| adam_g | jroll, oh cool. does that require some changes elsewhere to make those files available on tarballs.o.o? | 00:02 |
|---|---|---|
| jroll | adam_g: already landed | 00:02 |
| jroll | adam_g: this came out of sdague not liking us unpacking the tarball and whantot | 00:02 |
| adam_g | jroll, weird, okay | 00:03 |
| jroll | adam_g: so once that and another change land... need infra/config changes to do CI on IPA | 00:03 |
| jroll | which I've based off your refactor | 00:04 |
| adam_g | yeah, let me ping them now about that | 00:04 |
| jroll | idk if you have time to babysit that or want me to take it | 00:04 |
| jroll | oh, cool :) | 00:04 |
| jroll | happy to help push that through if needed | 00:04 |
| adam_g | maybe if you peste^Wping them too it will help :) | 00:04 |
| jroll | heh, I've pinged them enough today :P | 00:05 |
| jroll | but happy to do so tomorrow :D | 00:05 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Draft a 'pxe_logpower' driver https://review.openstack.org/98950 | 00:07 |
| *** romcheg1 has quit IRC | 00:08 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-python-agent: Add versioning to Agent decommission https://review.openstack.org/106859 | 00:09 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-python-agent: Adding support for decommissioning https://review.openstack.org/104379 | 00:09 |
| Haomeng | mrda: night:) | 00:13 |
| mrda | night Haomeng! | 00:14 |
| mrda | you must be up late? | 00:14 |
| Haomeng | my timezone is +8, now it is 8am:) | 00:17 |
| Haomeng | guess you are in night:) | 00:17 |
| mrda | Haomeng: It's 10am here for me | 00:19 |
| Haomeng | mrda: :) | 00:20 |
| mrda | I'm +9.5 | 00:20 |
| Haomeng | mrda: :) | 00:20 |
| *** igordcard has quit IRC | 00:24 | |
| *** rloo has quit IRC | 00:28 | |
| *** bandicot has quit IRC | 00:29 | |
| *** Ugallu has quit IRC | 00:46 | |
| *** chuckC has quit IRC | 00:50 | |
| *** GheRivero has quit IRC | 00:50 | |
| *** GheRivero has joined #openstack-ironic | 00:51 | |
| *** killer_prince is now known as lazy_prince | 01:29 | |
| lazy_prince | hi all.. pls take a bit of time to review https://review.openstack.org/#/c/103483/ and provide your valuable comments.. | 01:38 |
| *** rloo has joined #openstack-ironic | 01:42 | |
| *** nosnos has joined #openstack-ironic | 01:49 | |
| rloo | mrda: did you see https://review.openstack.org/#/c/111428/ failed pep8 | 01:49 |
| mrda | yes | 01:49 |
| mrda | rloo: am prepping a new fix and uploading now | 01:49 |
| mrda | sigh | 01:49 |
| mrda | thank you for noticing | 01:49 |
| rloo | thx mrda. | 01:50 |
| rloo | just missing the from nova.i18n import _ line. | 01:50 |
| rloo | talk to you next week mrda. I'm signing out now. (yay) | 01:50 |
| mrda | have fun rloo! | 01:51 |
| *** rloo has quit IRC | 01:51 | |
| mrda | wow, 111428 fails in check, but doesn't fail locally. I for one welcome our new Heisenbug overlords. | 01:55 |
| *** ryanpetrello has quit IRC | 01:56 | |
| *** ryanpetrello has joined #openstack-ironic | 02:00 | |
| *** eghobo has quit IRC | 02:11 | |
| *** harlowja is now known as harlowja_away | 02:27 | |
| *** lazy_prince is now known as killer_prince | 02:28 | |
| *** dkehnx has quit IRC | 02:37 | |
| *** dlaube has quit IRC | 02:49 | |
| *** eghobo has joined #openstack-ironic | 02:51 | |
| *** zz_naotok is now known as naotok | 03:03 | |
| *** vinbs has joined #openstack-ironic | 03:09 | |
| *** eghobo has quit IRC | 03:10 | |
| *** dlaube has joined #openstack-ironic | 03:22 | |
| *** dlaube has quit IRC | 03:23 | |
| *** wanyen has quit IRC | 03:28 | |
| *** mgoddard has quit IRC | 03:28 | |
| *** eghobo has joined #openstack-ironic | 03:45 | |
| *** pcrews has quit IRC | 04:01 | |
| *** Poornima has joined #openstack-ironic | 04:23 | |
| *** pcrews has joined #openstack-ironic | 04:41 | |
| *** rushiagr_away is now known as rushiagr | 04:47 | |
| *** pcrews has quit IRC | 04:48 | |
| *** kevinbenton has quit IRC | 05:03 | |
| *** kevinbenton has joined #openstack-ironic | 05:06 | |
| *** krtaylor has quit IRC | 05:12 | |
| *** nikunj2513 has joined #openstack-ironic | 05:15 | |
| *** rakesh_hs has joined #openstack-ironic | 05:20 | |
| *** krtaylor has joined #openstack-ironic | 05:20 | |
| *** killer_prince is now known as lazy_prince | 05:50 | |
| *** yuanying has joined #openstack-ironic | 05:59 | |
| *** yuanying_ has quit IRC | 05:59 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/118548 | 06:08 |
| *** k4n0 has joined #openstack-ironic | 06:25 | |
| *** Haomeng has quit IRC | 06:27 | |
| *** Haomeng has joined #openstack-ironic | 06:30 | |
| *** nikunj2513 has quit IRC | 06:34 | |
| *** nikunj2513 has joined #openstack-ironic | 06:36 | |
| *** eghobo has quit IRC | 06:37 | |
| *** bluex-pl has joined #openstack-ironic | 07:08 | |
| *** athomas has joined #openstack-ironic | 07:08 | |
| *** jistr has joined #openstack-ironic | 07:15 | |
| *** chenglch has joined #openstack-ironic | 07:27 | |
| *** bluex-pl has quit IRC | 07:30 | |
| *** jcoufal has joined #openstack-ironic | 07:32 | |
| *** igordcard has joined #openstack-ironic | 07:44 | |
| *** andreykurilin_ has joined #openstack-ironic | 07:46 | |
| *** ifarkas has joined #openstack-ironic | 07:47 | |
| *** derekh has joined #openstack-ironic | 07:54 | |
| *** agordeev has quit IRC | 07:58 | |
| *** naotok is now known as zz_naotok | 07:58 | |
| *** agordeev has joined #openstack-ironic | 08:05 | |
| *** ndipanov has joined #openstack-ironic | 08:07 | |
| *** igordcard has quit IRC | 08:11 | |
| *** dtantsur|afk is now known as dtantsur | 08:15 | |
| dtantsur | Morning Ironic | 08:15 |
| mrda | Hey dtantsur | 08:19 |
| *** romcheg has quit IRC | 08:25 | |
| dtantsur | mrda, g'evening) | 08:29 |
| *** romcheg has joined #openstack-ironic | 08:29 | |
| *** lucasagomes has joined #openstack-ironic | 08:31 | |
| yuriyz | morning Ironic dtantsur mrda | 08:34 |
| dtantsur | yuriyz, morning | 08:34 |
| *** derekh has quit IRC | 08:37 | |
| mrda | hi yuriyz romcheg | 08:39 |
| lucasagomes | morning dtantsur mrda yuanying | 08:46 |
| lucasagomes | ouch | 08:46 |
| lucasagomes | yuriyz, | 08:46 |
| *** yongli has quit IRC | 08:48 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Remove unnecessary mapping from Agent drivers https://review.openstack.org/118326 | 08:54 |
| dtantsur | lucasagomes, morning. can we close ironic part of https://bugs.launchpad.net/ironic/+bug/1351010actually? | 08:54 |
| lucasagomes | dtantsur, +1 | 08:54 |
| dtantsur | ok | 08:54 |
| lucasagomes | dtantsur, I don't think we have anything else to do apart from doc it | 08:54 |
| *** bluex-pl has joined #openstack-ironic | 08:55 | |
| *** andreykurilin_ has quit IRC | 09:00 | |
| openstackgerrit | A change was merged to openstack/ironic: Use models.TimestampMixin from oslo.db https://review.openstack.org/119195 | 09:05 |
| *** Haomeng|2 has joined #openstack-ironic | 09:07 | |
| *** Haomeng has quit IRC | 09:08 | |
| *** dtantsur is now known as dtantsur|brb | 09:10 | |
| mrda | Night all, see you Monday | 09:12 |
| *** k4n0 has quit IRC | 09:16 | |
| *** MattMan has joined #openstack-ironic | 09:18 | |
| *** nikunj2513 has quit IRC | 09:24 | |
| *** igordcard has joined #openstack-ironic | 09:27 | |
| openstackgerrit | A change was merged to openstack/ironic: mock.assert_called_once() is not valid method https://review.openstack.org/119234 | 09:29 |
| *** zz_naotok is now known as naotok | 09:31 | |
| *** bluex-pl has quit IRC | 09:31 | |
| *** GheRivero is now known as GheRivero_pto | 09:34 | |
| *** bluex-pl has joined #openstack-ironic | 09:37 | |
| *** pelix has joined #openstack-ironic | 09:39 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Log ERROR power state in node_power_action() https://review.openstack.org/117499 | 09:41 |
| *** bluex-pl has quit IRC | 09:41 | |
| *** bluex-pl has joined #openstack-ironic | 09:42 | |
| openstackgerrit | Vinay B S proposed a change to openstack/ironic: Adds instructions for deploying instances on real hardware https://review.openstack.org/118614 | 09:45 |
| *** chenglch has quit IRC | 09:54 | |
| *** vdrok_afk is now known as vdrok | 10:08 | |
| *** jcoufal has quit IRC | 10:16 | |
| *** derekh has joined #openstack-ironic | 10:23 | |
| *** bluex-pl has quit IRC | 10:26 | |
| lucasagomes | yuriyz, should we tag bug 1358816 to RC1? | 10:37 |
| yuriyz | lucasagomes, IMO we should | 10:38 |
| *** jcoufal has joined #openstack-ironic | 10:41 | |
| lucasagomes | yuriyz, ack, will tag it | 10:43 |
| *** Haomeng has joined #openstack-ironic | 10:47 | |
| *** Haomeng|2 has quit IRC | 10:47 | |
| *** dtantsur|brb is now known as dtantsur | 11:01 | |
| *** bluex-pl has joined #openstack-ironic | 11:04 | |
| *** bluex-pl has quit IRC | 11:09 | |
| *** bluex-pl has joined #openstack-ironic | 11:10 | |
| romcheg | Morning guys | 11:12 |
| romcheg | I'm not very familiar with this internals of hashing. Could someone please explain me why big endian is chosen here? https://review.openstack.org/#/c/118932/4/ironic/common/hash_ring.py | 11:14 |
| romcheg | It's on L85 | 11:14 |
| *** teju has joined #openstack-ironic | 11:20 | |
| teju | hi all... need help with boot provisioning using ironic....the provisioning state is stuck in wait call-back | 11:22 |
| romcheg | teju: For how long? | 11:24 |
| romcheg | teju: Hi, btw :) | 11:24 |
| teju | romcheg : it's been more than 20 min | 11:25 |
| romcheg | teju: Is tftp server up and runnign? | 11:25 |
| teju | romcheg : yes | 11:25 |
| openstackgerrit | A change was merged to openstack/ironic: Unbreak debugging via testr https://review.openstack.org/118896 | 11:25 |
| romcheg | teju: What about DHCP, can you see anything in the logs? | 11:25 |
| teju | romcheg : when i ran the command to deploy, the baremetal server is powered ON, but it failed to get boot image ....i get error pxe e53 | 11:26 |
| romcheg | teju: Could you please search your syslog to find something like DHCPDISCOVER from your node? | 11:27 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Translator functions cleanup part 1 https://review.openstack.org/116303 | 11:29 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Translator functions cleanup part 2 https://review.openstack.org/118587 | 11:29 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Translator functions cleanup part 3 https://review.openstack.org/118588 | 11:30 |
| romcheg | teju: AFAIR pxe e53 may be related to wrong configuration of the DHCP server | 11:30 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Translator functions cleanup part 4 https://review.openstack.org/118591 | 11:30 |
| teju | romcheg : i am running dnsmasq | 11:31 |
| *** Poornima has quit IRC | 11:31 | |
| romcheg | teju: It would be useful to run "neutron port-show <your_port_uuid>" to check the PXE configuration | 11:31 |
| *** Poornima has joined #openstack-ironic | 11:32 | |
| teju | romcheg : i followed this link : http://ma.ttwagner.com/bare-metal-deploys-with-devstack-and-ironic/#comment-2578 | 11:34 |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Log ERROR power state in node_power_action() https://review.openstack.org/117499 | 11:37 |
| *** vinbs has quit IRC | 11:38 | |
| dtantsur | teju, "NOTE: This guide is fairly dated (originally authored in January)." ;) | 11:39 |
| dtantsur | Matt started updating it, but I doubt he finished | 11:39 |
| *** lucasagomes is now known as lucas-hungry | 11:40 | |
| teju | dtantsur : ok....when i ran the curl command mentioned in that link, the server is powered ON, but it failed to get boot image.....may be i am missing some conf | 11:41 |
| teju | romcheg : http://pastebin.com/f2bnGhDv | 11:43 |
| romcheg | teju: So I see that DHCP seems to be accessible for the node (that happened to be a problem in the past). Could you please show the detailed neutron port info? | 11:44 |
| *** naotok is now known as zz_naotok | 11:46 | |
| *** Haomeng|2 has joined #openstack-ironic | 11:46 | |
| romcheg | you can find the Neutron vif _id in the Ironic Port extra specs | 11:46 |
| *** Haomeng has quit IRC | 11:48 | |
| teju | romcheg : i dont know whether I am making sense, but i am trying to boot provision without neutron...will this work? i mean, neutron is just for providing IP for my bare metal server...instead I am using dnsmasq for dhcp | 11:48 |
| romcheg | teju: Ah, makes sense | 11:48 |
| teju | romcheg : that should work...right? | 11:49 |
| romcheg | teju: than something has to set PXE options in the DHCP settings | 11:49 |
| romcheg | looks like the node ain't getting them | 11:50 |
| teju | romcheg : yes....provisioning state change from wait call back to deploy failed | 11:50 |
| romcheg | teju: because of the timeout, I believe | 11:51 |
| teju | romcheg : one basic question....pls dont mind,,,, do I need any physical connection from bare metal server to openstack server? | 11:52 |
| romcheg | teju: So the node has to be able to connect both to the DHCP server and to the conductor | 11:54 |
| romcheg | teju: and to the ironic api too | 11:54 |
| *** k4n0 has joined #openstack-ironic | 11:55 | |
| romcheg | teju: and of course to the tftp server too | 11:56 |
| teju | romcheg : yes....this gets complicated bcoz i am running all openstack services(keystone,glance,nova,neutron,ironic) in a VM | 11:57 |
| teju | romcheg : and I am trying to boot a real bare metal server | 11:57 |
| teju | romcheg : thanks for ur time....i will try to fix dhcp issues | 11:58 |
| romcheg | teju: That is not very complicated, I have the same kind of setup on my lab (the only difference is that I use neutron for configuring dhcp) | 11:58 |
| romcheg | teju: basically you can add another interface to the VM to the same network your bm node is | 11:58 |
| romcheg | teju: and set up all required PXE params to the dnsmasq | 12:00 |
| teju | romcheg : or I can run dhcp server in a physical server rather than in VM and connect this server to bare metal server...this may work...right? | 12:01 |
| *** openstackgerrit has quit IRC | 12:01 | |
| *** openstackgerrit has joined #openstack-ironic | 12:03 | |
| romcheg | teju: node still has to have access to the tftp server and to the ironic api to send the callback | 12:04 |
| teju | romcheg : yes, once the bare metal node gets IP from dhcp server, it will be able to talk to all openstack services, tftp, | 12:06 |
| teju | romcheg : bcoz all the IPs for physical server and VMs are in the same subnet | 12:07 |
| romcheg | teju: then it should work | 12:08 |
| teju | romcheg : once again...thanks... i will try to get pxe boot work | 12:09 |
| romcheg | teju: you're always welcome ;) | 12:09 |
| openstackgerrit | Derek Higgins proposed a change to openstack/ironic: Revert "Nova review updates for _node_resource" https://review.openstack.org/119357 | 12:11 |
| teju | No VIFs found for node f8908f2c-2abf-4ec5-b25c-726793f4983b when attempting to update Neutron DHCP BOOT options. | 12:11 |
| derekh | Tripleo Ironic CI tests are failing : bug 1365750 : the revert ^^ seems to fix it | 12:11 |
| dtantsur | derekh, hmm.. this change was requested during Nova review, I guess you want to talk to nova folks... | 12:12 |
| dtantsur | mrda, lucas-hungry ^^^ | 12:12 |
| *** teju has left #openstack-ironic | 12:14 | |
| derekh | dtantsur: hmm, maybe it depends on a nova patch thats yet to merge ? | 12:14 |
| dtantsur | IIRC yes | 12:14 |
| dtantsur | don't hav e link at hand | 12:14 |
| *** bluex-pl has quit IRC | 12:18 | |
| *** nosnos has quit IRC | 12:19 | |
| *** nosnos has joined #openstack-ironic | 12:20 | |
| *** nosnos has quit IRC | 12:24 | |
| *** bluex-pl has joined #openstack-ironic | 12:24 | |
| *** Poornima has quit IRC | 12:26 | |
| *** rakesh_hs has quit IRC | 12:34 | |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/python-ironicclient: Add sort_key and sort_dir parameters to *-list https://review.openstack.org/114555 | 12:36 |
| *** bigjools has quit IRC | 12:38 | |
| *** bigjools has joined #openstack-ironic | 12:40 | |
| *** bigjools has joined #openstack-ironic | 12:40 | |
| lucas-hungry | derekh, uuu, yup... actually that was something they required us to have | 12:45 |
| lucas-hungry | to canonicalize the arch | 12:45 |
| *** lucas-hungry is now known as lucasagomes | 12:45 | |
| *** viktors|afk is now known as viktors | 12:45 | |
| viktors | dtantsur: hi! | 12:46 |
| lucasagomes | I think it's on part 3 patch set 14 | 12:46 |
| dtantsur | viktors, hi | 12:46 |
| lucasagomes | this discussion started | 12:46 |
| lucasagomes | derekh, will grab the link | 12:46 |
| viktors | dtantsur: does someone suppose to fix bug 1347604 (ironic-dbsync fails on SQLite due to bad support of migrations) ? ) | 12:47 |
| derekh | lucasagomes: ok, we've reverted it in tripleo CI, I'll remove the revert once its merged | 12:47 |
| lucasagomes | derekh, https://review.openstack.org/#/c/111425/14/nova/virt/ironic/driver.py | 12:47 |
| dtantsur | viktors, depending on definition of "fix". I'll be happy to close it, once we fix the docs | 12:47 |
| lucasagomes | derekh, it was just approved in nova | 12:47 |
| derekh | lucasagomes: back in a few minutes, gotta get of train | 12:47 |
| derekh | lucasagomes: ok, will try it in a few minutes | 12:48 |
| lucasagomes | derekh, ack, right yeah we have to sync about it and find a consensus | 12:48 |
| romcheg | lucasagomes, dtantsur, could you please check out my question about endianess ^ | 12:48 |
| lucasagomes | romcheg, sorry didn't see that | 12:48 |
| lucasagomes | lemme see | 12:48 |
| *** derekh has quit IRC | 12:48 | |
| romcheg | lucasagomes: thanks! | 12:49 |
| viktors | dtantsur: yes, I mean documentation fixes. My English is far from perfect, so I'll be glad, if someone with good english will fix the docs :) | 12:49 |
| dtantsur | viktors, then unassign yourself with the comment and someone will pick it | 12:50 |
| viktors | dtantsur: ok, will do | 12:50 |
| lucasagomes | romcheg, I think ur comment is accurate. I'm not sure why big endian, but I don't think if that makes any difference too | 12:51 |
| lucasagomes | romcheg, I know where that came from tho, I've the article somewhere about the consitent hashing ring | 12:52 |
| lucasagomes | where we based ours on | 12:52 |
| lucasagomes | will grab the link | 12:52 |
| romcheg | lucasagomes: let's wait for lifeless then :) | 12:52 |
| lucasagomes | (if I find) | 12:52 |
| lucasagomes | romcheg, we may find the answer in the article, out of the top of my head I don't remmeber | 12:52 |
| lucasagomes | romcheg, here we go http://greg.brim.net/page/building_a_consistent_hashing_ring.html | 12:53 |
| romcheg | lucasagomes: thanks! | 12:53 |
| *** Haomeng|2 has quit IRC | 12:54 | |
| romcheg | lucasagomes: at least I can see they use big endian too :D | 12:55 |
| lucasagomes | romcheg, heh yeah | 12:57 |
| *** Haomeng|2 has joined #openstack-ironic | 12:57 | |
| * romcheg starts suspecting someone did copy&paste | 12:57 | |
| jbjohnso__ | well I just reviewed redfish specification and gave feedback... I frankly don't see it working in the intended role very well | 13:10 |
| *** derekh has joined #openstack-ironic | 13:11 | |
| *** pelix has quit IRC | 13:13 | |
| *** Haomeng|2 has quit IRC | 13:25 | |
| *** Haomeng|2 has joined #openstack-ironic | 13:27 | |
| *** pcrews has joined #openstack-ironic | 13:33 | |
| *** dhellmann is now known as dhellmann_ | 13:40 | |
| *** bigjools has quit IRC | 13:42 | |
| rushiagr | lucasagomes: hi.. I have resubmitted the patch https://review.openstack.org/#/c/118554/ (Move metadata filtering logic to utils.py), incorporating your comments. Please have a relook when you have time | 13:44 |
| rushiagr | lucasagomes: and thanks for the review :) | 13:44 |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Implement hardware discovery setup in PXE driver https://review.openstack.org/110031 | 13:44 |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: Add newly_discovered column to Node object https://review.openstack.org/107389 | 13:44 |
| *** bigjools has joined #openstack-ironic | 13:45 | |
| * lucasagomes feels like he should go to #openstack-nova | 13:45 | |
| lucasagomes | rushiagr, sure will do! thanks for addressing it | 13:45 |
| rushiagr | lucasagomes: heh | 13:45 |
| *** r-daneel has joined #openstack-ironic | 13:53 | |
| *** SpamapS has quit IRC | 13:59 | |
| marios | quick q: when talking about actual bare metal machines and discovery, will we still create an entry in Ironic for any new machines (populating only ipmi creds i guess). i.e. same as for the virt case? will we still need to use MAC to match against discovered info? | 13:59 |
| marios | (ok that was 2 questions) | 14:00 |
| *** SpamapS_ has joined #openstack-ironic | 14:02 | |
| *** datajerk has joined #openstack-ironic | 14:03 | |
| dtantsur | marios, if you mwan what I'm prototyping (there's no official discovery in Ironic), then 1. yes, 2. no | 14:03 |
| dtantsur | * mean actually | 14:03 |
| marios | dtantsur: thanks! (just noticed those reviews earlier in the chat, do they answer my questions?) | 14:04 |
| dtantsur | marios, that' my code, but it's in the middle of refactoring | 14:05 |
| marios | dtantsur: i guess that means the 'maintenance mode' could/will be used for baremetal (at least right now) | 14:05 |
| dtantsur | marios, likely will | 14:06 |
| marios | dtantsur: am asking because i'm converting the dnsmasq script to element, so we can change the logic there | 14:06 |
| marios | dtantsur: k thanks matey | 14:06 |
| marios | dtantsur: (one last thing, will be easier once i get a review out and we can discuss specifics, but): doing it this way is quite expensive (to get a list of MAC)... get nodes where maintenance=False (OK), but then for each get ports (expensive) | 14:07 |
| marios | dtantsur: so wanted to check if there would be difference in baremetal case that might make it better in this respect | 14:08 |
| dtantsur | marios, for real thing (not virt) you don't have to check for maintenance, just black-list everything in ports | 14:08 |
| marios | dtantsur: I C. NO MAC! | 14:08 |
| marios | dtantsur: yes, there will be no longer any need to match with the MAC right | 14:09 |
| marios | dtantsur: so ironic entry won't have a mac (or port? or port without mac, whatever) for new nodes | 14:09 |
| dtantsur | right. at least that's how I see it | 14:09 |
| marios | dtantsur: yes seems right. which is great. (now just need to be able to distinguish if we are talking about vms or bm and am in business, as the script will now depend) | 14:10 |
| marios | dtantsur: thanks again very helpful | 14:12 |
| dtantsur | np) | 14:12 |
| *** ndipanov is now known as ndipanoff | 14:16 | |
| *** Haomeng|2 has quit IRC | 14:18 | |
| jroll | good morning ironic :) | 14:22 |
| romcheg | lucasagomes: in your last comment on 117499 you said that you are approving the patch but I don't see +A. Didn't you miss it? | 14:26 |
| *** Haomeng|2 has joined #openstack-ironic | 14:27 | |
| romcheg | Morning jroll! | 14:27 |
| lucasagomes | romcheg, oh it it was patch set 7 | 14:27 |
| lucasagomes | yuriyz, posted a new one rebasing it | 14:27 |
| lucasagomes | will re-approve | 14:27 |
| lucasagomes | done | 14:28 |
| romcheg | Hmm, gerrit can apply old scores after a rebase again :) | 14:28 |
| Shrews | mordred: ooh, dox sounds interesting | 14:30 |
| mordred | Shrews: it needs a little bit more work... | 14:31 |
| jroll | romcheg: morning :) | 14:31 |
| jroll | and lucasagomes and Shrews and mordred and dtantsur :) | 14:31 |
| mordred | morning jroll | 14:31 |
| Shrews | mordred: interestingly, i had a very similar idea a month or so back (using docker for tests) | 14:32 |
| Shrews | jroll: morning | 14:32 |
| dtantsur | jroll, Shrews, mordred, morning | 14:32 |
| Shrews | hey dtantsur o/ | 14:33 |
| jroll | Shrews: seen something like this before? http://logs.openstack.org/65/119165/3/check/check-tempest-dsvm-virtual-ironic/3720214/logs/screen-ir-api.txt.gz?level=WARNING | 14:33 |
| jroll | all sorts of weird port manipulation problems | 14:33 |
| mordred | Shrews: it makes TOTAL sense for things like unittests/locally for the dev | 14:33 |
| mordred | Shrews: especially since it's got the speed and caching thing sorted | 14:33 |
| Shrews | mordred: exactly. i had it in mind for running ironic scenario tests locally instead of in tempest | 14:33 |
| mordred | Shrews: it's too bad iscsi isn't namespaced... | 14:33 |
| Shrews | jroll: looking | 14:34 |
| Shrews | jroll: O.o | 14:34 |
| Shrews | jroll: is this happening in several reviews, or just this one? | 14:35 |
| jroll | Shrews: just this one afaik? | 14:36 |
| jroll | I might just recheck it | 14:36 |
| jroll | mordred: maybe we should replace iscsi from ironic :) | 14:36 |
| jroll | owait | 14:36 |
| jroll | doing that already | 14:36 |
| jroll | unfortunately, I don't think it will actually go away :/ | 14:37 |
| NobodyCam | Good morning Ironic | 14:39 |
| NobodyCam | TGIF! | 14:39 |
| jroll | morning NobodyCam, TGIF indeed | 14:40 |
| NobodyCam | morning mordred :) | 14:40 |
| NobodyCam | morning jroll :) | 14:40 |
| jroll | Shrews: rechecking | 14:41 |
| Shrews | jroll: this is really odd. checking for that first chassis id in ir-cond shows nothing. it's like the wrong api server was being used or something | 14:42 |
| jroll | Shrews: it's like openstack doesn't work or something | 14:43 |
| *** dhellmann_ is now known as dhellmann | 14:43 | |
| Shrews | jroll: lol. the first node though (38f3eb1a) was actually successfully deleted according to ir-cond. so weird | 14:43 |
| jroll | yeah. completely lost. | 14:44 |
| lucasagomes | jroll, yo morning | 14:45 |
| jroll | heya :) | 14:46 |
| Shrews | jroll: so, these are normal | 14:50 |
| Shrews | jroll: the tempest test deletes the node, then attempts to get it again to make sure that NotFound is raised | 14:50 |
| jroll | Shrews: I was looking at this weirdness: http://logs.openstack.org/65/119165/3/check/check-tempest-dsvm-virtual-ironic/3720214/logs/screen-ir-api.txt.gz?level=WARNING#_2014-09-05_04_17_44_214 | 14:51 |
| jroll | and the couple lines after | 14:51 |
| jroll | Shrews: and, the test failed, to be clear | 14:51 |
| Shrews | jroll: yeah. looking at that one now (the earlier ones left me confused, though) | 14:51 |
| Shrews | the *real* error is in the baremetal scenario test: Details: Timed out waiting to get Ironic node by instance id 33822cbf-59d5-4870-acad-44bde8961f87 | 14:53 |
| jroll | right | 14:54 |
| Shrews | that i've seen before, though rarely | 14:55 |
| * Shrews blames a slow node | 14:55 | |
| *** k4n0 has quit IRC | 14:57 | |
| jroll | we're only 2 +A votes away from nova driver landing :) | 15:01 |
| NobodyCam | woo hoo | 15:01 |
| *** bluex-pl has quit IRC | 15:05 | |
| *** Guest18621 is now known as annegentle | 15:16 | |
| *** pelix has joined #openstack-ironic | 15:17 | |
| *** datajerk has quit IRC | 15:17 | |
| *** vdrok is now known as vdrok_afk | 15:24 | |
| dtantsur | NobodyCam, morning and TGIF | 15:32 |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Implement hardware discovery setup in PXE driver https://review.openstack.org/110031 | 15:33 |
| dtantsur | actually calling it a day for today :) have a nice weekend | 15:36 |
| *** dtantsur is now known as dtantsur|afk | 15:36 | |
| NobodyCam | morning dtantsur|afk | 15:37 |
| NobodyCam | have a good weekend dtantsur|afk | 15:38 |
| dtantsur|afk | :) | 15:38 |
| NobodyCam | just some general fyi. I will be not online next Friday! | 15:40 |
| *** slagle_ is now known as slagle | 15:41 | |
| lazy_prince | JoshNang: I am trying to use https://github.com/rackerlabs/ironic-neutron-plugin | 15:42 |
| lazy_prince | JoshNang: As per https://github.com/rackerlabs/ironic-neutron-plugin/blob/master/etc/neutron.conf.dist, i am using mechanism_drivers = ironic | 15:43 |
| lazy_prince | JoshNang: but it throws error... | 15:44 |
| jroll | morgabra: ^^ | 15:44 |
| lucasagomes | jrist, w00t | 15:47 |
| lucasagomes | morning NobodyCam | 15:47 |
| lucasagomes | dtantsur|afk, g'night | 15:47 |
| jroll | lazy_prince: morgabra might know more about that, I could help you later today too | 15:47 |
| jrist | I need to change my nick, don't I jroll, lucasagomes ? | 15:48 |
| jroll | hehehe, somebody does | 15:48 |
| lucasagomes | jrist, oh gosh again | 15:48 |
| lucasagomes | lol | 15:49 |
| jroll | OMG | 15:49 |
| jroll | oh | 15:49 |
| jroll | that one was intentional :P | 15:49 |
| lucasagomes | jrist, how are u doing :) | 15:49 |
| lucasagomes | yeah that was intentional jroll :D | 15:49 |
| jrist | I was born in 1981, so can I lay claim, jroll? | 15:49 |
| jroll | damn. | 15:49 |
| jrist | also mine is alphabetically in front | 15:49 |
| jrist | lucasagomes: great, you? | 15:49 |
| * jroll is a youngin, I guess | 15:49 | |
| *** jistr has quit IRC | 15:49 | |
| lucasagomes | jrist, all is well :D | 15:49 |
| jrist | haha | 15:50 |
| jrist | making me feel old | 15:50 |
| NobodyCam | 81 gah /me is old | 15:52 |
| jroll | I won't tell you when I was born, then :P | 15:52 |
| * NobodyCam was born before 1970 :-p | 15:52 | |
| lucasagomes | heh | 15:55 |
| lucasagomes | 89 FTW! | 15:55 |
| NobodyCam | gah /me gets his cain and hearing aid and trundles away | 15:56 |
| jroll | yay, I'm not the youngest | 15:56 |
| jroll | LOL | 15:56 |
| NobodyCam | lol | 15:56 |
| lucasagomes | ;( | 15:57 |
| NobodyCam | hehehe | 15:57 |
| jroll | this passed jenkins now | 15:57 |
| jroll | https://review.openstack.org/#/c/118693/ | 15:57 |
| jroll | it's already in nova, one of y'all want to approve it? | 15:57 |
| lazy_prince | morgabra: here is the complete log in case if you need it. http://paste.openstack.org/show/106614/ | 15:58 |
| jroll | oooo colors | 15:58 |
| NobodyCam | jroll: landing | 15:59 |
| jroll | lazy_prince: to be clear, you pip install'd this plugin, right? | 15:59 |
| jroll | NobodyCam: thanks! | 15:59 |
| *** dlaube has joined #openstack-ironic | 15:59 | |
| lazy_prince | nope.. i pulled it from github | 16:00 |
| lazy_prince | jroll: devstack did that for me | 16:00 |
| jroll | are you sure devstack pip installed it? | 16:01 |
| jroll | you need to do that | 16:01 |
| lazy_prince | actually devstack did setup neutron for me and then i stopped it and used the instructions from https://github.com/rackerlabs/ironic-neutron-plugin | 16:01 |
| jroll | or python setup.py install | 16:01 |
| jroll | or whatever | 16:01 |
| jroll | to load the entrypoints | 16:01 |
| jroll | export PYTHONPATH=$PYTHONPATH:/path/to/ironic-neutron-plugin | 16:02 |
| jroll | might also work, idk if you did that | 16:02 |
| lazy_prince | i am sorry.. which plugin, ironic-neutron-plugin (neutron plugin), i copied it manually after stopping neutron service to neutron directory.. | 16:03 |
| jroll | yeah | 16:03 |
| jroll | you need to cd to ironic-neutron-plugin | 16:03 |
| jroll | python setup.py install | 16:03 |
| jroll | that's how the entrypoint in setup.cfg get loaded, so that stevedore knows about them | 16:04 |
| lazy_prince | okay let me do that.. | 16:04 |
| jroll | lazy_prince: what sort of switch will you be using this with? | 16:04 |
| lazy_prince | I will be using HP switch.. | 16:05 |
| jroll | also... we still need to upstream our code that uses this | 16:05 |
| jroll | ok, you'll likely have to write a driver, then, too | 16:05 |
| jroll | in here: https://github.com/rackerlabs/ironic-neutron-plugin/tree/master/ironic_neutron_plugin/drivers | 16:05 |
| lazy_prince | I have finished the driver code for it | 16:05 |
| jroll | neat :) | 16:05 |
| lazy_prince | and wanted to test end-to-end.. thats where i am getting into issues.. | 16:06 |
| jroll | right | 16:06 |
| jroll | the upstreaming part of this is going to be interesting | 16:07 |
| lazy_prince | python setup.py install failed with error about EnvironmentError: mysql_config not found | 16:07 |
| jroll | there's other people basically writing this in neutron | 16:07 |
| jroll | apt-get install libmysqlclient-dev | 16:07 |
| *** igordcard has quit IRC | 16:07 | |
| jroll | (strangely, that should already be installed in devstack, I think) | 16:08 |
| lazy_prince | well.. i am trying to see if it works for me.. it is not meant for production.. when the support comes from neutron, we will use that.. | 16:08 |
| jroll | right, ok | 16:08 |
| jroll | would you be interested in our current code that interacts with this? | 16:09 |
| NobodyCam | lol - this is cute: https://www.youtube.com/watch?v=YoB8t0B4jx4&t=35 | 16:09 |
| jroll | we still haven't put it in gerrit, because it likely won't land as-is | 16:09 |
| lazy_prince | yup.. | 16:09 |
| jroll | ok, I'll try to get that up next week | 16:10 |
| lucasagomes | NobodyCam, lol | 16:10 |
| lucasagomes | poor dog hah | 16:10 |
| NobodyCam | hahaha | 16:10 |
| jroll | lol that's awesome | 16:11 |
| * jroll goes shopping for one | 16:11 | |
| NobodyCam | :) | 16:11 |
| NobodyCam | looks like they had a lot of fun with it | 16:11 |
| lazy_prince | jroll: after running 'apt-get install libmysqlclient-dev' followed by 'python setup.py install' worked | 16:12 |
| jroll | cool :) | 16:12 |
| * jroll goes afk for a while | 16:12 | |
| lazy_prince | thanks for the guidance.. now will start testing.. | 16:12 |
| jroll | bbiab | 16:12 |
| lazy_prince | jroll: and will wait for the current code | 16:13 |
| jroll | ok :) | 16:13 |
| Shrews | NobodyCam: awww. that poor puppy | 16:14 |
| NobodyCam | [s]he seems to be having fun | 16:15 |
| Shrews | or will be left with an inferiority complex thinking no one likes her | 16:16 |
| jroll | lazy_prince: this is the very short version of how onmetal does it: https://gist.github.com/jimrollenhagen/9f80ba59d2a031a130f8 | 16:16 |
| *** vinbs has joined #openstack-ironic | 16:17 | |
| jroll | bbl | 16:18 |
| *** mordred has quit IRC | 16:18 | |
| *** mordred has joined #openstack-ironic | 16:18 | |
| *** vinbs has quit IRC | 16:22 | |
| lazy_prince | jroll: i am doing something similar.. i pulled it from the ironic patch written by JoshNag. thanks to him.. | 16:24 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Proxy Nova baremetal commands to Ironic https://review.openstack.org/116316 | 16:24 |
| JayF | lazy_prince: JoshNang works with both myself and jroll, so that's the same code pretty much :P | 16:24 |
| lazy_prince | JayF: something failed with neutron while running devstack. let me c.. | 16:26 |
| lazy_prince | here is the relevant devstack log http://paste.openstack.org/show/106629/ | 16:29 |
| *** ndipanoff has quit IRC | 16:30 | |
| lucasagomes | aight I will call it a day | 16:36 |
| lucasagomes | have a good night everybody! | 16:36 |
| lucasagomes | enjoy the weekend | 16:36 |
| NobodyCam | have a good weekend lucasagomes | 16:36 |
| *** lucasagomes is now known as lucas-dinner | 16:37 | |
| morgabra | lazy_prince: hey, sorry. Did you run setup.py at all? | 16:38 |
| *** eghobo has joined #openstack-ironic | 16:38 | |
| morgabra | lazy_prince: try 'python setup.py develop' | 16:38 |
| lazy_prince | hi morgabra | 16:38 |
| morgabra | it builds the config file that stevedore reads for drivers | 16:38 |
| lazy_prince | morgabra: i did python setup.py install.. and it solved my problem. | 16:39 |
| morgabra | cool, sorry for the late reply. hit me up if you run into anything else | 16:39 |
| lazy_prince | morgabra: jroll helped me get past. but i have another problem.. http://paste.openstack.org/show/106629/ | 16:40 |
| lazy_prince | morgabra: well.. in another 20mins i will call it a day.. till then i will continue to bug you.. | 16:41 |
| morgabra | neutron router-interface-add 8f96ab0c-da3d-49f0-8bfd-a1caa697fd0b 6380f776-ba25-4e4a-83a7-8c961ddc5eec fails, hmm | 16:41 |
| lazy_prince | have you seen this before..? | 16:42 |
| * NobodyCam does a quick run to starbuck..bbiaf | 16:42 | |
| morgabra | I haven't tried to run this in devstack at all yet | 16:42 |
| morgabra | I didn't change any extensions that ml2 loads | 16:43 |
| lazy_prince | oh.. by the way, i have intgrated your plugin with devstack and i am running it with devstack.. could that be the issue..? | 16:43 |
| morgabra | so if it is supposed to work for ml2 I'm not sure what's up | 16:43 |
| morgabra | well, maybe. Can you look in the neutron application log for the error? | 16:43 |
| lazy_prince | File "/usr/local/lib/python2.7/dist-packages/oslo/db/sqlalchemy/exc_filters.py", line 235, in _raise_mysql_table_doesnt_exist_asis | 16:44 |
| morgabra | hrm, I'm not sure what that's about tbh. | 16:45 |
| *** derekh has quit IRC | 16:45 | |
| morgabra | I know the latest version of neutron changed to not auto-sync db models | 16:45 |
| morgabra | so maybe devstack isn't using the one specified in requirements.txt | 16:46 |
| lazy_prince | let me get you complete trace.. | 16:46 |
| *** rushiagr is now known as rushiagr_away | 16:47 | |
| lazy_prince | morgabra: got the log. here it is http://paste.openstack.org/show/106636/ | 16:50 |
| morgabra | yeah, db didn't get synced | 16:50 |
| lazy_prince | can i do that manually..? | 16:51 |
| morgabra | you could try 'neutron-db-manage --config-file /path/to/neutron.conf upgrade head | 16:52 |
| morgabra | but if it were using the correct version of neutron it would create the tables automatically when the service starts | 16:53 |
| morgabra | (I need to update this at some point :/) | 16:53 |
| lazy_prince | aha.. i am using the Top of Trunk of Neutron.. | 16:53 |
| lazy_prince | can you let me know when you update it..? I will be waiting for it.. | 16:54 |
| lazy_prince | morgabra: now that you are here, let me ask you few things.. | 16:56 |
| lazy_prince | morgabra: https://github.com/openstack-dev/devstack/blob/master/lib/neutron_plugins/ml2#L51 | 16:56 |
| lazy_prince | do you know what values I should use for your plugin..? | 16:56 |
| lazy_prince | morgabra: and on L52 in the same file | 16:57 |
| *** penick has joined #openstack-ironic | 16:58 | |
| morgabra | hmm. I'm not sure how devstack works there. It's just pointing to another config file | 16:58 |
| lazy_prince | morgabra: yup.. but if i do not use config file devstack fails.. | 16:59 |
| lazy_prince | morgabra: and since you are extending ml2_plugin, it makes sense to use ml2 as base for devstack too.. | 17:00 |
| *** dguerri is now known as dguerri`afk | 17:00 | |
| lazy_prince | i could be worng too.. on 2nd thought, this is all neutron related.. should we switch to neutron channel so that others can also participate..? | 17:01 |
| *** viktors is now known as viktors|afk | 17:01 | |
| morgabra | well, this is kind of a one-off thing | 17:01 |
| lazy_prince | okay.. i will leave now.. can i catch you on Monday same time..? | 17:02 |
| *** penick has quit IRC | 17:02 | |
| morgabra | yeah I'll be around | 17:02 |
| lazy_prince | okay.. c ya on Monday then.. | 17:02 |
| *** openstackgerrit has quit IRC | 17:04 | |
| * NobodyCam is back | 17:05 | |
| *** harlowja_away is now known as harlowja | 17:05 | |
| *** jcoufal has quit IRC | 17:05 | |
| *** penick has joined #openstack-ironic | 17:11 | |
| *** HenryG is now known as HenryThe8th | 17:12 | |
| *** MattMan has left #openstack-ironic | 17:16 | |
| *** pelix has quit IRC | 17:17 | |
| *** dhellmann is now known as dhellmann_ | 17:23 | |
| *** lazy_prince is now known as killer_prince | 17:31 | |
| *** rushiagr_away is now known as rushiagr | 17:44 | |
| *** igordcard has joined #openstack-ironic | 17:52 | |
| *** kfox1111 has joined #openstack-ironic | 17:53 | |
| kfox1111 | so, you can have some nova compute's configured with ironic and some with kvm, then you can use host aggregates to associate flavors to hypervisors, and one nova schedular will then be happy with the setup? | 17:57 |
| *** dhellmann_ is now known as dhellmann | 18:02 | |
| *** eghobo has quit IRC | 18:04 | |
| NobodyCam | kfox1111: I have seen it work with compute zones, but that was like 8 months to a year ago | 18:06 |
| JayF | kfox1111: In our deployment, we segreate out different hypervisors into different nova cells, but afaik there's nothing that requires that other than making our management practices a bit more sane :) | 18:08 |
| *** penick has quit IRC | 18:09 | |
| *** openstackgerrit has joined #openstack-ironic | 18:11 | |
| *** penick has joined #openstack-ironic | 18:21 | |
| *** yjiang5 has joined #openstack-ironic | 18:23 | |
| yjiang5 | Hi, all, are there any administrator guide for ironic? I didn't find any in https://wiki.openstack.org/wiki/Ironic . (Is this the right channel, or I need to go to #openstack channel?) | 18:25 |
| NobodyCam | this is the Ironic channel | 18:25 |
| NobodyCam | yjiang5: have you seen http://docs.openstack.org/developer/ironic/#admin-guide | 18:26 |
| yjiang5 | NobodyCam: Aha, thanks. Didn't realize it's in the title of "developer documentation". Will have a look. | 18:27 |
| NobodyCam | ya, we'll get around to moving it around later, lol | 18:28 |
| yjiang5 | NobodyCam: cool, thanks. | 18:29 |
| kfox1111 | Ok. thanks for the help. | 18:30 |
| *** romcheg1 has joined #openstack-ironic | 18:38 | |
| *** ifarkas has quit IRC | 18:44 | |
| *** lucas-dinner has quit IRC | 18:45 | |
| * NobodyCam wanders afk for a few min | 18:54 | |
| *** openstackgerrit has quit IRC | 19:01 | |
| *** openstackgerrit has joined #openstack-ironic | 19:03 | |
| *** rushiagr is now known as rushiagr_away | 19:05 | |
| NobodyCam | anyone happen to know if BOOTx64.EFI is used for 32 and 64 bit arch's? | 19:10 |
| kfox1111 | whats the status of ironic integration/baremetal deprecation? | 19:14 |
| NobodyCam | kfox1111: as in... the nova driver has yet to land but we're very close. see https://review.openstack.org/#/c/111223 as the start of the nova patches | 19:17 |
| kfox1111 | ah. cool. | 19:17 |
| kfox1111 | can you run a juno ironic with an icehouse everything else? | 19:21 |
| NobodyCam | hummm, not sure, I think the nova api changed at some point that would break things | 19:24 |
| kfox1111 | ah. then I'll probably just wait for juno to play then... | 19:27 |
| *** penick has quit IRC | 19:27 | |
| NobodyCam | kfox1111: you can devstack for just playing with ironic | 19:31 |
| jroll | yeah, nova api has changed | 19:34 |
| kfox1111 | didn't meen that kind of playing. playing with a deployment with the intent of using it soonish. I believe it will work, just want to play with how to actually deploy it for use. | 19:34 |
| kfox1111 | jroll: thanks. | 19:34 |
| jroll | kfox1111: just run master of everything :D | 19:34 |
| kfox1111 | jroll: "What could go wrong (TM)" :) | 19:35 |
| jroll | :) | 19:35 |
| kfox1111 | "imagine everything on fire" ;) | 19:35 |
| jroll | it's really not that bad | 19:35 |
| kfox1111 | probably not. just been bit one too many times by bad updates. | 19:36 |
| *** andreykurilin_ has joined #openstack-ironic | 19:36 | |
| jroll | right | 19:36 |
| kfox1111 | "Oh yeah, we upgraded the database schema.... you need to sync that..." | 19:36 |
| jroll | yeah, good testing is required | 19:36 |
| kfox1111 | I'm interested in if something like solum or tripleo will help with that. | 19:37 |
| jroll | yeah, dunno | 19:38 |
| kfox1111 | Openstack's got great testing infrastructure, but it doesn't help to ensure what was tested, is whats tested when we deploy it. | 19:38 |
| jroll | right, you need to do downstream testing as well | 19:38 |
| kfox1111 | would be great if it was easy to reuse the testing infrastructure locally. | 19:38 |
| jroll | it's all in puppet and whatnot | 19:39 |
| kfox1111 | or alternately, tripleo tests things, and then releases the exact images they tested. | 19:39 |
| * jroll cringes | 19:39 | |
| JayF | It's a problem of basically unlimited space | 19:40 |
| JayF | people should be responsible enough with versioning that everything at any version works with everything at any version | 19:40 |
| JayF | but that's as likely to happen as world peace :) | 19:40 |
| JayF | so I'd maybe just wait until juno-rc1 is cut, use that to play around with, then for prod you can use juno | 19:40 |
| kfox1111 | jayf: you only need to keep n most up to date versions of the images or something. not every possible image. | 19:41 |
| JayF | or run master, like we (basically) do, but have to have stronger internal testing processes | 19:41 |
| JayF | kfox1111: if you're talking about the tripleo images idea, I really dislike that for a multitude of reasons | 19:41 |
| kfox1111 | yeah. | 19:41 |
| JayF | not the least of which being that tripleo is /a/ way to deploy Openstack not /the/ way to deploy Openstack :) | 19:41 |
| kfox1111 | jayf: sure. but that shouldn't preclude them from providing images for those who don't want to build them themselves. it lowers the bar to entry greatly. | 19:42 |
| JayF | I firmly disagree, but honestly, I dislike image-based deployment as a pattern generally :) | 19:42 |
| kfox1111 | you should be able to deploy openstack with puppet on redhat using rdo, or trunk on ubunto, or build your own images, or just get some prebuilt ones. all options are good in different ways. | 19:43 |
| *** penick has joined #openstack-ironic | 19:43 | |
| kfox1111 | JayF: we've been back and forth on it for a long time. both have advantages. | 19:44 |
| kfox1111 | images boot much faster, and easy to relyably get exactly what you want. | 19:44 |
| JayF | but lots and lots of overhead just to patch one thing | 19:44 |
| JayF | like, say, heartbleed | 19:44 |
| kfox1111 | kickstart style deployment is easier to deploy but more chance of random spring comes out. | 19:44 |
| *** chuckC has joined #openstack-ironic | 19:45 | |
| JayF | The magic isn't in using images | 19:45 |
| JayF | the magic is in the validation | 19:45 |
| kfox1111 | yeah. thats where something like disk image builder is very interesting. you can take one image, do a yum upgrade in it easily, and make another. you can then fix heartblead fairly easily while still doing images. | 19:45 |
| JayF | just by using images you can *try* to push validation work back upstream | 19:45 |
| *** andreykurilin_ has quit IRC | 19:45 | |
| *** andreykurilin_ has joined #openstack-ironic | 19:46 | |
| jroll | rolling upgrades are more complex with image | 19:46 |
| kfox1111 | I see the kickstart style image building, as absolutely needed in order to manage creating/updating images. images though have other advantages. I think using one to build the other has a lot of benifits. | 19:46 |
| kfox1111 | jroll: true. but ideally, you have a tool like heat that will do the rolling updates for you. | 19:47 |
| jroll | also the, 'how do I distribute updates to 40k hypervisor hosts?' question is a bit harder with full images | 19:47 |
| *** penick has quit IRC | 19:47 | |
| jroll | distribute updates quickly* | 19:48 |
| kfox1111 | true: but also harder the other direction too. which of my 40k hypervisor hosts got bad update X and need fixing? | 19:48 |
| NobodyCam | jroll: isn't that what tripleO addresses? | 19:48 |
| jroll | what bad update x? | 19:48 |
| jroll | don't push bad updates to production :P | 19:48 |
| jroll | or have a way to pluck the version out? | 19:48 |
| NobodyCam | no 40 k node updates | 19:48 |
| jroll | NobodyCam: what does tripleo address? | 19:48 |
| jroll | why? | 19:48 |
| jroll | I have 40k compute nodes, they all need new n-cpu code | 19:49 |
| jroll | right? | 19:49 |
| NobodyCam | ya so perserve ephermal and update them | 19:49 |
| NobodyCam | :-p | 19:49 |
| NobodyCam | (/me is giving jroll a hard time) | 19:50 |
| JayF | that was a high quality troll | 19:50 |
| JayF | I felt my blood pressure going up | 19:50 |
| NobodyCam | hehehe | 19:50 |
| NobodyCam | :-p | 19:50 |
| jroll | lolol | 19:50 |
| NobodyCam | it is friday | 19:50 |
| jroll | high five | 19:50 |
| kfox1111 | There was an idea floated at some point that was a light weight image update. rather then do a full image replacement, patch it with the diff between two images. kind of half way inbetween. | 19:51 |
| kfox1111 | jroll: ideally you would never push out a bad version of something. stuff happens though. :/ | 19:51 |
| kfox1111 | :) | 19:51 |
| jroll | right, have a way to get a version | 19:52 |
| *** chuckC has quit IRC | 19:52 | |
| *** chuckC has joined #openstack-ironic | 19:53 | |
| *** athomas has quit IRC | 20:02 | |
| kfox1111 | we currently use cobbler + anaconda to deploy bare metal nodes. we use cobbler to sync rpmsets and have production/testing repo sets. going forward is pretty easy with this setup, but if we ever have to roll something back, its kind of a pain. | 20:05 |
| *** penick has joined #openstack-ironic | 20:05 | |
| NobodyCam | kfox1111: I've interviewed with folk who never rolled back only forward (thou that was many years ago) | 20:09 |
| jroll | ^^ | 20:09 |
| jroll | roll forward or bust | 20:09 |
| NobodyCam | actually they had a pretty cool roll out | 20:09 |
| kfox1111 | we often will push forward since its so hard to roll back. but it causes outages that are longer then they should be. can be hours instead of minutes. thats the drawback to that approach. when something unexpected goes wrong, it is down time in which you have to figure out what unexpected happened, rather then roll back to a known state quickly, and take nondown time to figure it out. | 20:11 |
| NobodyCam | the code did automatic a/b testing, so all servers had all the code but only some would run the new code and if blewup theyd just stop serving that rev | 20:12 |
| kfox1111 | ah. a/b is a kind of image. ;) | 20:12 |
| NobodyCam | kfox1111: http://en.wikipedia.org/wiki/A/B_testing | 20:13 |
| kfox1111 | yeah. | 20:16 |
| *** romcheg1 has quit IRC | 20:18 | |
| jroll | adam_g: check out https://review.openstack.org/#/c/112693 when you have a chance? | 20:54 |
| jroll | adam_g: relies on https://review.openstack.org/#/c/119473 | 20:54 |
| openstackgerrit | linggao proposed a change to openstack/ironic: Minor fixes for ipminative console support https://review.openstack.org/119475 | 20:54 |
| adam_g | jroll, sure will take a look in a bit | 20:55 |
| jroll | thanks :) | 20:55 |
| jroll | just want to make sure that does what I think it does (see commit message) | 20:55 |
| *** r-daneel has quit IRC | 21:12 | |
| kfox1111 | so, how does cells and neutron work. do you have a neutron per cell or is it shared. Is that an issue with ironic and kvm in the same cloud? | 21:24 |
| *** openstackgerrit has quit IRC | 21:31 | |
| *** openstackgerrit has joined #openstack-ironic | 21:33 | |
| adam_g | jroll, looks good but i think you've forgotten to add check-tempest-dsvm-ironic-agent_ssh-src to projects.yaml? | 21:34 |
| *** andreykurilin_ has quit IRC | 21:43 | |
| *** eghobo has joined #openstack-ironic | 22:02 | |
| NobodyCam | gah friday if fighting to distract me at every turn | 22:09 |
| jroll | adam_g: exactly why I was looking for more eyes | 22:17 |
| jroll | NobodyCam: you have no idea, so many context switches today :/ | 22:17 |
| NobodyCam | :( | 22:25 |
| NobodyCam | just fyi: I am here in but brain is shutting down :-p | 22:26 |
| *** dlaube has quit IRC | 22:29 | |
| lifeless | romcheg: hi, you wanted me? | 22:40 |
| mikal | Please brace for impact | 22:50 |
| jroll | :o | 22:50 |
| greghaynes | mikal: O_o | 22:50 |
| NobodyCam | :/ | 22:50 |
| jroll | oshi | 22:50 |
| jroll | last nova patch has +2 | 22:50 |
| jroll | err | 22:50 |
| jroll | 2x +2 | 22:50 |
| jroll | this is happening | 22:50 |
| NobodyCam | omg no | 22:50 |
| NobodyCam | devananda: ^^^^^^^^^^ | 22:51 |
| jroll | everything does | 22:51 |
| JayF | http://i.kinja-img.com/gawker-media/image/upload/s--Gwlhw4RX--/qtr9wvgsujidg5zri4im.gif | 22:51 |
| mikal | NobodyCam: want me to -2?!? | 22:51 |
| * jroll hands mikal the giant red button | 22:51 | |
| * NobodyCam looks for his GUN | 22:51 | |
| mikal | NobodyCam: https://twitter.com/dolphm/status/505787977512275968/photo/1 | 22:52 |
| jroll | nice | 22:55 |
| jroll | my kind of party | 22:55 |
| mikal | Oh, good news | 22:56 |
| mikal | This is going to fail | 22:56 |
| mikal | And its my fault | 22:56 |
| NobodyCam | mikal: lol | 22:56 |
| mikal | Remember, vote one flailing guy | 22:56 |
| *** harlowja is now known as harlowja_away | 23:00 | |
| mikal | Ok, new plan | 23:01 |
| mikal | We're going to have a nap instead | 23:01 |
| mikal | Or bacon | 23:01 |
| mikal | One or the other | 23:02 |
| jroll | why you gotta tease us like that | 23:02 |
| kfox1111 | bacon! :) | 23:02 |
| NobodyCam | ya the answer has to be "Bacon" | 23:05 |
| mikal | jroll: cause I hate freedome | 23:05 |
| mikal | Also, IT BEGINS AGAIN | 23:05 |
| jroll | lol | 23:06 |
| adam_g | http://iheartjimmy.com/wp-content/uploads/2012/07/tumblr_m4zmrtA1Bf1qheqxz.gif | 23:07 |
| NobodyCam | lol adam_g that is so 70's | 23:08 |
| * mikal wanders off now that he's done a thing | 23:10 | |
| *** eguz has joined #openstack-ironic | 23:10 | |
| *** eghobo has quit IRC | 23:13 | |
| jroll | mikal: why's it going to fail? | 23:14 |
| jroll | (other than it's running openstack CI) | 23:15 |
| *** dividehex has quit IRC | 23:15 | |
| NobodyCam | jroll: boo | 23:16 |
| *** igordcard has quit IRC | 23:16 | |
| NobodyCam | more like because it has to make its way through the gate | 23:16 |
| mikal | jroll: the first time had a pep8 fail in my blocker patch | 23:16 |
| mikal | jroll: the second time around fixes that | 23:16 |
| NobodyCam | gate / gauntlet | 23:16 |
| jroll | mikal: lol. | 23:16 |
| * mikal runs away to eat the bacon of victory | 23:18 | |
| mikal | In the muffin of pep8 shame | 23:18 |
| *** dividehex has joined #openstack-ironic | 23:21 | |
| NobodyCam | lol | 23:22 |
| *** penick has quit IRC | 23:23 | |
| *** harlowja_away is now known as harlowja | 23:31 | |
| mrda | Just to chime in | 23:37 |
| mrda | \o/ | 23:37 |
| * mrda wonders whether he should weekend, or sit here watching +A'd patches make their way through check and gate | 23:42 | |
| jroll | it's da weekend | 23:42 |
| mrda | jroll: is that a question or a statement? | 23:44 |
| jroll | statement | 23:44 |
| jroll | I am like... 20 minutes away from weekend | 23:45 |
| mrda | but this is my baby :) | 23:45 |
| jroll | hehe | 23:45 |
| jroll | let mikal babysit it for the weekend :) | 23:45 |
| mrda | And I have 2 hours until I have to coach a basketball team :) | 23:45 |
| jroll | lol if you think it can get through any check/gate jobs in two hours | 23:45 |
| mrda | I can only hope | 23:46 |
| mrda | maybe check | 23:46 |
| mrda | for a few of them | 23:46 |
| mrda | check is under 90 minutes right now | 23:46 |
| mrda | o.O | 23:46 |
| jroll | yeah, but the gate was super backed up last I checked | 23:48 |
| mrda | nah, gate is only 16 hours! | 23:49 |
| jroll | only. | 23:49 |
| mrda | better than 30+ hours for I | 23:49 |
| jroll | I guess | 23:49 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
