| *** matsuhashi has joined #openstack-ironic | 00:30 | |
| *** rloo has quit IRC | 00:33 | |
| *** rloo has joined #openstack-ironic | 00:34 | |
| *** rloo has quit IRC | 00:36 | |
| *** rloo has joined #openstack-ironic | 00:36 | |
| *** rloo has quit IRC | 00:38 | |
| *** rloo has joined #openstack-ironic | 00:38 | |
| *** takadayuiko has joined #openstack-ironic | 00:47 | |
| *** rloo has quit IRC | 00:47 | |
| *** rloo has joined #openstack-ironic | 00:49 | |
| lifeless | adam_g: flake8 I think - https://review.openstack.org/#/c/96577/ | 00:49 |
|---|---|---|
| *** igordcard_ has quit IRC | 00:49 | |
| lifeless | adam_g: (haven't looked closely, but both python versions failed...) | 00:49 |
| adam_g | lifeless, well, theres test failures too. im actually building a venv to look now | 00:49 |
| lifeless | ah, nuts | 00:49 |
| *** rloo has quit IRC | 00:55 | |
| *** Penick has quit IRC | 00:55 | |
| *** rloo has joined #openstack-ironic | 00:55 | |
| *** rloo has quit IRC | 00:59 | |
| *** rloo has joined #openstack-ironic | 01:00 | |
| adam_g | lifeless, should be better after this merges https://review.openstack.org/#/c/96576/1 | 01:00 |
| *** ellenh has quit IRC | 01:05 | |
| *** takadayuiko has quit IRC | 01:05 | |
| *** bandicot has quit IRC | 01:14 | |
| *** rloo has quit IRC | 01:17 | |
| *** rloo has joined #openstack-ironic | 01:18 | |
| *** dwalleck has joined #openstack-ironic | 01:29 | |
| *** dwalleck has quit IRC | 01:34 | |
| *** dwalleck has joined #openstack-ironic | 01:34 | |
| *** nosnos has joined #openstack-ironic | 01:47 | |
| *** eguz has joined #openstack-ironic | 01:52 | |
| *** eghobo has quit IRC | 01:56 | |
| *** rloo has quit IRC | 02:06 | |
| *** zdin0bot has joined #openstack-ironic | 02:28 | |
| *** harlowja_ is now known as harlowja_away | 02:41 | |
| *** matsuhashi has quit IRC | 02:45 | |
| *** shakayumi has quit IRC | 02:58 | |
| *** eghobo has joined #openstack-ironic | 03:19 | |
| *** praefect has joined #openstack-ironic | 03:32 | |
| *** nosnos has quit IRC | 03:40 | |
| *** rameshg87 has joined #openstack-ironic | 03:42 | |
| *** matsuhashi has joined #openstack-ironic | 03:47 | |
| *** k4n0 has joined #openstack-ironic | 03:54 | |
| *** matsuhashi has quit IRC | 03:54 | |
| *** matsuhashi has joined #openstack-ironic | 03:54 | |
| *** zdin0bot has quit IRC | 03:55 | |
| *** matsuhashi has quit IRC | 03:57 | |
| *** matsuhashi has joined #openstack-ironic | 03:58 | |
| *** Jatin360 has joined #openstack-ironic | 03:59 | |
| *** matsuhashi has quit IRC | 04:03 | |
| *** matsuhashi has joined #openstack-ironic | 04:03 | |
| *** Jatin360 has quit IRC | 04:06 | |
| *** takadayuiko has joined #openstack-ironic | 04:10 | |
| takadayuiko | Hello Ironic :D | 04:10 |
| k4n0 | 0/ | 04:20 |
| *** zdin0bot has joined #openstack-ironic | 04:26 | |
| *** jcoufal has joined #openstack-ironic | 04:28 | |
| *** nosnos has joined #openstack-ironic | 04:42 | |
| *** dwalleck_ has joined #openstack-ironic | 04:45 | |
| *** dwalleck has quit IRC | 04:48 | |
| *** lokesh184 has joined #openstack-ironic | 04:55 | |
| *** rakesh_hs has joined #openstack-ironic | 05:04 | |
| *** radsy has quit IRC | 05:16 | |
| *** lazy_prince has joined #openstack-ironic | 05:20 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 05:49 | |
| *** vinbs has joined #openstack-ironic | 05:58 | |
| *** matsuhashi has quit IRC | 06:30 | |
| *** matsuhashi has joined #openstack-ironic | 06:31 | |
| *** pradipta_away is now known as pradipta | 06:45 | |
| GheRivero | morning all | 06:59 |
| *** matsuhashi has quit IRC | 07:11 | |
| *** sysexit has joined #openstack-ironic | 07:11 | |
| *** matsuhas_ has joined #openstack-ironic | 07:16 | |
| *** matsuhas_ has quit IRC | 07:16 | |
| *** matsuhashi has joined #openstack-ironic | 07:17 | |
| *** takadayuiko has quit IRC | 07:19 | |
| mrda | hi GheRivero | 07:22 |
| *** pradipta is now known as pradipta_away | 07:25 | |
| *** ifarkas has joined #openstack-ironic | 07:34 | |
| *** mrda is now known as mrda_awak | 07:36 | |
| mrda_awak | Night Ironic | 07:36 |
| *** max_lobur has joined #openstack-ironic | 07:37 | |
| *** eghobo has quit IRC | 07:39 | |
| *** davidlenwell has quit IRC | 07:55 | |
| *** davidlenwell has joined #openstack-ironic | 07:55 | |
| *** kevinbenton has quit IRC | 07:56 | |
| *** dguerri has quit IRC | 07:56 | |
| *** jgrimm has quit IRC | 07:56 | |
| *** NobodyCam has quit IRC | 07:56 | |
| *** sysexit has quit IRC | 07:57 | |
| *** devananda has quit IRC | 07:57 | |
| *** jrist has quit IRC | 07:57 | |
| *** devananda has joined #openstack-ironic | 07:58 | |
| *** dguerri has joined #openstack-ironic | 08:01 | |
| *** kevinbenton has joined #openstack-ironic | 08:01 | |
| *** jrist has joined #openstack-ironic | 08:03 | |
| *** NobodyCam has joined #openstack-ironic | 08:04 | |
| *** sysexit has joined #openstack-ironic | 08:04 | |
| *** jgrimm has joined #openstack-ironic | 08:04 | |
| *** lsmola has quit IRC | 08:07 | |
| vinbs | Hello Ironic | 08:11 |
| GheRivero | hi | 08:11 |
| *** lsmola has joined #openstack-ironic | 08:13 | |
| *** romcheg has joined #openstack-ironic | 08:14 | |
| *** derekh_ has joined #openstack-ironic | 08:16 | |
| *** martyntaylor has joined #openstack-ironic | 08:17 | |
| *** martyntaylor has quit IRC | 08:17 | |
| *** jistr has joined #openstack-ironic | 08:19 | |
| vinbs | I have an all in one node openstack ice house setup and I'm trying to deploy an instance on a baremetal node using ironic | 08:20 |
| vinbs | but when I try to launch an instance | 08:21 |
| vinbs | the provision_state is "deploy failed" and the last_error is "Failed to deploy. Error: 'kernel_id'" | 08:21 |
| vinbs | I'm not able to figure out what the problem is | 08:22 |
| vinbs | any help on this would be great | 08:22 |
| *** martyntaylor has joined #openstack-ironic | 08:23 | |
| *** ndipanov has joined #openstack-ironic | 08:28 | |
| *** lucasagomes has joined #openstack-ironic | 08:32 | |
| *** lazy_prince2 has joined #openstack-ironic | 08:33 | |
| *** lazy_prince has quit IRC | 08:35 | |
| *** lazy_prince2 has quit IRC | 08:37 | |
| *** lokesh184 has quit IRC | 08:40 | |
| *** lazy_prince has joined #openstack-ironic | 08:40 | |
| *** lokesh184 has joined #openstack-ironic | 08:41 | |
| *** max_lobur has quit IRC | 08:42 | |
| *** radsy has joined #openstack-ironic | 08:46 | |
| *** radsy has joined #openstack-ironic | 08:46 | |
| *** dtantsur|afk is now known as dtantsur | 08:49 | |
| dtantsur | Morning Ironic! | 08:50 |
| romcheg | Morning dtantsur! How're you today? | 08:50 |
| lucasagomes | morning dtantsur romcheg | 08:50 |
| dtantsur | romcheg, morning, much better, thank you! | 08:50 |
| lucasagomes | dtantsur, feeling better? | 08:50 |
| lucasagomes | and morning Ironic :) | 08:50 |
| dtantsur | lucasagomes, morning :) yeah, better | 08:50 |
| lucasagomes | ^^ good good | 08:51 |
| romcheg | vinbs: could you please post conductor's log to http://paste.openstack.org and give us a link? | 08:51 |
| dtantsur | ... still I'm looking forward to PTO in 1 week.. I'm quite tired, haven't been on vacation since September | 08:51 |
| romcheg | dtantsur: w00t! | 08:51 |
| vinbs | romcheg, let me do that | 08:51 |
| dtantsur | vinbs, and make sure you have proper deploy kernel and ramdisk assigned to flavor | 08:53 |
| vinbs | dtantsur, the deploy kernel and ramdisk I mention in ironic node-create and the ones assigned to the flavor are the same | 08:55 |
| *** igordcard has joined #openstack-ironic | 08:59 | |
| vinbs | romcheg, here's ironic-conductor log http://paste.openstack.org/show/82162/ | 09:00 |
| lucasagomes | dtantsur, nice! yeah I also need some holidays | 09:01 |
| lucasagomes | dtantsur, r u travelling somewhere? | 09:01 |
| dtantsur | this June I'll be partly in Brno, partly in Prague wit my parent who arrive from Russia to see me | 09:01 |
| dtantsur | in July I'm going to Paris | 09:02 |
| dtantsur | will be fun summer, I hope :) | 09:02 |
| romcheg | Guys, let's go to Lviv for a few days :) | 09:02 |
| *** lokesh184 has quit IRC | 09:03 | |
| dtantsur | romcheg, I would, I like Lviv a lot and my wife grew up there | 09:03 |
| dtantsur | romcheg, but one with Russian passport has no chances of getting there now :( | 09:03 |
| romcheg | dtantsur: I don't think so. There are problems for guys coming from Russia to western regions. If you travel from Czech Republic to Lviv, they should let you go | 09:04 |
| romcheg | s/western/eastern | 09:05 |
| *** lokesh184 has joined #openstack-ironic | 09:05 | |
| *** lazy_prince has quit IRC | 09:06 | |
| *** lazy_prince2 has joined #openstack-ironic | 09:07 | |
| lucasagomes | dtantsur, nice! | 09:10 |
| dtantsur | romcheg, afaik they care about citizenship, not about particular flight :( not sure of course | 09:11 |
| dtantsur | iirc now it's officially forbidden for Russian male 16-55 (or smth) to attend Ukraine | 09:11 |
| vinbs | romcheg, dtantsur, This question on ask.openstack explains the same problem I'm facing https://ask.openstack.org/en/question/30707/failed-to-deploy-error-kernel_id/ | 09:12 |
| vinbs | with more details | 09:12 |
| romcheg | vinbs: /me is looking at the log | 09:12 |
| dtantsur | I really hope the situation is going to be solved for good, and we'll have chance of having a meet-up in Lviv | 09:12 |
| romcheg | dtantsur: Not if you travel with family or have an invite. I can call and ask them, if you need | 09:13 |
| dtantsur | would be nice to know | 09:13 |
| dtantsur | my wife's parent were unable to go to Lviv to visit their parents | 09:14 |
| *** Mikhail_D_ltp has quit IRC | 09:23 | |
| lucasagomes | vinbs, the image you've in glance has the kernel_id and ramdisk_id property set? | 09:24 |
| lucasagomes | e.g | 09:24 |
| lucasagomes | vinbs, http://paste.openstack.org/show/82167/ | 09:24 |
| lucasagomes | as we are pxe booting the image kernel and ramdisk should be extracted from it | 09:26 |
| lucasagomes | (we will support local boot as well, where it won't be needed, but right now we don't) | 09:26 |
| vinbs | lucasagomes, no I don't see those two properties set | 09:27 |
| lucasagomes | vinbs, right ironic relies on it | 09:28 |
| vinbs | these are the images I have http://paste.openstack.org/show/82168/ | 09:29 |
| vinbs | lucasagomes, how do I set these properties? | 09:30 |
| *** lazy_prince2 has quit IRC | 09:31 | |
| lucasagomes | vinbs, when doing a glance image-create you can pass it using --property | 09:31 |
| lucasagomes | vinbs, there's a tripleo script that does it all | 09:31 |
| *** lazy_prince has joined #openstack-ironic | 09:31 | |
| lucasagomes | vinbs, | 09:31 |
| lucasagomes | https://github.com/openstack/tripleo-incubator/blob/master/scripts/load-image | 09:31 |
| *** rameshg87 has quit IRC | 09:31 | |
| lucasagomes | but in Ironic itself, we need to improve the usability | 09:31 |
| *** max_lobur has joined #openstack-ironic | 09:32 | |
| romcheg | lucasagomes: That seems to be a bug | 09:32 |
| lucasagomes | vinbs, did you open a bug in Ironic about it? | 09:32 |
| lucasagomes | romcheg, yes | 09:32 |
| vinbs | lucasagomes, let me add these properties and try | 09:32 |
| lucasagomes | vinbs, ack | 09:32 |
| vinbs | no I haven't opened a bug for this | 09:32 |
| vinbs | should I? | 09:32 |
| *** sysexit has quit IRC | 09:32 | |
| lucasagomes | vinbs, would be good :) lemme see if there's one already open | 09:33 |
| vinbs | lucasagomes, sure.. thanks :) | 09:34 |
| lucasagomes | vinbs, seems there's none open about it, I will file one, feel free to add more logs stuff to it since you mighty have it handy there | 09:36 |
| vinbs | lucasagomes, sure | 09:37 |
| *** pelix has joined #openstack-ironic | 09:39 | |
| vinbs | lucasagomes, Do I need to set any dhcp options manually? | 09:45 |
| lucasagomes | vinbs, neutron should set the dhcp for u | 09:46 |
| vinbs | lucasagomes, does it need special configuration in neutron? | 09:48 |
| vinbs | I haven't specifically set any dhcp options after setting up ironic | 09:48 |
| lucasagomes | vinbs, hmm that's the area I know least in OpenStack hmm I don't know if u have to set any configuration in neutron, I would say yes you probably have but I gotta check what devstack is doing in order to know what | 09:49 |
| * lucasagomes will take a look in a bit | 09:50 | |
| *** takadayuiko has joined #openstack-ironic | 09:57 | |
| *** sysexit has joined #openstack-ironic | 10:04 | |
| lucasagomes | vinbs, https://review.openstack.org/#/c/96710, hope it helps with this error | 10:14 |
| lucasagomes | if the image does not contain those properties it's going to fail at validation time and not after the deploy has started | 10:15 |
| lucasagomes | and give a better error message as well | 10:15 |
| *** matsuhashi has quit IRC | 10:16 | |
| vinbs | lucasagomes, adding those two properties to the image solved the issue | 10:17 |
| lucasagomes | vinbs, cool :) | 10:17 |
| vinbs | now I'm trying to get dhcp working | 10:17 |
| lucasagomes | vinbs, right... sometimes neutron just drives me crazy | 10:17 |
| vinbs | my server is not able to receive any dhcp offers from openstack | 10:18 |
| lucasagomes | vinbs, if u want, you can start up ur own dhcp server | 10:18 |
| lucasagomes | for tests | 10:18 |
| lucasagomes | vinbs, sudo dnsmasq --conf-file= --port=69 --enable-tftp --tftp-root=/tftpboot --dhcp-boot=pxelinux.0 --bind-interfaces --pid-file=/var/run/dnsmasq.pid --interface=ens9 --dhcp-range=192.0.2.32,192.0.2.48 | 10:18 |
| lucasagomes | that's how I started my dhcp server manually before | 10:18 |
| vinbs | yes.. starting a dnsmasq manually has worked for me too before with devstack | 10:19 |
| vinbs | I'll try to get dnsmasq created by neutron to work with the ironic node | 10:20 |
| *** matsuhashi has joined #openstack-ironic | 10:21 | |
| *** matsuhashi has quit IRC | 10:27 | |
| *** matsuhashi has joined #openstack-ironic | 10:28 | |
| takadayuiko | vinbs, lucasagomes: I'm trying to baremetal provisioning with Ironic, too :D Your conversation is very useful to me! | 10:29 |
| lucasagomes | takadayuiko, glad it helps! | 10:32 |
| lucasagomes | takadayuiko, and good afternoon/evening there :) | 10:32 |
| takadayuiko | lucasagomes, Thx, it's my first baremetal provisioning! | 10:35 |
| lucasagomes | takadayuiko, nice! please let us know if u have any problems deploying it ( or success :) ) | 10:36 |
| takadayuiko | lucasagomes, yes, it's about time for dinner :-) | 10:37 |
| lucasagomes | takadayuiko, :) bon appetit! | 10:38 |
| *** sysexit has quit IRC | 10:40 | |
| takadayuiko | lucasagomes, thank you, there are many many problems :-@ Especially I lack netwotk knowledge, so I have to study about them... | 10:40 |
| lucasagomes | takadayuiko, heh +1, ur not alone, I also have to study more about networks :( | 10:43 |
| lucasagomes | btw you might want to take a look at https://www.coursera.org/course/sdn | 10:43 |
| *** dtantsur is now known as dtantsur|lunch | 10:43 | |
| lucasagomes | I think you might be able to join it even tho it already started | 10:43 |
| *** BadCub has quit IRC | 10:44 | |
| takadayuiko | Wao e-learning?? | 10:45 |
| *** BadCub has joined #openstack-ironic | 10:45 | |
| lucasagomes | takadayuiko, yeah, it's a free e-course about sdn | 10:48 |
| lucasagomes | seems grand | 10:48 |
| takadayuiko | Oh, think I'll try :-) | 10:50 |
| lucasagomes | :) | 10:52 |
| *** BadCub has quit IRC | 10:54 | |
| *** BadCub has joined #openstack-ironic | 10:55 | |
| *** lokesh184 has quit IRC | 10:57 | |
| *** matsuhashi has quit IRC | 11:00 | |
| *** shausy has joined #openstack-ironic | 11:00 | |
| *** matsuhashi has joined #openstack-ironic | 11:00 | |
| *** BadCub has quit IRC | 11:01 | |
| *** NobodyCam has quit IRC | 11:01 | |
| *** NobodyCam has joined #openstack-ironic | 11:02 | |
| *** BadCub has joined #openstack-ironic | 11:04 | |
| *** matsuhashi has quit IRC | 11:05 | |
| shausy | Im using openstack icehouse to validate cinder 3par FC, Create /delete works fine but when I try to attach the volume im getting an exception " VolumeBackendAPIException: Bad or unexpected response from the storage volume backend API: Unable to fetch connection information from backend: 'wwpns'" | 11:07 |
| shausy | Do i need configuer host information in 3par prior to testing this | 11:07 |
| lucasagomes | shausy, wrong channel? maybe asking at #openstack-cinder? | 11:08 |
| shausy | sorry | 11:08 |
| lucasagomes | shausy, :) no worries | 11:09 |
| *** matsuhashi has joined #openstack-ironic | 11:09 | |
| *** BadCub_ has joined #openstack-ironic | 11:10 | |
| *** BadCub has quit IRC | 11:13 | |
| takadayuiko | DHCP issue was solved! Yeah! | 11:14 |
| lucasagomes | w00t! | 11:16 |
| lucasagomes | it's lunch time for me :) brb | 11:16 |
| *** lucasagomes is now known as lucas-hungry | 11:16 | |
| takadayuiko | enjoy your lunch! | 11:18 |
| *** openstackgerrit has joined #openstack-ironic | 11:26 | |
| *** jcoufal has quit IRC | 11:28 | |
| *** jcoufal has joined #openstack-ironic | 11:28 | |
| takadayuiko | Good bye everyone ;) | 11:49 |
| *** takadayuiko has quit IRC | 11:49 | |
| *** radsy has quit IRC | 11:49 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 11:56 | |
| *** matsuhashi has quit IRC | 12:00 | |
| *** matsuhashi has joined #openstack-ironic | 12:00 | |
| *** NobodyCam has quit IRC | 12:02 | |
| *** matsuhashi has quit IRC | 12:02 | |
| *** NobodyCam has joined #openstack-ironic | 12:02 | |
| *** matsuhashi has joined #openstack-ironic | 12:02 | |
| *** vinbs has quit IRC | 12:05 | |
| *** NobodyCam has quit IRC | 12:08 | |
| *** NobodyCam has joined #openstack-ironic | 12:09 | |
| *** matsuhashi has quit IRC | 12:12 | |
| *** matsuhashi has joined #openstack-ironic | 12:13 | |
| *** Alexei_9871 has joined #openstack-ironic | 12:13 | |
| *** matsuhashi has quit IRC | 12:13 | |
| *** lucas-hungry is now known as lucasagomes | 12:17 | |
| *** rakesh_hs has quit IRC | 12:20 | |
| *** jdob has joined #openstack-ironic | 12:20 | |
| *** openstackgerrit has quit IRC | 12:22 | |
| *** openstackgerrit has joined #openstack-ironic | 12:23 | |
| *** jbjohnso has joined #openstack-ironic | 12:26 | |
| *** dtantsur|lunch is now known as dtantsur | 12:27 | |
| Shrews | morning | 12:39 |
| *** nosnos has quit IRC | 12:40 | |
| *** k4n0 has quit IRC | 12:42 | |
| dtantsur | morning, Shrews | 12:44 |
| Shrews | o/ | 12:45 |
| lucasagomes | morning Shrews | 12:52 |
| lucasagomes | any updates on mounting the ephemeral partition ? | 12:52 |
| Shrews | lucasagomes: Ng tested the latest patch that just duplicates nova-bm (using /dev/sda1) and it works. | 12:54 |
| NobodyCam | good mornig ironic | 12:54 |
| Shrews | lucasagomes: i'm fixing the unit tests now so we can get that pushed through for a short-term fix | 12:55 |
| lucasagomes | Shrews, nice! yeah | 12:57 |
| lucasagomes | NobodyCam, morning | 12:57 |
| romcheg | Morning NobodyCam Shrews and everyone else who only have morning when in Europe we are about to have Friday's beer :) | 12:57 |
| NobodyCam | :) really bad internerts today | 12:57 |
| *** igordcard has quit IRC | 12:57 | |
| NobodyCam | morning lucasagomes & romcheg | 12:58 |
| NobodyCam | and Shrews | 12:58 |
| *** igordcard has joined #openstack-ironic | 12:58 | |
| Shrews | romcheg: i am still recovering from thursday night's beer! | 12:58 |
| Shrews | hey NobodyCam | 12:58 |
| NobodyCam | wow that must have been some beer | 12:59 |
| Shrews | hehe. met up with fungi to celebrate a little bit :) | 12:59 |
| lucasagomes | heh nice | 13:02 |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 13:04 |
| dtantsur | morning, NobodyCam | 13:06 |
| NobodyCam | good morning dtantsur | 13:08 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Add delay to ipmi commands https://review.openstack.org/96558 | 13:10 |
| jbjohnso | so eventlet.monkey_patch() breaks a lot of stuff... surprised people survive it... | 13:19 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Wipe any metadata from a nodes disk on first deploy https://review.openstack.org/93133 | 13:19 |
| NobodyCam | anybody have a few free minutes to review ^^^^ | 13:19 |
| NobodyCam | morning jbjohnso | 13:19 |
| * romcheg is looking | 13:19 | |
| jbjohnso | NobodyCam, morning... | 13:20 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: PXE driver to validate the requested image in Glance https://review.openstack.org/96710 | 13:20 |
| jbjohnso | anyway, fyi, as it stands, pyghmi.command.session.os = eventlet.patcher.original('os') at least unbreaks that... | 13:20 |
| lucasagomes | NobodyCam, yes will do | 13:21 |
| NobodyCam | doh fogot the new conf on 96498 | 13:21 |
| *** datajerk has quit IRC | 13:22 | |
| *** sysexit has joined #openstack-ironic | 13:23 | |
| lucasagomes | dtantsur, ew, the "Ignored by this driver" was a bad copy/paste, that field is ignored by some methods and I just copy and pasted :( | 13:25 |
| NobodyCam | and of course generate smaple dosen't like my mac | 13:25 |
| dtantsur | :) | 13:25 |
| *** shausy has quit IRC | 13:26 | |
| openstackgerrit | Aleksandr Gordeev proposed a change to openstack/ironic-python-agent: Introduce MainlineHardwareManager https://review.openstack.org/92847 | 13:33 |
| *** matty_dubs|gone is now known as matty_dubs | 13:37 | |
| *** jdob has quit IRC | 14:00 | |
| romcheg | NobodyCam: I have a question about 93133 | 14:01 |
| NobodyCam | sure | 14:05 |
| romcheg | NobodyCam: Does mkfs change the partition table? | 14:05 |
| romcheg | NobodyCam: I'm aware of the case when preserve_ephemeral is True but sizes of root and swap partitions have different sizes | 14:06 |
| romcheg | NobodyCam: Not 100% sure how that works | 14:06 |
| NobodyCam | mkfs will not change the partition table | 14:07 |
| romcheg | What concerns me is L273 here https://review.openstack.org/#/c/93133/44/ironic/drivers/modules/deploy_utils.py | 14:07 |
| NobodyCam | I could check the sizes and fail if they are different | 14:09 |
| romcheg | Should it fail in that case? | 14:09 |
| NobodyCam | ? how do you see that case being handled? | 14:09 |
| romcheg | Well, fail is the easiest case I think | 14:10 |
| romcheg | *option | 14:10 |
| NobodyCam | I guess I could wipe partitions >1 and recreate | 14:11 |
| Shrews | lifeless: lucasagomes: https://review.openstack.org/96498 is good-to-go to unblock triple0 | 14:11 |
| Shrews | tripleO, even | 14:11 |
| romcheg | NobodyCam: If someone decided to make swap 2 gigs bigger and root 2 gigs smaller, it could be handled, however. | 14:12 |
| *** rwsu has joined #openstack-ironic | 14:12 | |
| devananda | 'away | 14:20 |
| devananda | g'morning, all | 14:20 |
| NobodyCam | good morning devananda | 14:21 |
| romcheg | Morning devananda! | 14:21 |
| *** shakamunyi has joined #openstack-ironic | 14:22 | |
| *** jcoufal has quit IRC | 14:25 | |
| NobodyCam | romcheg: so I understand if --preserve_ephemeral is set on rebuild then wipe swap and root partitions and recreate them. | 14:26 |
| lucasagomes | devananda, morinig | 14:26 |
| lucasagomes | morning* | 14:26 |
| * devananda reads scrollback | 14:26 | |
| *** lazy_prince has quit IRC | 14:26 | |
| NobodyCam | if that is correct could you please post that to the review. | 14:26 |
| devananda | lucasagomes: the lack of those properties on a nova flavor is, IMHO, just a doc bug -- at least as far as icehouse is concerned | 14:26 |
| romcheg | NobodyCam: That sounds reasonable. At least we should not trust external data and verify it | 14:26 |
| lucasagomes | devananda, not in flavor, it's in glance | 14:27 |
| devananda | lucasagomes: er, right. /me hasn't had coffee | 14:27 |
| lucasagomes | devananda, that's not the deploy k&r they r the instance image k&r | 14:27 |
| *** datajerk has joined #openstack-ironic | 14:27 | |
| lucasagomes | devananda, but yeah, it's doc problem as well | 14:27 |
| lucasagomes | would be good to have a sanity check and a better error message than "KeyError: kernel_id" | 14:27 |
| NobodyCam | brb | 14:28 |
| *** romcheg has left #openstack-ironic | 14:28 | |
| lucasagomes | dtantsur, ping re unittest decorator for the firewall tihngs | 14:28 |
| *** romcheg has joined #openstack-ironic | 14:28 | |
| dtantsur | lucasagomes, here | 14:28 |
| lucasagomes | dtantsur, so I could mock the class, mock.patch.object(firewall.FirewallDriver, <method name>) | 14:28 |
| lucasagomes | as a decorator | 14:28 |
| devananda | romcheg: root/swap size shouldn't change when doing rebuild because the flavor doesn't change (only the image) | 14:29 |
| dtantsur | lucasagomes, I guess you can. But anyway you'd better not use .start() for this :) | 14:29 |
| lucasagomes | but idk if it works, cause when I check if it always fail with: Expected to be called once. Called 0 times. | 14:29 |
| lucasagomes | dtantsur, right, so mocking the object (instance) works | 14:29 |
| lucasagomes | but the class directly doesn't | 14:30 |
| lucasagomes | dtantsur, use with then? | 14:30 |
| dtantsur | lucasagomes, I would really want to know, why class does not work in this case, but `with` should be ok | 14:30 |
| romcheg | devananda: That's true if Ironic is used by Nova. However, we plan to allow use it alone, don't we? | 14:30 |
| lucasagomes | dtantsur, http://paste.openstack.org/show/82188/ | 14:31 |
| lucasagomes | dtantsur, will try some more with the class see if I can figure out | 14:31 |
| dtantsur | oh, btw, morning, devananda! | 14:31 |
| lucasagomes | dtantsur, just checking if u already have any idea of how to make it work with the class, or if I'm doing something wrong | 14:32 |
| devananda | romcheg: fair point. so we should fail if root size has changed -- except, it *will* change because of growroot | 14:32 |
| dtantsur | lucasagomes, should work... I don't know this code well, I have to admit. | 14:33 |
| dtantsur | Os if you fail with class, got with `with` :) | 14:33 |
| lucasagomes | dtantsur, ack | 14:33 |
| NobodyCam | devananda: is that ^^ in regard to 93133 | 14:34 |
| lucasagomes | dtantsur, cheers | 14:34 |
| *** comstud is now known as bearhands | 14:34 | |
| devananda | lucasagomes, NobodyCam, dtantsur, romcheg - if you might be able to attend the ironic midcycle in Portland in July, please go vote on exact dates here: https://docs.google.com/forms/d/1ws6iezBwQHvvEP5_9lOI2MclJOLFi75XwbPfji4ea_M/viewform | 14:35 |
| NobodyCam | portland a lot easier then Raleigh | 14:36 |
| lucasagomes | devananda, ack, so that's already decided that we are doing with nova and not with tripleo etc...? | 14:36 |
| *** openstackgerrit has quit IRC | 14:36 | |
| dtantsur | devananda, oh... I'm afraid, I'll have troubles leaving EU in July, I'll probably be on temporary visa with my residence permit being on prolongation :( | 14:36 |
| *** openstackgerrit has joined #openstack-ironic | 14:37 | |
| romcheg | dtantsur: Isn't it possible to make a multi-entry visa? | 14:38 |
| NobodyCam | bbt..brb | 14:38 |
| lucasagomes | I just can't before OSCON, I'm planning to get some holidays in july till 15th | 14:39 |
| devananda | other folks who will try to make the midcycle, also please feel free to vote ^ -- Shrews, adam_g, SpamapS, GheRivero, russel_h, JayF, JoshNang, yuriyz, Ng, matty_dubs, jroll | 14:39 |
| dtantsur | romcheg, it's what I have, but I'll be without it in July, while my new visa application will be considered | 14:39 |
| romcheg | dtantsur: :( | 14:39 |
| romcheg | I will be moving to the EU this summer, but do not know exact dates yet. | 14:40 |
| * dtantsur is so jealous of folks who don't have to go through all these visa BS | 14:40 | |
| devananda | lucasagomes: I think, given our goal of landing the nova virt driver, that it's better for us to colocate with nova this time | 14:40 |
| matty_dubs | Hmm, so it looks like we're maybe going to have a fractured meetup? | 14:40 |
| Shrews | dtantsur: So, on 96498, in the tests, there are lots of calls to mock start/stop calls. If we were to eventually replace all of those with context managers, that would be *lots* of indenting. How would you suggest handling that? | 14:40 |
| lucasagomes | devananda, I agree | 14:40 |
| devananda | but it also sounds like those dates in general don't work for a bunch of you :( | 14:40 |
| bearhands | agree | 14:40 |
| dtantsur | Shrews, using @decorators, that's what Mikhail_D_ltp does in his patches | 14:40 |
| *** dhellmann_ is now known as dhellmann | 14:40 | |
| NobodyCam | aug would be better for /me | 14:40 |
| bearhands | although that might mean i'm in the nova stuff the whole time | 14:40 |
| dtantsur | +1 to aug | 14:40 |
| Shrews | dtantsur: i can't seem to make the decorators work for my changes (not a mock expert). care to give me an example? | 14:41 |
| devananda | so the problem with august is that it's essentially too late | 14:41 |
| devananda | FF is at the end of august | 14:41 |
| *** romcheg has left #openstack-ironic | 14:41 | |
| bearhands | yeah. | 14:41 |
| * Ng is starting to wonder if there should be a single, all-projects mid-cycle sprint | 14:41 | |
| devananda | meeting any later than the first week of august is really too late, unless it's a doc meeting ... | 14:42 |
| bearhands | nova's mid-cycle isn't a sprint | 14:42 |
| bearhands | but actual more design-summit style | 14:42 |
| bearhands | real agenda of topics to discuss | 14:42 |
| bearhands | but we just discuss them until we're done | 14:42 |
| dtantsur | Shrews, you can mock class, not instance | 14:42 |
| bearhands | instead of having a time limit like at the real conf. | 14:42 |
| dtantsur | Shrews, if that does not work for you, you can use context managers for this particular case, we'll fix it globally afterwards | 14:43 |
| dtantsur | Shrews, see e.g. https://review.openstack.org/#/c/86536/ | 14:43 |
| devananda | bearhands: I think we spent most of the last midcycle doing design discussions and hacking on devstack/tempest things (at least I did) | 14:43 |
| bearhands | that might make the colocation less useful.. although there certainly could be an ironic topic. | 14:43 |
| bearhands | yeah, ours was basically all design | 14:43 |
| bearhands | ours = nova | 14:44 |
| bearhands | i'd propose an ironic topic at the nova midcycle meetup | 14:44 |
| devananda | bearhands: gotcha. in as much as we need to continue addressing the federated hypervisor problem, I think a few of us need to be there | 14:45 |
| bearhands | that would mean we would essentially talk about "how do we get it merged" | 14:45 |
| bearhands | and we talk about it until we have it figured out | 14:45 |
| bearhands | nod | 14:45 |
| devananda | bearhands: so we already have a plan for that :) | 14:45 |
| bearhands | ok | 14:45 |
| devananda | bearhands: it actually just needs to happen now (and not get changed half way through, again) | 14:45 |
| *** romcheg has joined #openstack-ironic | 14:45 | |
| bearhands | well, if anything comes up and there's shit in the way :) | 14:45 |
| devananda | right | 14:45 |
| romcheg | Gah, router went down | 14:46 |
| bearhands | fyi, i'm getting back to my ironic cleanups here | 14:46 |
| bearhands | i went back to nova to cleanup some object stuff there | 14:46 |
| bearhands | so I can work on getting the base stuff into oslo | 14:47 |
| bearhands | so I can pull it into ironic | 14:47 |
| bearhands | heh | 14:47 |
| devananda | bearhands: so it sounds like a few representative cores from ironic would be helpful for any planning/unblocking discussions | 14:47 |
| devananda | great. i'd love to see that oslo'ified at some point | 14:47 |
| bearhands | yeah, i have some cleanups in ironic itself coming yet | 14:47 |
| devananda | awesome | 14:47 |
| bearhands | that are outside oslo, but.. oslo stuff will be coming soon too, i hope | 14:48 |
| bearhands | also been a little tied up with some rackspace stuff | 14:48 |
| bearhands | and i took a mini vacation :) | 14:48 |
| devananda | lucasagomes, dtantsur, romcheg - if you guys can't make the nova midcycle, would it be helpful to have a sprint in EU in august? | 14:48 |
| dtantsur | devananda, sounds cool to me | 14:49 |
| lucasagomes | devananda, I probably can make it (just not before OSCON), but having a sprint in EU would be great as well! | 14:49 |
| romcheg | devananda: EU is always easier for me because I can do it without my company | 14:49 |
| romcheg | devananda: As I told I'm planning to move to EU with my partner this summer but I do not know exact dates. | 14:50 |
| romcheg | I hope the sprint/midcycle won't be conflicting with the relocation | 14:51 |
| devananda | Shrews: how's EU the first week of August for you? | 14:51 |
| lucasagomes | NobodyCam, ^ | 14:52 |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 14:52 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Implement security groups and firewall filtering methods https://review.openstack.org/96466 | 14:52 |
| devananda | lucasagomes: if nova meeting is after OSCON, it's less likely I'll be able to make it | 14:52 |
| lucasagomes | devananda, oh :/ hmm | 14:52 |
| devananda | also, i'll be offline the last few weeks of August | 14:53 |
| lucasagomes | devananda, even directly after it? | 14:53 |
| devananda | probably 8/18 - 9/3 | 14:53 |
| lucasagomes | I see | 14:53 |
| GheRivero | morning all | 14:53 |
| dtantsur | morning GheRivero | 14:53 |
| lucasagomes | yeah the one in EU then sounds pretty good to me | 14:53 |
| romcheg | Good morning GheRivero! | 14:53 |
| devananda | lucasagomes: the weekend right after OSCON is a possibility -- 7/25 - 7/27 | 14:53 |
| romcheg | +1 for the EU | 14:53 |
| devananda | lucasagomes: but later than that, no | 14:53 |
| NobodyCam | I need to check on dates | 14:53 |
| NobodyCam | morning GheRivero | 14:54 |
| lucasagomes | devananda, gotcha | 14:54 |
| lucasagomes | devananda, my problem is from 9 to 15 of july | 14:54 |
| devananda | lucasagomes: ah :( | 14:54 |
| devananda | lucasagomes: 7/14 - 7/16 are the ideal dates for me. I'll be in portland anyway from 7/18 - 7/24 | 14:55 |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 14:55 |
| lucasagomes | devananda, yeah if it's decided the dates are going to collide I will try to sort it our somehow | 14:56 |
| lucasagomes | shift my holidays to another date or so | 14:56 |
| NobodyCam | romcheg: can you add coment to 93133 of how you'd like to that case handled? I like it there so others can see it | 14:56 |
| lucasagomes | problem is that I'm planning something together with a friend so it's not that easy | 14:56 |
| lucasagomes | but anyway, workable | 14:56 |
| lucasagomes | s/our/out | 14:56 |
| romcheg | devananda: Will it be possible to make an invitation(s) to the sprint to make getting a visa easier? | 14:56 |
| Shrews | dtantsur: you are too fast | 14:57 |
| Shrews | :) | 14:57 |
| dtantsur | :) | 14:57 |
| romcheg | NobodyCam: sure! | 14:57 |
| lucasagomes | romcheg, the visa u got is not valid for some years? | 14:58 |
| romcheg | lucasagomes: I mean the sprint in the EU | 14:58 |
| lucasagomes | ah sorry | 14:58 |
| romcheg | lucasagomes: I have no multi-entry Shengen visas | 14:58 |
| romcheg | Schengen even | 14:58 |
| romcheg | yuriyz neither | 14:59 |
| lucasagomes | gotcha | 14:59 |
| lucasagomes | idk how it works, but I can check with RH about those invitation letters if needed as well | 15:00 |
| romcheg | That depends on the place we are going to have that sprint | 15:00 |
| lucasagomes | yeah | 15:00 |
| lucasagomes | (will vote for Ireland :P) | 15:01 |
| romcheg | lucasagomes: Lviv :-P | 15:01 |
| *** igordcard_afk has quit IRC | 15:01 | |
| lucasagomes | heh | 15:01 |
| NobodyCam | yuriyz: I dont get the attempt always <3 comment on 96558 the next line is attempt = attempt+1 ??? | 15:02 |
| dtantsur | lucasagomes, afaik RedHat should be an organizer, to send a letter | 15:04 |
| dtantsur | romcheg, what I did with EuroPython is getting tourist visa. I specially confirmed with italian embassy in Moscow | 15:04 |
| lucasagomes | dtantsur, yeah :/ we gotta sort out the venue yet | 15:05 |
| ifarkas | devananda, ping | 15:05 |
| matty_dubs | Is Cancun on the table? | 15:05 |
| lucasagomes | lol | 15:05 |
| devananda | ifarkas: pong, sortya | 15:05 |
| dtantsur | romcheg, I mean, from their perspective, chatting with folks on IT is not a business trip :) | 15:05 |
| yuriyz | NobodyCam, if attempt 3 after l165 we raise (l171) or break (l173) | 15:05 |
| * devananda goes semi-afk to get coffee | 15:05 | |
| ifarkas | devananda, hey, dtantsur mentioned you were asking if I am working on https://bugs.launchpad.net/ironic/+bug/1314961 | 15:05 |
| devananda | ifarkas: hi! yep -- it's a High pri bug, with no activity in since you assigned it to yourself | 15:06 |
| dtantsur | lucasagomes, -1 for Ireland, sorry. It's not Schengen | 15:06 |
| ifarkas | devananda, I got distracted with something else in the meantime but I am back on it, so I am definitely working on that bug | 15:06 |
| lucasagomes | dtantsur, :( | 15:06 |
| ifarkas | devananda, I will finish it soon | 15:06 |
| romcheg | dtantsur: I have several embassies here. Invitation makes it easier - you don't have to have all the confirmations from the hotels (some embassies do not acceps booking.com's confirmations) | 15:06 |
| devananda | ifarkas: ok, thanks | 15:06 |
| NobodyCam | ahh the break is if the command worked | 15:06 |
| ifarkas | devananda, np | 15:06 |
| yuriyz | and we dont return to l164 after attempt == 3 | 15:06 |
| ifarkas | devananda, btw, I am also interested in a sprint in the EU | 15:07 |
| devananda | yuriyz: if exception AND attempt == 3: raise (break out of loop) | 15:07 |
| dtantsur | romcheg, not accepting booking.com is insane, never saw it Oo | 15:07 |
| devananda | yuriyz: if no exception AND attempt >= 3, leave while loop anyway | 15:07 |
| *** dwalleck has joined #openstack-ironic | 15:07 | |
| devananda | ok - /me really afk for coffee now | 15:08 |
| NobodyCam | :) | 15:08 |
| yuriyz | devananda, ok | 15:08 |
| dtantsur | romcheg, can't speak for every country, but business visas mostly require accomodation as well | 15:08 |
| dtantsur | re meet-up, I would vote for Brno (suddenly :), Prague, Vienna, Bratislava - these are the closest to me. But any Schengen is more or less ok, I guess | 15:16 |
| lucasagomes | +1 Vienna | 15:20 |
| romcheg | +1 Vienna, Amsterdam, Prague, Oslo :) | 15:20 |
| lucasagomes | maybe we should put Amsterdam on the table as well, there's many international flights going there | 15:20 |
| lucasagomes | I mean, it's easier to ppl to get a directly flight to ams | 15:21 |
| * jroll wants to go to europe | 15:22 | |
| jroll | devananda: thanks for the survey about portland | 15:22 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Nova Ironic Driver to not get deploy r&k from flavor https://review.openstack.org/95551 | 15:22 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add migration script to instance_info https://review.openstack.org/96136 | 15:22 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Move PXE instance level parameters to instance_info https://review.openstack.org/94855 | 15:23 |
| NobodyCam | brb | 15:26 |
| dtantsur | lucasagomes, yeah, VIE and AMS should be easy ones | 15:28 |
| *** martyntaylor has left #openstack-ironic | 15:29 | |
| lucasagomes | cool, I'm grand with any of those | 15:29 |
| romcheg | I'd actually rather make a meet up somewhere in a village Switzerland in the mountains but that sounds surrealistic :) | 15:30 |
| dtantsur | romcheg, there should be one nice place near the lake in Austria. Never been there, but a lot of people like :) | 15:31 |
| dtantsur | romcheg, this: http://en.wikipedia.org/wiki/Klagenfurt | 15:31 |
| dtantsur | ah, maybe now, one moment | 15:32 |
| dtantsur | now = no | 15:32 |
| jroll | romcheg++ | 15:32 |
| dtantsur | ok, it's called "klopeiner see" and I can't find any reasonable link for it >_< | 15:33 |
| romcheg | dtantsur: Returning from Interlaken was feeling like returning to the place where absolutely nothing is good :) | 15:34 |
| matty_dubs | I dunno, I like the idea of us meeting at one of the "coffee" shops in Amsterdam to write code. And then afterwards, looking back at what in the world we created. | 15:34 |
| lucasagomes | I would like to propose to make it in brazil and not eu, we can make it in a place called lençois maranhenses: http://goo.gl/Wu4wvh | 15:34 |
| lucasagomes | jk :) | 15:35 |
| dtantsur | nice! | 15:35 |
| lucasagomes | matty_dubs, lol hah yeah | 15:35 |
| dtantsur | matty_dubs, we'll get the driver merged easily afterwards! | 15:35 |
| matty_dubs | lucasagomes: I know you have helped me with this about 47 times already, but remind me what I need to delete to fix "ImportError: cannot import name messaging" ? | 15:35 |
| lucasagomes | matty_dubs, heh rm -rf ironic/openstack/common/rpc | 15:36 |
| matty_dubs | Ooh, wait, that doesn't exist. Something else must be wrong. | 15:37 |
| lucasagomes | matty_dubs, is olso messaging installed? | 15:37 |
| lucasagomes | pip install oslo.messaging | 15:37 |
| matty_dubs | Ha! | 15:38 |
| matty_dubs | I'm circling back to an older dev box, and it's clearly missing some of this stuff. | 15:39 |
| matty_dubs | OK, that was it. I was convinced it was the "I need to delete a directory, but I forget which" problem. I'm in business now; thanks! | 15:39 |
| lucasagomes | :D | 15:40 |
| *** lsmola has quit IRC | 15:41 | |
| *** derekh_ has quit IRC | 15:46 | |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Add delay to ipmi commands https://review.openstack.org/96558 | 15:50 |
| *** eghobo has joined #openstack-ironic | 15:50 | |
| NobodyCam | brb | 15:51 |
| *** matty_dubs is now known as matty_dubs|lunch | 15:57 | |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: PXE to pass hints to ImageCache on how much space to reclaim https://review.openstack.org/94371 | 15:59 |
| *** bandicot has joined #openstack-ironic | 16:01 | |
| *** eghobo has quit IRC | 16:03 | |
| *** eghobo has joined #openstack-ironic | 16:03 | |
| openstackgerrit | Stig Telfer proposed a change to openstack/ironic-specs: BP spec: ironic-snmp-power-driver https://review.openstack.org/96816 | 16:05 |
| NobodyCam | oh that sounds neat-o ^^ | 16:08 |
| *** datajerk has quit IRC | 16:12 | |
| *** dwalleck has quit IRC | 16:14 | |
| *** ellenh has joined #openstack-ironic | 16:18 | |
| *** shakamunyi has quit IRC | 16:20 | |
| *** ifarkas has quit IRC | 16:22 | |
| openstackgerrit | A change was merged to openstack/ironic: Pass kwargs to ClientWrapper's call() method https://review.openstack.org/92384 | 16:22 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/96228 | 16:26 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/88722 | 16:26 |
| *** bandicot has quit IRC | 16:29 | |
| lucasagomes | NobodyCam, giving a heads up, keystoneclient updated some stuff in their config options | 16:29 |
| lucasagomes | NobodyCam, https://review.openstack.org/#/c/96694/ | 16:29 |
| lucasagomes | devananda, ^ | 16:30 |
| *** dwalleck has joined #openstack-ironic | 16:30 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/96263 | 16:32 |
| *** jrist has quit IRC | 16:32 | |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Add delay for ipmitool commands https://review.openstack.org/96558 | 16:35 |
| *** bandicot has joined #openstack-ironic | 16:36 | |
| NobodyCam | lucasagomes: is that for items that https://github.com/openstack/ironic/commit/6901cbfd7113f770d0b32a99131411fe15dc3d56 didn't cover? | 16:38 |
| lucasagomes | NobodyCam, oh, hmm not sure lemme check | 16:40 |
| lucasagomes | maybe it's failing only for me locally due the version of keystoneclient? | 16:40 |
| lucasagomes | lemme check | 16:40 |
| lucasagomes | I thought it would be a pip mirroring, that's the reason it wasn't failing in gate | 16:40 |
| openstackgerrit | Stig Telfer proposed a change to openstack/ironic-specs: Blueprint ironic-snmp-power-driver: Power driver for SNMP-enabled smart PDUs https://review.openstack.org/96838 | 16:43 |
| NobodyCam | brb | 16:44 |
| *** bandicot has quit IRC | 16:46 | |
| lucasagomes | NobodyCam devananda , the config was due my env, abandoning | 16:53 |
| *** zdiN0bot has joined #openstack-ironic | 16:55 | |
| *** jrist has joined #openstack-ironic | 16:55 | |
| *** rameshg87 has joined #openstack-ironic | 16:56 | |
| lucasagomes | alright i will call it a day | 16:57 |
| lucasagomes | have a good night everybody | 16:57 |
| lucasagomes | enjoy the weekend | 16:57 |
| lucasagomes | monday is holiday for me, but I will try to make it for the meeting | 16:57 |
| *** lucasagomes is now known as lucas-dinner | 16:57 | |
| *** romcheg has quit IRC | 16:59 | |
| *** jistr has quit IRC | 16:59 | |
| *** matty_dubs|lunch is now known as matty_dubs | 17:00 | |
| *** jdob has joined #openstack-ironic | 17:03 | |
| *** harlowja_away is now known as harlowja_ | 17:15 | |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 17:20 |
| Shrews | dtantsur: good idea on skipping the save(). ^^^ that addresses your comments | 17:23 |
| dtantsur | good! | 17:23 |
| *** martyntaylor has joined #openstack-ironic | 17:25 | |
| *** rameshg87 has quit IRC | 17:25 | |
| dtantsur | Shrews, left a couple more comments | 17:27 |
| *** max_lobur has quit IRC | 17:28 | |
| *** dwalleck has quit IRC | 17:28 | |
| Shrews | dtantsur: save() is on the instance | 17:29 |
| Shrews | not client | 17:29 |
| dtantsur | Shrews, oops, sorry, it's Friday evening :) | 17:29 |
| dtantsur | disregard that comment | 17:29 |
| *** datajerk has joined #openstack-ironic | 17:31 | |
| *** eghobo has quit IRC | 17:31 | |
| Shrews | dtantsur: i'll change it, but just curious... why does assertEqual() order matter? | 17:31 |
| dtantsur | Shrews, better error message | 17:31 |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 17:31 |
| Shrews | ah | 17:31 |
| dtantsur | it will say smth like "expected {0}, got {1}" | 17:31 |
| dtantsur | IIRC | 17:31 |
| dtantsur | Shrews, reviewed | 17:34 |
| dtantsur | ... and now I guess I can finally call it a day! | 17:34 |
| Shrews | dtantsur: thanks! have a great weekend | 17:35 |
| dtantsur | g'night everyone! :) Have a nice weekend | 17:35 |
| NobodyCam | night dtantsur | 17:36 |
| *** dtantsur is now known as dtantsur|afk | 17:37 | |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Add delay for ipmitool commands https://review.openstack.org/96558 | 17:41 |
| NobodyCam | oh htats unique our conf file must have two blank lines at the end :-p | 17:42 |
| *** dwalleck has joined #openstack-ironic | 17:43 | |
| *** ndipanov has quit IRC | 17:43 | |
| *** sseago__ has quit IRC | 17:44 | |
| Shrews | must it? | 17:44 |
| Shrews | how very odd | 17:44 |
| NobodyCam | :-p | 17:45 |
| *** dwalleck has quit IRC | 17:47 | |
| *** martyntaylor has left #openstack-ironic | 17:52 | |
| openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Work around iscsiadm delete failures https://review.openstack.org/96577 | 17:57 |
| NobodyCam | lol 93133 back to the the starting point of how to remove a partition when parted is failing thinking its mounted when its not! | 17:59 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 18:01 |
| *** sseago has joined #openstack-ironic | 18:06 | |
| openstackgerrit | Stig Telfer proposed a change to openstack/ironic-specs: Power driver for SNMP-enabled smart PDUs https://review.openstack.org/96838 | 18:08 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding swift temp url support https://review.openstack.org/81391 | 18:09 |
| devananda | Shrews: left a comment on https://review.openstack.org/#/c/96498/8/ironic/nova/virt/ironic/driver.py | 18:18 |
| Shrews | ack | 18:18 |
| devananda | NobodyCam: on https://review.openstack.org/#/c/96558/6, I would prefer the options be done differently | 18:20 |
| *** Mikhail_D_ltp has quit IRC | 18:20 | |
| NobodyCam | sure | 18:20 |
| devananda | NobodyCam: see ironic/drivers/modules/ipminative.py -- there's already an [ipmi] group with a retry_timeout setting | 18:20 |
| devananda | NobodyCam: so we're only missing is retry_interval | 18:22 |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Set instance default_ephemeral_device https://review.openstack.org/96498 | 18:23 |
| NobodyCam | ok | 18:24 |
| devananda | NobodyCam: btw, have you looked into ipmitool docs before ? | 18:27 |
| NobodyCam | devananda: interval or attempts? | 18:27 |
| mgagne | I installed ironic with devstack a couple hours ago. I tried to delete an instance and I'm getting this error in nova-compute: ValueError: Circular reference detected | 18:27 |
| devananda | NobodyCam: is there a reason not to use "-N $sec -R $retries" options? | 18:27 |
| mgagne | Can someone help me debug this? Where should I start? | 18:27 |
| mgagne | And I also have this one in ironic-api: MultipleResultsFound: Multiple rows were found for one() | 18:28 |
| NobodyCam | other then we have always used the one attempt per command and handled the looping with in python | 18:29 |
| devananda | mgagne: that ^ indicates there is duplication of records in ironic's database. something created the same instance more than once, would be my guess. can you unstack and restart devstack? | 18:29 |
| devananda | NobodyCam: right -- perhaps ipmitool is smarter than we are :) | 18:29 |
| mgagne | devananda: ok, will try. My logs, nova list and ironic node-list: http://paste.openstack.org/show/82218/ | 18:30 |
| devananda | mgagne: yea - the problem is you have 3 nodes in ironic, all with the same instance_uuid | 18:31 |
| mgagne | devananda: right, how can I make it never happen in the future? | 18:31 |
| NobodyCam | devananda: want is you ipmi tool version | 18:31 |
| * NobodyCam mac does not have those options: (py27)nobodycamair:ironic NobodyCam$ ipmitool -V | 18:32 | |
| NobodyCam | ipmitool version 2.5b1 | 18:32 |
| NobodyCam | s/you/your/ | 18:32 |
| devananda | NobodyCam: ipmitool version 1.8.13 | 18:32 |
| NobodyCam | lol | 18:33 |
| devananda | mgagne: i'm not sure /how/ it happened, so i can't tell you how to prevent it yet. if you can reproduce it repeatedly, that would help | 18:34 |
| mgagne | devananda: as you could have guessed, I'm new to Ironic and today is my first experience with it. Will try to reproduce if I can =) | 18:34 |
| devananda | mgagne: ack. I haven't seen that error before so unfortunately i can't help a whole lot more than "huh, I'd love to know how that happened so we can fix it" :) | 18:35 |
| devananda | NobodyCam: so "man ipmitool" doesn't have any options like | 18:36 |
| devananda | -N <sec> | 18:36 |
| devananda | Specify nr. of seconds between retransmissions of lan/lanplus messages. Defaults are 2 seconds for lan and 1 second for lanplus interfaces. | 18:36 |
| devananda | ? | 18:36 |
| devananda | *on your mac | 18:36 |
| mgagne | devananda: right, we are diving headfirst in ironic next week, hopefully we will be able to provide feedbacks and bug fixes :) | 18:37 |
| NobodyCam | I have no option for -N or -R | 18:37 |
| NobodyCam | wnat me to paste you my manpage | 18:37 |
| devananda | NobodyCam: um, no. the ipmitool manpage is HUGE | 18:38 |
| NobodyCam | lol :-p | 18:38 |
| devananda | they might not be visible in --help | 18:38 |
| devananda | NobodyCam: and it might not be visible in the manpage as a top-level option | 18:38 |
| devananda | NobodyCam: or mac may just have a terrible implementation | 18:39 |
| NobodyCam | (py27)nobodycamair:ironic NobodyCam$ man ipmitool |grep '\-m' -m <local_address> | 18:39 |
| NobodyCam | (py27)nobodycamair:ironic NobodyCam$ man ipmitool |grep '\-N' | 18:39 |
| NobodyCam | (py27)nobodycamair:ironic NobodyCam$ | 18:39 |
| *** sysexit has quit IRC | 18:39 | |
| devananda | :( | 18:40 |
| NobodyCam | that may be why we did out own looping | 18:40 |
| jroll | random thing: has anybody thought about having the scheduler check if a node is already assigned to an instance? | 18:40 |
| jroll | rather than just trying to schedule a node and then finding out? | 18:40 |
| jroll | and having 3 failures tell nova it couldn't schedule? | 18:41 |
| devananda | jroll: compute won't pick a node if it's already associated, and scheduler should see that node has having no available resoruces and so shouldn't pick it | 18:41 |
| devananda | there's a brief race there (same for nova-bm) | 18:41 |
| jroll | hmm | 18:42 |
| jroll | oh right, there's the exact filter thing | 18:42 |
| devananda | jroll: but the problem in mgagne's case is the scheduler picked /different/ nodes each time | 18:42 |
| devananda | and each one failed | 18:42 |
| jroll | I might not have that configured | 18:42 |
| jroll | heh, I'm talking about my own issues, not mgagne :P | 18:42 |
| devananda | jroll: ah, yea. if your flavor != node properties AND not using exact_match AND the host_manager is broken -- then you'll get scheduler picking nodes that are already in use | 18:42 |
| openstackgerrit | A change was merged to openstack/ironic: list_instances() to return a list of instances names https://review.openstack.org/92385 | 18:43 |
| jroll | "host_manager is broken" hm? | 18:43 |
| devananda | jroll: using the wrong host_manager | 18:43 |
| jroll | ah | 18:43 |
| jroll | ok | 18:43 |
| jroll | thanks deva | 18:44 |
| devananda | jroll: hth :) | 18:44 |
| devananda | jbjohnso: looking at pyghmi, i see a way to change the wait time for commands, but not a way to set the interval between retries | 18:45 |
| devananda | jbjohnso: am i missing something? | 18:45 |
| NobodyCam | devananda: on the ipminative retry_timeout this the amount of time to attempt to run command x so adding retry_interval means will run command x more than the current limit of three times, is that how your seeing it too? | 18:48 |
| devananda | NobodyCam: I think it's a ~10 line patch... gimme a moment | 18:49 |
| *** pelix has quit IRC | 18:50 | |
| JayF | devananda: btw, you may see ellenh picking up some bugs. She's working with our team this summer. Anything low hanging bugs that are interesting and good for new people might be worth throwing in her/our direction. I think she's working on https://bugs.launchpad.net/ironic/+bug/1309110 (ipa bug) now as a start | 18:59 |
| devananda | JayF: great | 18:59 |
| devananda | NobodyCam: so there's two levels to this issue, I think | 19:01 |
| devananda | NobodyCam: one is simply retrying the same command too fast | 19:01 |
| NobodyCam | ok | 19:02 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add retry delay to ipmitool commands https://review.openstack.org/96902 | 19:02 |
| devananda | NobodyCam: I think ^ will help there | 19:02 |
| devananda | NobodyCam: the other problem is that /different/ commands may be issued too quickly, even when they succeed | 19:03 |
| devananda | NobodyCam: eg, validate; power off; validate; power on; | 19:03 |
| devananda | NobodyCam: imagine if all four of those are issued -- and each one succeeds. what will be the interval between each IPMI command? | 19:04 |
| openstackgerrit | Jarrod Johnson proposed a change to stackforge/pyghmi: Fix pyghmicons script crash on idle https://review.openstack.org/96903 | 19:04 |
| devananda | lifeless, Ng: IIUC, that is also causing your problems (different IPMI commands are issued rapidly, the first (one? few?) succeed, then the BMC falls over) | 19:05 |
| NobodyCam | ok so I need to see what that dose with my ipmi | 19:06 |
| *** zdiN0bot has quit IRC | 19:06 | |
| NobodyCam | *ipmitool | 19:06 |
| openstackgerrit | A change was merged to stackforge/pyghmi: Fix pyghmicons script crash on idle https://review.openstack.org/96903 | 19:06 |
| devananda | NobodyCam: so the module-scope array you added is addressing that second case | 19:06 |
| devananda | NobodyCam: but my point is, these are separate problems. the error handling shouldn't be done in python -- it should be retried by ipmitool. the global "last time this BMC was poked" is all we need to do in python | 19:07 |
| *** Penick has joined #openstack-ironic | 19:08 | |
| devananda | because, as far as I can tell, ipmitool /already/ does that, even if you dont tell it to | 19:08 |
| devananda | with the default options, I'm seeing retransmits | 19:08 |
| devananda | at least in some really simple tests | 19:08 |
| lifeless | devananda: we don't know if its the same retried, or different ones, but the Holly folk had to make Holly leave an inter-command-gap or the same thing happened on the same hardware. | 19:10 |
| devananda | lifeless: good data point, thanks | 19:10 |
| lifeless | devananda: which reminds me | 19:10 |
| *** Alexei_9871 has quit IRC | 19:10 | |
| lifeless | devananda: next time Wan-yen asks about what would make iLO better for Ironic : more robust IPMI please :) | 19:11 |
| NobodyCam | I think my patch handled both of those cases. it tracked the last "poke" to the bmc addresse and basic rate limited the retrys reguardless of if it was the same or different commands being snet | 19:11 |
| *** Penick_ has joined #openstack-ironic | 19:11 | |
| NobodyCam | *sent even | 19:11 |
| devananda | lifeless: heh heh heh | 19:11 |
| devananda | NobodyCam: that aspect of your patch is great. my point is that the error handling shouldn't be in your patch | 19:12 |
| lifeless | devananda: sadly serious :) | 19:12 |
| devananda | NobodyCam: if you don't mind, i can rebase yours on mine and demonstrate | 19:12 |
| devananda | lifeless: i know :( | 19:12 |
| NobodyCam | but when I look my ipmitool will not support the options your adding | 19:12 |
| NobodyCam | please do | 19:12 |
| *** Penick has quit IRC | 19:12 | |
| *** Penick_ is now known as Penick | 19:12 | |
| devananda | NobodyCam: https://developer.apple.com/library/mac/documentation/Darwin/Reference/Manpages/man1/ipmitool.1.html vs http://linux.die.net/man/1/ipmitool | 19:14 |
| devananda | NobodyCam: that claims to be from the same source, but it's got a different list of options, clearly | 19:15 |
| *** ellenh has quit IRC | 19:15 | |
| devananda | NobodyCam: man page from the project source repo includes -N -C -- http://ipmitool.cvs.sourceforge.net/viewvc/ipmitool/ipmitool/doc/ipmitool.1 | 19:18 |
| devananda | NobodyCam: so apple is doing something :P | 19:18 |
| devananda | ok, i'll do a rebase on 96558 as soon as I get some more coffee :) | 19:18 |
| devananda | bbiaf | 19:18 |
| mgagne | devananda: my coworker mislead me by telling me linuxbridge was working fine with ironic in devstack. it does not =) | 19:19 |
| NobodyCam | and morning lifeless :) | 19:19 |
| mgagne | devananda: everything works fine now with openvswitch, I can ssh to my baremetal instance, something y coworker never managed to do =) | 19:20 |
| devananda | mgagne: \o/ | 19:20 |
| Ng | devananda: yep, that change seems very sensible to me | 19:24 |
| *** sysexit has joined #openstack-ironic | 19:27 | |
| NobodyCam | devananda: fyi freeBSD does not list -N or _r ... need to check how up to date that is. http://manned.org/ipmitool/e7216a6d | 19:33 |
| NobodyCam | s/_r/-R/ | 19:33 |
| *** max_lobur has joined #openstack-ironic | 19:37 | |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Add __init__.py for nova scheduler filters https://review.openstack.org/96912 | 19:40 |
| bearhands | ^^^ This is a masterpiece | 19:41 |
| bearhands | btw | 19:41 |
| jroll | nice | 19:41 |
| bearhands | haa | 19:41 |
| bearhands | it's my best work | 19:41 |
| devananda | back | 19:45 |
| devananda | bearhands: you could always just copy those files into nova's tree .... | 19:45 |
| JayF | even his bear hands aren't large enough to carry the whole driver into nova himself | 19:47 |
| bearhands | devananda: Yes, could. | 19:47 |
| bearhands | but I think they should work just by installing ironic pkg | 19:47 |
| bearhands | you have the __init__ in scheduler so that host manager loads fine | 19:48 |
| bearhands | but it's missing in the filters dir | 19:48 |
| NobodyCam | devananda: when you rebase note !!! pep8 failure ./ironic/drivers/modules/ipmitool.py:61:1: E302 expected 2 blank lines, found 1 | 19:49 |
| devananda | ack | 19:49 |
| devananda | bearhands: still have to copy the new clusteredcomputemanager class | 19:50 |
| *** blamar has quit IRC | 19:51 | |
| bearhands | hm. | 19:51 |
| bearhands | not sure i'm seeing that for some reason | 19:51 |
| bearhands | checkin | 19:51 |
| bearhands | somehow that is loading okay out of tree | 19:52 |
| devananda | mmm, breakfast is here | 19:52 |
| devananda | bearhands: rly? | 19:52 |
| bearhands | maybe we added an __init__.py | 19:52 |
| devananda | hah | 19:52 |
| bearhands | but yeah | 19:52 |
| NobodyCam | its 3pm here :-p | 19:52 |
| bearhands | no | 19:52 |
| bearhands | er | 19:52 |
| devananda | bearhands: always makes me happy when things that i thougth were broken, actually work fine | 19:53 |
| bearhands | it's upstream | 19:53 |
| bearhands | __init__.py for compute dir | 19:53 |
| bearhands | so ya, it's only missing for the scheduler filters dir | 19:53 |
| bearhands | after that... everything works completely fine being out of nova tree | 19:54 |
| NobodyCam | jbjohnso: off the top of your head does pyghmi support retry_interval | 19:58 |
| jbjohnso | NobodyCam, it can be made to... it defaults to trying to be a bit automagic.. | 19:59 |
| jbjohnso | e.g. at least I think I made it to have a pretty quick timeout on get channel auth cap, and then a long timeout for other packets since there is evidence that the connection works | 19:59 |
| NobodyCam | just thought of it as we are putting that logic into ipmitool | 20:00 |
| jbjohnso | primarily toward the goal of having SOL sessions survive a shared nic outage that is transient | 20:00 |
| jbjohnso | do you pass it to ipmitool or are you invoking ipmitool separately? | 20:00 |
| NobodyCam | s/into/around/ | 20:00 |
| NobodyCam | well actually a little of both | 20:01 |
| jbjohnso | so the default for get channel auth cap is | 20:01 |
| jbjohnso | first retry is between 0.5 and 1.0 seconds (randomly) | 20:02 |
| jbjohnso | then between 1.5 and 2.0 | 20:02 |
| jbjohnso | then between 2.5 and 3 | 20:02 |
| jbjohnso | and then give up | 20:02 |
| jbjohnso | if a packet is lost in the midst of a sessien | 20:02 |
| jbjohnso | then it is 0.5-1, then 1.5 to 2, then 2.5 to 3, then 3.5 to 4, then 4.5 to 5, then 5.5 to 6, then give up | 20:03 |
| jbjohnso | so it tries to tolerate an 18 second outage in the midst of a session | 20:05 |
| jbjohnso | much more than that and the ipmi device has probably deallocated the session anyway (by spec an ipmi session shouldn't survive the interval that would be induced by spanning tree in block-by-default for example) | 20:06 |
| NobodyCam | jbjohnso: per this bug intel says there should be a 5 second delay https://bugs.launchpad.net/ironic/+bug/1320513 | 20:07 |
| jbjohnso | well that one is specifically talking about a certain class of things | 20:08 |
| jbjohnso | e.g. it would be considered bad form to set a lan channel parameter and then be upset if retrieval of the parameter did not complete within 5 seconds | 20:08 |
| jbjohnso | of course, even 5 seconds isn't enough for that stuff in some cases | 20:09 |
| jbjohnso | now if exec is failing on the session establishment and slots are consumed | 20:10 |
| jbjohnso | then the bmc is supposed to recover after 60 seconds on it's own... | 20:10 |
| jbjohnso | though hammering a bad login is also bad form | 20:11 |
| jbjohnso | which is why bad auth data is bailed out of... | 20:11 |
| jbjohnso | e.g. in confluent, it doesn't check for credential changes in that case no sooner than a 20 second interval | 20:11 |
| jbjohnso | anyway, the particular change there would be in the driver, since pyghmi in this specific regard is like ipmitool, it tries and decides once and only once if connection is possible | 20:13 |
| jbjohnso | and it is up to the calling library to decide the right course (e.g. in confluent it checks to see if maybe someone changes the bmc later, but also hooks a config change if someone changes the password to check immediately) | 20:14 |
| jbjohnso | so an occasional polling except for an event driven check for reconfigure | 20:14 |
| jbjohnso | incidentally, one thing about ipmitool style polling is that it is a tad bit rougher on a bmc than confluent style interrogation | 20:15 |
| jbjohnso | e.g. intel bmcs that try to do an auditlog competently will have their SEL contain mountains of messages | 20:15 |
| *** tteggel_ has joined #openstack-ironic | 20:16 | |
| jbjohnso | since every ipmitool invocation is log in, do request, logout | 20:16 |
| jbjohnso | and confluent model is log in... do request......... .do request..... do request.... maybe log out one day.... repair session if broken.... | 20:16 |
| NobodyCam | :-p | 20:17 |
| jbjohnso | which incidentally takes a 'power status' command from about 14 packets to 2 | 20:17 |
| jbjohnso | and get health is harder to enumerate, but you save about 80% of the time or more depending on implementation this way | 20:18 |
| jbjohnso | I'm might be the only person in the world with this fixation on obscure facets of ipmi usage... | 20:18 |
| jroll | jbjohnso: you should start a startup to fix all the ipmi things | 20:19 |
| devananda | jbjohnso: Magic 8-ball says .... there is a distinct possibility :) | 20:19 |
| jroll | including bmc | 20:19 |
| jroll | because, as you know, it's all horrible | 20:19 |
| jbjohnso | well, it's not all bad... it could be a number of other specifications that border on uselessness out of being too vague... | 20:19 |
| devananda | NobodyCam: so unit tests fail because they're expecting a specific set of options to ipmitool ... ugh. otherwise i fixed it | 20:20 |
| russell_h | I was telling pquerna this the other day | 20:20 |
| jbjohnso | the security decision to have RAKP2 with the first use of Kuid in a HMAC is dubious | 20:20 |
| russell_h | we should all quit | 20:20 |
| jroll | jbjohnso: in practice, not in theory :P | 20:20 |
| russell_h | and build good servers | 20:20 |
| russell_h | in other words | 20:20 |
| *** SpamapS_ has joined #openstack-ironic | 20:20 | |
| russell_h | startup idea: go head-to-head with hp, dell, etc, with no discernible strategic advantage | 20:21 |
| jbjohnso | and the fact that generic discrete sensors are impossible to reliably decide if 'assertion' is 'critical', 'warning', or even if it's good, bad, or completely innocuous... | 20:21 |
| jbjohnso | and the authorization model is impossibly convoluted for anyone's needs... | 20:21 |
| *** SpamapS has quit IRC | 20:21 | |
| *** tteggel has quit IRC | 20:21 | |
| jbjohnso | and there's no way to really tell a retry from a new request... | 20:21 |
| *** SpamapS_ is now known as SpamapS | 20:21 | |
| jroll | russell_h: the advantage is everyone that actually runs these computers will beg the heck out of their bosses to buy them :P | 20:21 |
| jbjohnso | other than that though, not too shabby... | 20:21 |
| *** SpamapS has quit IRC | 20:22 | |
| *** SpamapS has joined #openstack-ironic | 20:22 | |
| matty_dubs | russell_h: LOL, I was advocating something similar here | 20:22 |
| devananda | jbjohnso: we should just use SSH and put a full linux kernel in there ;) | 20:22 |
| jbjohnso | I like the fact that '0 4 0 1' always means 'turn on the identify light, we don't care what vendor you are, this is the only meaning it can have and it must mean this' | 20:22 |
| * devananda is obviously trolling | 20:22 | |
| matty_dubs | Imagine if you could VNC to a BMC, instead of needing some wacky Java applet? | 20:22 |
| jbjohnso | matty_dubs, we actually had that once upon a time... | 20:23 |
| jbjohnso | then I'm convinced a vendor went golfing with someone... and lo, java required | 20:23 |
| matty_dubs | lol | 20:23 |
| matty_dubs | "You know what would make this better? If it required Java." | 20:24 |
| jbjohnso | I'm trying to one up that... | 20:24 |
| NobodyCam | devananda: fixed? 96902 or 96558 or even both | 20:24 |
| jbjohnso | I've gotten our servers to work with confluent out of the cardboard in the next release | 20:24 |
| devananda | NobodyCam: both. fixing unit tests now | 20:24 |
| jbjohnso | so no java, no vnc, just text... | 20:25 |
| jbjohnso | all without slowing down the people who don't like text consoles (which was the previous gripe preventing this before) | 20:25 |
| jbjohnso | anyway, graphics consoles are lame | 20:26 |
| jbjohnso | copy and paste sucks, use more bandwidth than needed | 20:26 |
| jbjohnso | even windows has EMS | 20:26 |
| jbjohnso | esxi, windows, linux, all have very serviceable serial consoles nowadays | 20:27 |
| jbjohnso | sure, you can't run notepad over text, but hey | 20:27 |
| jbjohnso | (waiting for them to add aalib support to their text console...) | 20:27 |
| jbjohnso | I wonder what notepad would look like through aalib... | 20:28 |
| *** zdiN0bot has joined #openstack-ironic | 20:28 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 20:28 | |
| jbjohnso | but if you have BMC unique passwords generated in a fashion like 'base64.b64encode(os.urandom(15))', select a quality vendor, use it mostly within a datacenter... then it is a pretty respectable thing | 20:31 |
| jbjohnso | a vendor largely steering clear of generic discrete sensors for anything trying to indicate a potentially 'bad' thing.. | 20:32 |
| *** ellenh has joined #openstack-ironic | 20:33 | |
| jbjohnso | I think that pretty much is a scenario tolerant to the criticisms... | 20:33 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Enforce a minimum time between all IPMI commands https://review.openstack.org/96558 | 20:34 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Let ipmitool natively retry commands https://review.openstack.org/96902 | 20:34 |
| devananda | NobodyCam: ^ | 20:34 |
| NobodyCam | also devananda once / and if you have a free second.. would you mind poping by 93133 | 20:35 |
| * NobodyCam looking | 20:35 | |
| jbjohnso | devananda, is that time between ipmi operations or between retries of ipmi operations? | 20:35 |
| devananda | jbjohnso: both | 20:36 |
| devananda | jbjohnso: first patch only does it for retries. second patch makes it for time between any operation | 20:36 |
| jbjohnso | 5 seconds between ipmi operations would make 'power query' take 70 seconds... | 20:36 |
| devananda | jbjohnso: not if it succeeds onthe first try | 20:36 |
| jbjohnso | well, mabe actually 35... | 20:36 |
| jbjohnso | well, one 'power query' in ipmitool is 7 transmitted packets | 20:37 |
| devananda | jbjohnso: sure. because ipmitool doesn't reuse sessions | 20:37 |
| jbjohnso | unless you mean 5 seconds between invocations of ipmitool and sequential ipmi packets within one exec are just one thing.. | 20:37 |
| devananda | right :) | 20:37 |
| jbjohnso | ok, so the 'get sensor reading' command would not get so slowed | 20:37 |
| jbjohnso | since you need a gobton of those to assess health... | 20:38 |
| devananda | jbjohnso: this is trying to solve the "we poke the BMC too quickly and it falls over" problem | 20:38 |
| jbjohnso | a metric gobton even | 20:38 |
| jbjohnso | the only solution to that is better bmc vendors ;) | 20:38 |
| devananda | jbjohnso: granted, it is a somewhat naive solution, but it gives operators a knob to tune if they know their hardware is faster/slower | 20:38 |
| jbjohnso | I have seen el cheapo bmc stuff fall over without anything seemingly happening at all | 20:39 |
| NobodyCam | devananda: my only concern is folks with wacko versions of ipmitool like /me | 20:41 |
| NobodyCam | :-p | 20:41 |
| jbjohnso | wacko version of ipmitool? | 20:41 |
| devananda | jbjohnso: yea, NobodyCam's running on a mac, which doesn't use the ipmitool from sourceforge | 20:41 |
| NobodyCam | I do not have the -N nore -R options | 20:41 |
| jbjohnso | oh, compile it then | 20:41 |
| NobodyCam | lol... deva did you see my note that bsd may have same ipmitool | 20:42 |
| jbjohnso | hmm, sles11 version doesn't have it | 20:42 |
| NobodyCam | oh that will be a issue for me | 20:42 |
| NobodyCam | lol | 20:42 |
| *** jdob has quit IRC | 20:43 | |
| jbjohnso | rhel7 has it.. | 20:43 |
| jbjohnso | and rhel6 | 20:43 |
| jbjohnso | well, rhel6 with updates... | 20:43 |
| NobodyCam | devananda: sels11 is going to hurt | 20:43 |
| devananda | jbjohnso: http://linux.die.net/man/1/ipmitool is the manpage for latest ipmitool, afaict | 20:43 |
| devananda | NobodyCam: compile it :) | 20:43 |
| NobodyCam | lol | 20:44 |
| devananda | and complain to distros | 20:44 |
| jbjohnso | well, have a good weekend | 20:44 |
| devananda | they're claiming a higher rev # than their upstream package provides | 20:44 |
| devananda | jbjohnso: you too! | 20:44 |
| NobodyCam | you too jbjohnso | 20:44 |
| jbjohnso | time to go out on the boat | 20:44 |
| *** jbjohnso has left #openstack-ironic | 20:45 | |
| *** max_lobur has quit IRC | 20:52 | |
| devananda | bbiaf | 20:53 |
| NobodyCam | devananda: -1'd 96902 for doc do you agree? | 20:53 |
| *** ellenh has quit IRC | 20:54 | |
| *** openstackgerrit has quit IRC | 21:01 | |
| *** ellenh has joined #openstack-ironic | 21:01 | |
| *** lucas-dinner has quit IRC | 21:13 | |
| *** sysexit has quit IRC | 21:18 | |
| devananda | back | 21:18 |
| NobodyCam | wb | 21:18 |
| *** Mikhail_D_ltp has left #openstack-ironic | 21:20 | |
| devananda | NobodyCam: on 93133, i continue to wonder, if destroy_disk_metadata fails, what should the behavior be? | 21:21 |
| devananda | NobodyCam: as it stands, it just logs and blindly continues | 21:21 |
| NobodyCam | I believe the orginal logic was that removing labels could fail and that was ok, but with the use of DD that noy in fact be incorrect logic | 21:23 |
| NobodyCam | s/noy/not/ | 21:24 |
| devananda | not incorrect? | 21:25 |
| NobodyCam | s/incorrect/correct/ | 21:25 |
| NobodyCam | :-p | 21:25 |
| NobodyCam | it's late on friday and I'm on vary slow bandwidth | 21:26 |
| NobodyCam | Very even | 21:26 |
| devananda | :p | 21:27 |
| NobodyCam | what are your thoughts on wiping the swap and root when rebuilding | 21:27 |
| *** tteggel_ has quit IRC | 21:29 | |
| devananda | NobodyCam: I don't think that's necessary because it's the same tenant | 21:30 |
| *** tteggel has joined #openstack-ironic | 21:30 | |
| NobodyCam | the concern was someone enlarging swap and / or shrink root or viseversa | 21:32 |
| NobodyCam | but to me thats kinda redepoly note rebuild | 21:33 |
| NobodyCam | s/note/not/ | 21:33 |
| *** jgrimm has quit IRC | 21:36 | |
| devananda | right | 21:36 |
| devananda | that's no longer the same instance | 21:36 |
| NobodyCam | ya | 21:36 |
| *** zdiN0bot has quit IRC | 21:37 | |
| NobodyCam | that's a snikey way to pass data between instances lol... | 21:37 |
| NobodyCam | snekey ?? | 21:37 |
| devananda | sneaky? | 21:37 |
| NobodyCam | :-p | 21:37 |
| devananda | snakey? | 21:37 |
| devananda | crikey? | 21:38 |
| *** zdiN0bot has joined #openstack-ironic | 21:39 | |
| *** Haomeng has joined #openstack-ironic | 21:40 | |
| NobodyCam | devananda: I'm going to keep the try blocks and add a raise to them so we still get the node id in hte log | 21:41 |
| *** zdiN0bot has left #openstack-ironic | 21:41 | |
| *** matty_dubs is now known as matty_dubs|gone | 21:42 | |
| devananda | NobodyCam: adding a raise there will abort the whole deploy/rebuild | 21:43 |
| devananda | maybe that's the right thing to do | 21:43 |
| NobodyCam | yes | 21:44 |
| devananda | NobodyCam: one othe roption | 21:45 |
| devananda | NobodyCam: instead of re-generating the part table, assuming that it's the same as it was | 21:45 |
| devananda | what if we detected the part table from the device? | 21:45 |
| devananda | during rebuild | 21:46 |
| devananda | then it wouldn't matter what the instance_info[root/swap/ephemeral] values are | 21:46 |
| NobodyCam | we could dump the partition table easy then reconstruct part_dict from that | 21:50 |
| NobodyCam | then lol back to a seperate make_parted_dict function :-p | 21:51 |
| devananda | if preserve: detect part, write new root image | 21:51 |
| NobodyCam | reformat swap? | 21:51 |
| devananda | else: wipe start&end, write new partitions, write new root image | 21:51 |
| NobodyCam | if there | 21:52 |
| devananda | NobodyCam: that seems cleaner to me | 21:52 |
| NobodyCam | ack can I ask that you comment about thouse ideas to the review ... as I am trying to avoid the ping/pong effect | 21:53 |
| NobodyCam | and I agree I like that too | 21:53 |
| devananda | NobodyCam: http://paste.openstack.org/show/82246/ | 21:54 |
| devananda | like that | 21:54 |
| devananda | NobodyCam: AIUI this patch originally just tried to do one thing -- wipe start/end of disk when deploying | 21:55 |
| devananda | NobodyCam: and when it tried to start doing TWO things, it became really hard to land | 21:56 |
| devananda | a good example of why sometimes, it's best to make two patches ;) | 21:56 |
| NobodyCam | ahh see I have always tring to workaround the parted issue. my thought pattern was not focused enough and allowed reviews to side track it | 21:58 |
| devananda | NobodyCam: comment posted | 21:59 |
| NobodyCam | ach | 22:00 |
| NobodyCam | ack even | 22:00 |
| *** ellenh has quit IRC | 22:02 | |
| *** ellenh has joined #openstack-ironic | 22:05 | |
| devananda | NobodyCam: found the problem -- mac os and suse are packaging openipmi | 22:12 |
| devananda | which provides the same package (ipmitool) as the official ipmitool project | 22:13 |
| NobodyCam | ahh ha... great find | 22:13 |
| devananda | openipmi was used in rhel3, too. but they're on ipmitool now | 22:13 |
| NobodyCam | but we point that out in the docs so people know what to look for | 22:14 |
| devananda | yep | 22:15 |
| NobodyCam | but we SHOULD... | 22:15 |
| NobodyCam | :-p | 22:15 |
| NobodyCam | I'll split 93133 up this weekend | 22:15 |
| jroll | do we ever expect someone to run ironic on os x or rhel 3? | 22:17 |
| jroll | or I should ask, is it absurd of me to not want to support that? | 22:17 |
| devananda | jroll: i dont really want to support that either | 22:18 |
| devananda | i mean -- rhel 6 is reasonable ... rhel 3 ??? | 22:18 |
| jroll | right. | 22:18 |
| jroll | like, if you're running rhel 3, you have bigger problems than automating bare metal | 22:18 |
| NobodyCam | jroll: what sles11 | 22:19 |
| NobodyCam | bsd | 22:19 |
| * devananda searches pbone | 22:19 | |
| jroll | huh? | 22:19 |
| NobodyCam | suse 11 has openipi | 22:20 |
| NobodyCam | openipmi | 22:20 |
| jroll | hrm | 22:20 |
| NobodyCam | as does freeBSD | 22:20 |
| devananda | NobodyCam: so try removing openipmi and installing ipmitool | 22:20 |
| devananda | NobodyCam: pbone says SLES11 has *both* | 22:20 |
| devananda | :) | 22:20 |
| devananda | it just may default to the other one | 22:21 |
| NobodyCam | :) | 22:21 |
| jroll | NobodyCam: maybe fair, but I would document "hey don't use openipmi" rather than support openipmi | 22:21 |
| devananda | ipmitool has been around since 2003 | 22:21 |
| devananda | if some distro doesn't have support for it, that's a bigger problem | 22:21 |
| *** romcheg has joined #openstack-ironic | 22:21 | |
| devananda | hmm, bot is down again | 22:24 |
| devananda | I just updated the patch | 22:24 |
| NobodyCam | looks like I can brew install it (testing now) if so that a easy fix for devs, I'll have to figure out how to test on sles11 | 22:27 |
| NobodyCam | yep brew install works | 22:30 |
| NobodyCam | nobodycamair:ironic NobodyCam$ ipmitool -V | 22:30 |
| NobodyCam | ipmitool version 2.5b1 | 22:30 |
| NobodyCam | nobodycamair:ironic NobodyCam$ /usr/local/bin/ipmitool -V | 22:30 |
| NobodyCam | ipmitool version 1.8.13 | 22:30 |
| NobodyCam | easy fix for mac | 22:31 |
| *** openstackgerrit has joined #openstack-ironic | 22:32 | |
| * devananda fixes pep8 error | 22:34 | |
| devananda | NobodyCam: great :) | 22:34 |
| NobodyCam | humm now we need a patch to make https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ipmitool.py#L131 location adjustable | 22:34 |
| devananda | NobodyCam: no .... | 22:34 |
| devananda | NobodyCam: fix your system | 22:34 |
| NobodyCam | lol :-p | 22:34 |
| *** openstackgerrit has quit IRC | 22:35 | |
| * devananda updates patch again | 22:35 | |
| devananda | now with doc changes :) | 22:36 |
| *** openstackgerrit has joined #openstack-ironic | 22:36 | |
| NobodyCam | TY :) | 22:36 |
| NobodyCam | hehehehe | 22:36 |
| *** sseago has quit IRC | 22:49 | |
| *** blamar has joined #openstack-ironic | 23:01 | |
| *** Haomeng|2 has quit IRC | 23:01 | |
| *** sseago has joined #openstack-ironic | 23:12 | |
| *** zdiN0bot has joined #openstack-ironic | 23:21 | |
| *** sseago has quit IRC | 23:27 | |
| *** zdiN0bot has quit IRC | 23:32 | |
| *** igordcard has quit IRC | 23:37 | |
| *** sseago has joined #openstack-ironic | 23:39 | |
| * devananda starts the weekend 7 minutes early | 23:53 | |
| openstackgerrit | Ellen Hui proposed a change to openstack/ironic-python-agent: Tries to advertise valid default IP https://review.openstack.org/96980 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!