh has quit IRC | 17:10 | |
| *** fellypefca has quit IRC | 17:10 | |
| shoaib | To configure Baremetal instances with Nova, Do i need to set "baremetal_scheduler_default_filters" or any specific filters in nova.conf ? Thanks | 17:12 |
| *** tesseract-RH has quit IRC | 17:12 | |
| TheJulia | shoaib: the requested instance needs to match the hardware | 17:14 |
| TheJulia | err, requested flavor | 17:14 |
| TheJulia | the filters, if memory serve, with baremetal result in needing to do exact matching | 17:15 |
| TheJulia | since you can have only one instance per "hypervisor" with the ironic nova virt driver | 17:15 |
| shoaib | TheJulia: When you say, match the hardware, Does it mean, that I have to use all resources available on HW, or just a subset of Available memory , CPUs etc on Hardware ? | 17:18 |
| TheJulia | Exact matching, so if the hardware is 4 cpus, then the nova flavor needs to be 4vcpus | 17:19 |
| TheJulia | if it has 8192MB of ram, then the flavor needs to have that much ram | 17:19 |
| shoaib | TheJulia: Maybe this is a silly question, but if my Hardware node is available state, does ironic has the capability to determine how much HW CPUs, specs are available ? | 17:23 |
| shoaib | TheJulia : or can ironic retrieve that data through IPMI using some CLI command ? Thanks | 17:23 |
| TheJulia | shoaib: That is what we have inspector for and the inspect provision state verb | 17:24 |
| TheJulia | however, that is a separate service and configuration | 17:24 |
| *** racedo has joined #openstack-ironic | 17:24 | |
| TheJulia | IPMI differs so much that there is no real use to trying to get that data out, since I think it would be vendor/chipset specific too. | 17:25 |
| TheJulia | shoaib: inspector will also create ports for the node | 17:25 |
| TheJulia | which is important for PXE/iPXE | 17:25 |
| shoaib | TheJulia: so when i created ironic port create , I also dont see that port in the "neutron port-list" , Does this ironic port gets populated in neutron after Nova boots up an instance ? | 17:27 |
| TheJulia | so the neutron port is virtual, the ironic port is physical. Ironic only creates ports in neutron to do things like attaching to cleaning networks. Nova creates a port for an instance, and tells ironic about that, and then ironic tells neutron to attach that when required. | 17:28 |
| *** makowals has joined #openstack-ironic | 17:28 | |
| TheJulia | so yes, as a part of boot/deploy | 17:29 |
| *** mat128 has joined #openstack-ironic | 17:31 | |
| shoaib | TheJulia: Can you kindly point me to a good set of instructions to start, use ironic-inspector... Cause I don't know about the specs of the HW node (customer specs), I guess using Ironic inspector maybe the quickest way for me | 17:34 |
| *** mgoddard has joined #openstack-ironic | 17:34 | |
| *** mat128 has quit IRC | 17:34 | |
| *** rcernin has joined #openstack-ironic | 17:34 | |
| shoaib | TheJulia: Can the ironic inspector be used as a standalone tool, to use a few CLi commands and use it to retrieve HW specs (CPUD, Memory etc) ? Thanks | 17:34 |
| TheJulia | https://docs.openstack.org/ironic-inspector should have links to installation and use documentation | 17:35 |
| TheJulia | it can't really be used as a stand-alone tool, in terms of retrieving data, it is intended to really search/publish back into ironic | 17:35 |
| *** alex_xu has quit IRC | 17:37 | |
| *** amarao has joined #openstack-ironic | 17:39 | |
| *** alex_xu has joined #openstack-ironic | 17:42 | |
| dtrainor | hmm, 'm getting farther, i can get ironic to recognize the node defined with node-create/node-update etc, but when I try to inspect, I get: EBUG wsme.api [req-59f676bc-f1f8-4dc7-9edc-f6f26b5471c6 a99d2bccad954320ac7fb744f0d8afb1 a7bfd3a5e4ef42e398ebd44f6a96d563 - - -] Client-side error: Driver pxe_snmp does not support management (disabled or not implemented). | 17:43 |
| TheJulia | dtrainor: is pxe_snmp on the enabled_drivers list in ironic.conf | 17:44 |
| openstackgerrit | Ruby Loo proposed openstack/ironic master: [admin docs] ironic -> openstack baremetal CLI https://review.openstack.org/494335 | 17:44 |
| dtrainor | it is, and openstack-ironic-* has been restarted | 17:44 |
| TheJulia | oh | 17:45 |
| TheJulia | oh | 17:45 |
| dtrainor | oh dear | 17:45 |
| * TheJulia looks at somthing | 17:45 | |
| dtrainor | based on what i'm learning, i would wager that some functions of pxe_snmp are simply not developed in to ironic yet to support all the capabilities | 17:46 |
| *** jcoufal has joined #openstack-ironic | 17:46 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic master: [admin docs] ironic -> openstack baremetal CLI https://review.openstack.org/494335 | 17:46 |
| *** trown|lunch is now known as trown | 17:46 | |
| TheJulia | dtrainor: https://github.com/openstack/ironic/blob/master/ironic/drivers/pxe.py#L101 | 17:46 |
| dtrainor | oh. | 17:47 |
| dtrainor | ok, that makes sense | 17:47 |
| TheJulia | management is used to say "Hey, PXE boot new time you power on" or "boot from disk now" | 17:47 |
| dtrainor | riiight | 17:47 |
| TheJulia | and I guess, if you were up for some low level programming and wanted to define an OID... | 17:48 |
| TheJulia | it could be a thing.... | 17:48 |
| shoaib | Hi, I have a question about the sequence in which Baremetal node boot ... when i do ironic-node create, and provide it with the deploy_image, and power is up, is it at this stage that the node is supposed to bootup using tftp and PXe ? | 17:48 |
| dtrainor | i am using properties/capabilities=boot_option:local, wonder if that's making a sad | 17:48 |
| dtrainor | good call, great clue, thank you | 17:48 |
| shoaib | Or do i need to issue a "nova boot command with a user_image" to kickstart the deploy_boot (kernel/ramdisk) | 17:48 |
| TheJulia | dtrainor: likely part of the problem | 17:48 |
| dtrainor | i'll run with that. i appreciate it. | 17:49 |
| TheJulia | shoaib: so adding a node and moving it from manageable to available engages cleaning which does use the deployment image/ramdisk | 17:49 |
| TheJulia | well, deployment kernel/ramdisk | 17:49 |
| TheJulia | shoaib: When deploying, nova boot triggers the state change to deploy the node, which boots the deployment kernel/ramdisk, writes the user image to disk, and reboots the node | 17:50 |
| shoaib | TheJulia: See when i created a node using ironic-create, I didnt manually change the state to available, it came up as available" | 17:50 |
| shoaib | TheJulia: and then i powered my node, at which stage it sends a DHCP request, but doesnt get any response back | 17:50 |
| TheJulia | shoaib: ahh, so that was changed in api version 1.9 | 17:51 |
| *** livelace has joined #openstack-ironic | 17:51 | |
| shoaib | TheJulia: upto this stage, I hadn't issued "nova boot--" command, and thats where I'm confused | 17:51 |
| TheJulia | shoaib: try using the latest api version, 'latest' should cause it to be autonegotated | 17:51 |
| dtrainor | hrm, TheJulia when i verify the node, I see: | inspect | None | not supported | | 17:52 |
| TheJulia | yeah, that kind of makes sense, although it should be supportable | 17:52 |
| TheJulia | hmmmmmmm | 17:52 |
| TheJulia | Where is a milan when we have a question | 17:53 |
| dtrainor | hehe | 17:53 |
| TheJulia | so, basically, if there is an inspect interface, I think you would likely get further | 17:53 |
| dtrainor | as in, there's no inspect interface in pxe_snmp ? | 17:54 |
| *** yangyapeng has joined #openstack-ironic | 17:54 | |
| shoaib | TheJulia: So at what stage does the DHCP response comes in play, Is it the "nova boot" that triggers a DHCP response to be sent back to the BareMetal node that was enrolled >? | 17:54 |
| TheJulia | dtrainor: maybe something like https://github.com/openstack/ironic/blob/master/ironic/drivers/ipmi.py#L72 | 17:54 |
| TheJulia | dtrainor: no, there is not from what I saw | 17:55 |
| dtrainor | oh i see what you did there | 17:55 |
| TheJulia | dtrainor: https://github.com/openstack/ironic/blob/master/ironic/drivers/pxe.py#L101 and look 21 lines below :) | 17:55 |
| dtrainor | yep i saw that, cool | 17:56 |
| TheJulia | shoaib: there are a few steps in between.... | 17:56 |
| dtrainor | i'll hack on this for a minute, thanks again | 17:56 |
| TheJulia | dhcp response is set by ironic upon attaching the port | 17:56 |
| *** racedo has quit IRC | 17:56 | |
| TheJulia | so when ironic powers the node up, the correct reply is sent by neutron | 17:56 |
| TheJulia | well, whatever neutron's dhcp server is, typically dnsmasq | 17:57 |
| *** yufei has joined #openstack-ironic | 17:58 | |
| *** livelace has quit IRC | 17:58 | |
| shoaib | TheJulia: See thats my problem, I did do an ironic port-create, after creating an ironic node, this is all before nova-boot, and I dont see this ironic port attaching to neutron, which you are saying is very very important to get a DHCP response | 17:58 |
| *** yangyapeng has quit IRC | 17:58 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic master: [admin docs] ironic -> openstack baremetal CLI https://review.openstack.org/494335 | 17:58 |
| TheJulia | replies are going to be delayed, I have to jump on a call | 17:59 |
| shoaib | TheJulia: So Nova boot is never going to work for me, if my Node doesn't even get a DHCP response, right ? As neutron doesn't have the port created with the MAC Address of the baremetal node in its list ? | 17:59 |
| *** e0ne has joined #openstack-ironic | 18:00 | |
| *** yufei has left #openstack-ironic | 18:00 | |
| *** amarao has quit IRC | 18:04 | |
| *** e0ne has quit IRC | 18:13 | |
| *** e0ne has joined #openstack-ironic | 18:22 | |
| shoaib | TheJulia: Should nova-boot be triggered after the Node is up, available, and has a DHCP response received ? Thanks | 18:31 |
| *** makowals has quit IRC | 18:31 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic master: remove REST API examples from RAID doc https://review.openstack.org/496840 | 18:31 |
| openstackgerrit | Michael Johnson proposed openstack/python-ironicclient master: Add auto-generated CLI reference https://review.openstack.org/496841 | 18:33 |
| *** e0ne has quit IRC | 18:33 | |
| openstackgerrit | Michael Johnson proposed openstack/python-ironicclient master: Add auto-generated CLI reference https://review.openstack.org/496841 | 18:35 |
| *** jcoufal has quit IRC | 18:36 | |
| *** livelace has joined #openstack-ironic | 18:37 | |
| *** jcoufal has joined #openstack-ironic | 18:38 | |
| *** e0ne has joined #openstack-ironic | 18:40 | |
| *** livelace has quit IRC | 18:41 | |
| *** ChubYann has joined #openstack-ironic | 18:43 | |
| *** etingof has quit IRC | 18:44 | |
| *** rcernin has quit IRC | 18:47 | |
| *** johnsom has joined #openstack-ironic | 18:50 | |
| *** jtomasek_ has quit IRC | 18:52 | |
| TheJulia | shoaib: okay, back | 18:53 |
| TheJulia | shoaib: so, all attachments take place after the nova boot | 18:53 |
| TheJulia | you can set default dhcp replies for things like inspection in your dnsmasq config | 18:54 |
| TheJulia | ironic takes care of that neutron interaction for port creation, so if you have created a port in neutron manually for the node, you need to delete it | 18:54 |
| TheJulia | nova boot is what starts the process, so the node should be off, and booting the node should result in the power being turned on | 18:55 |
| *** yangyapeng has joined #openstack-ironic | 18:55 | |
| shoaib | TheJulia: I didn't create any port in neutron manually, I just created the ironic port.. | 18:56 |
| TheJulia | shoaib: so what exactly is your current state, and what is your expected state? | 18:57 |
| TheJulia | because I feel like there is a disconnect there | 18:57 |
| *** mgoddard has quit IRC | 18:57 | |
| shoaib | TheJulia: At this stage, I have an ironic node status as "available " { which came up by itself } , and power is on , but Nova boot keeps failing with this error : https://thepasteb.in/p/pghQ3v2r9RnSR | 18:58 |
| shoaib | This is my output from "ironic node-show " https://thepasteb.in/p/LghNnmW6p9jfZ | 18:58 |
| *** yangyapeng has quit IRC | 18:59 | |
| TheJulia | shoaib: so that looks like a pure scheduling issue. I'd verify that your nova-compute process is running with the ironic virt driver, and check that the scheduler process is seeing inventory before trying to boot a node | 18:59 |
| *** etingof has joined #openstack-ironic | 19:00 | |
| shoaib | TheJulia: I'm not sure about the expected state, frankly i expected my BareMetal Node to start booting with the Deployment ramdisk/kernel .. Instead the bare-metal node shows error saying that "No DHCP response received" | 19:00 |
| TheJulia | Okay, so you can power down the baremetal node, no sense in having it powered on right now | 19:01 |
| TheJulia | your issue, presently is entirely before ironic is told to deploy | 19:01 |
| shoaib | TheJulia: Allright, Can you also kindly clarify what you mean by "ironic takes care of that neutron interaction for port creation" ... Somebody here earlier mentioned that when you do ironic port-create, that port should have a ""vif_port_id" too so that it can be populated in the neutron-lists | 19:04 |
| TheJulia | you will only have a vif_port_id once nova has tried to initiate a deployment | 19:05 |
| shoaib | TheJulia: This is the output from my ironic port https://thepasteb.in/p/RghnqRYvwpJtz | 19:05 |
| TheJulia | but your not even getting that far because the scheduler is failing | 19:05 |
| shoaib | TheJulia: Oh ok, I get it .. So basically Nova is the root of the problem, alhtough i see nova-compute and the nova-schedular running | 19:06 |
| TheJulia | they may be running, but I suspect there will be errors in the logs | 19:07 |
| TheJulia | or it will just not be seeing the baremetal hardware as available resources | 19:07 |
| shoaib | TheJulia: also in the latest release, you mentioned that an ironic node created with status "enroll", and then a user has to manually set it to "deploy ", and "available " state, Is that correct ? Or is that only going to happen in Pike release ? | 19:09 |
| TheJulia | so, kind of | 19:11 |
| TheJulia | not deploy, that is what nova does | 19:11 |
| TheJulia | https://docs.openstack.org/ironic/latest/contributor/states.html might help show how it works | 19:12 |
| *** ijw has quit IRC | 19:13 | |
| TheJulia | shoaib: and it is not in the latest release, it has been in the api for... four releases I believe | 19:17 |
| shoaib | TheJulia: Ok, so basically as far as I understand this state machine, the state will go to "Deploy" after the "Available" state, and its nova that kicks it into that mode | 19:17 |
| TheJulia | correct | 19:18 |
| TheJulia | A human _can_ do it manually without nova, but the vast majority of users use nova | 19:18 |
| *** srobert has joined #openstack-ironic | 19:19 | |
| *** mat128 has joined #openstack-ironic | 19:20 | |
| *** ijw has joined #openstack-ironic | 19:21 | |
| shoaib | TheJulia: Where is the nova-compute supposed to run, on the bare-metal node, or on the Controller node ? | 19:23 |
| rloo | mariojv: if you're still interested, look what someone just submitted: https://review.openstack.org/#/c/496841/ | 19:23 |
| patchbot | patch 496841 - python-ironicclient - Add auto-generated CLI reference | 19:23 |
| TheJulia | shoaib: controller node | 19:25 |
| *** ijw has quit IRC | 19:28 | |
| *** ijw has joined #openstack-ironic | 19:29 | |
| *** baoli has joined #openstack-ironic | 19:34 | |
| *** baoli has quit IRC | 19:38 | |
| shoaib | TheJulia: Yeah in my case, looks like nova cant even see the baremetal node. In my hypervisor-list, all i see is the Ironic node, Shouldn't a Compute-Host name also be under the Hypervisor-list ? | 19:42 |
| *** mgoddard has joined #openstack-ironic | 19:44 | |
| TheJulia | Sounds kind of like your nova-compute node is misconfigured | 19:44 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic-ui master: Unit tests for set target raid configuration https://review.openstack.org/496857 | 19:45 |
| shoaib | TheJulia: I'm looking in the nova.conf file, and I also see "baremetal_enabled_filters" disabled, and "use_baremetal_filter=False" ... Would this impact nova from finding a baremetal node host ? | 19:46 |
| TheJulia | Maybe, we're reaching the edge of my Nova knoweldge. I would recommend checking out install docs | 19:47 |
| osdev | TheJulia: In the non-baremetal case, nova-compute runs on the host which is hosting the virtual machines; it seems in the baremetal/Ironic case we expect nova-compute to also run on the "controller" -("controller" here meaning where nova-scheduler, ironic-conductor) is currently running? (Or did I misinterpret?) | 19:53 |
| *** yangyapeng has joined #openstack-ironic | 19:56 | |
| TheJulia | We don't really care where it runs, as long as it is considered a running controller service as it is what queries and returns data from ironic to nova | 20:00 |
| *** yangyapeng has quit IRC | 20:01 | |
| TheJulia | Naturally, it can't run on an undeployed baremetal node since it is actually what is executing the virt driver | 20:01 |
| johnsom | rloo Hi, thanks for the bug. I do have a question about one of your comments however. The index you were requesting, would that be on the osc_plugin_cli.rst page? After you click into osc/v1/index.html the index is in the left nav | 20:06 |
| rloo | hi johnsom. give me a few secs to find it... | 20:07 |
| rloo | johnsom: so you're asking me about this page? https://docs.openstack.org/python-octaviaclient/latest/cli/index.html | 20:08 |
| rloo | johnsom: where i'd like to see that equivalent in ironic? | 20:08 |
| johnsom | rloo Yeah, would you like that tree in the http://docs-draft.openstack.org/41/496841/2/check/gate-python-ironicclient-docs-ubuntu-xenial/1fbc257//doc/build/html/cli/osc_plugin_cli.html#command-reference section? | 20:08 |
| rloo | johnsom: ideally no. but i think for now, it might be easiest. let me think. | 20:09 |
| johnsom | rloo On the new way, the list moves to the left nav, like here: http://docs-draft.openstack.org/27/496827/1/check/gate-python-octaviaclient-docs-ubuntu-xenial/9b45153//doc/build/html/cli/index.html | 20:09 |
| rloo | johnsom: right, but it is ugly. | 20:10 |
| rloo | johnsom: i mean, i guess when we get used to the stuff on the left nav | 20:10 |
| osdev | TheJulia: thanks, I suppose even if we had many baremetal nodes, the single instance of nova-compute (which could be running on a controller) would handle the returning of data from ironic to nova. hope I understood that corrrectly. | 20:10 |
| johnsom | rloo I don't disagree | 20:10 |
| rloo | johnsom: right now, i think that stuff is ugly so i'm not used to it | 20:10 |
| rloo | johnsom: the extra neurons i have to use to actually distinguish the two left panes or whatever :-( | 20:11 |
| johnsom | rloo I wish we had more control over the left nav behavior | 20:11 |
| TheJulia | osdev: correct, although you can run multiple nova-compute processes with ironic, they will distribute the baremetal nodes they are responsible for and acting upon across the running nova-compute services | 20:12 |
| rloo | johnsom: i'm just focussed on changing the RED for now... | 20:12 |
| rloo | johnsom: but that doesn't answer your question :) | 20:12 |
| rloo | johnsom: is the plan to get rid of https://docs.openstack.org/python-octaviaclient/latest/cli/index.html? | 20:12 |
| johnsom | rloo Well, that is what I did, but your comments are making me thing about it. The challenge is with that intermediate page the left nav gets odd | 20:13 |
| rloo | johnsom: cuz I find when I look at http://docs-draft.openstack.org/41/496841/2/check/gate-python-ironicclient-docs-ubuntu-xenial/1fbc257//doc/build/html/cli/osc_plugin_cli.html#command-reference I only see 5 things and there are way more commands than that. i know it isn't listing the commands but it almost seems that way. | 20:13 |
| johnsom | I can bump that up to go another level deep and see what happens | 20:14 |
| rloo | johnsom: ok. if it is too hard, i am happy to get what we get, which is better than what we had. but if I can ask, I'll ask :) | 20:14 |
| rloo | johnsom: am i correct in thinking that you are the only one adding these pages, so you are setting the precedent? | 20:15 |
| *** Goneri has quit IRC | 20:15 | |
| johnsom | rloo lol, ummm, I don't know if I would go that far... Just wanted good docs for octavia and got asked to do the same for ironic. | 20:16 |
| rloo | johnsom: ha ha! | 20:16 |
| johnsom | rloo So, it works fine with adding the layer down. | 20:16 |
| rloo | johnsom: sweet. | 20:16 |
| johnsom | rloo At some point that OSC page could be reworked, but that is a style thing IMO. | 20:17 |
| rloo | johnsom: we're all so busy that it is best to get it as 'right as we can' when we do it... | 20:18 |
| johnsom | rloo Yeah, immediately after this I need to go start beating on the octavia-dashboard, so I understand busy. | 20:19 |
| rloo | johnsom: I appreciate you offering to help out ironic! | 20:20 |
| openstackgerrit | Michael Johnson proposed openstack/python-ironicclient master: Add auto-generated CLI reference https://review.openstack.org/496841 | 20:20 |
| rloo | johnsom: if what I'm asking for is too much work, don't do it :) | 20:20 |
| johnsom | rloo Take a look at that and let me know what you think. | 20:20 |
| rloo | johnsom: will do, thx! | 20:20 |
| *** jkilpatr has quit IRC | 20:22 | |
| *** dprince has quit IRC | 20:26 | |
| *** rcernin has joined #openstack-ironic | 20:35 | |
| *** mgoddard has quit IRC | 20:43 | |
| *** jcoufal has quit IRC | 20:43 | |
| rloo | johnsom: thx, +2! | 20:43 |
| johnsom | rloo Cool, happy to help | 20:44 |
| rloo | johnsom: :) Now let's hope for more agreement! | 20:44 |
| *** shoaib has quit IRC | 20:48 | |
| *** e0ne has quit IRC | 20:52 | |
| *** yangyapeng has joined #openstack-ironic | 20:56 | |
| *** masber has quit IRC | 20:57 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic master: [TEST] [DO NOT MERGE] Testing tls-proxy enabling https://review.openstack.org/492664 | 21:00 |
| *** trown is now known as trown|outtypewww | 21:01 | |
| *** yangyapeng has quit IRC | 21:01 | |
| *** srobert has quit IRC | 21:10 | |
| *** racedo has joined #openstack-ironic | 21:10 | |
| *** crushil has quit IRC | 21:15 | |
| *** racedo has quit IRC | 21:20 | |
| *** mat128 has quit IRC | 21:23 | |
| *** bfournie has quit IRC | 21:25 | |
| *** cdearborn has quit IRC | 21:28 | |
| *** jkilpatr has joined #openstack-ironic | 21:29 | |
| *** crushil has joined #openstack-ironic | 21:32 | |
| *** ijw has quit IRC | 21:53 | |
| *** jlabarre has quit IRC | 21:53 | |
| *** ijw has joined #openstack-ironic | 21:56 | |
| *** yangyapeng has joined #openstack-ironic | 21:57 | |
| mariojv | thanks rloo and johnsom, +1ed it | 22:00 |
| rloo | mariojv: thx! :) | 22:00 |
| *** yangyapeng has quit IRC | 22:02 | |
| *** ijw has quit IRC | 22:07 | |
| *** ijw has joined #openstack-ironic | 22:20 | |
| *** crushil has quit IRC | 22:29 | |
| *** rpioso has quit IRC | 22:43 | |
| *** yangyapeng has joined #openstack-ironic | 22:58 | |
| *** mjturek has quit IRC | 23:00 | |
| *** yangyapeng has quit IRC | 23:02 | |
| *** athomas has quit IRC | 23:11 | |
| *** srobert has joined #openstack-ironic | 23:24 | |
| *** srobert has quit IRC | 23:26 | |
| *** karimb has quit IRC | 23:30 | |
| *** lennyb has quit IRC | 23:31 | |
| *** rloo has quit IRC | 23:33 | |
| *** baoli has joined #openstack-ironic | 23:34 | |
| *** baoli has quit IRC | 23:39 | |
| *** lennyb has joined #openstack-ironic | 23:41 | |
| *** charcol has joined #openstack-ironic | 23:44 | |
| *** ijw has quit IRC | 23:47 | |
| *** bnemec has joined #openstack-ironic | 23:54 | |
| *** ijw has joined #openstack-ironic | 23:58 | |
| *** yangyapeng has joined #openstack-ironic | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!