*** Sukhdev has quit IRC | 00:03 | |
*** achanda_ has joined #openstack-ironic | 00:03 | |
*** david-ly_ has joined #openstack-ironic | 00:04 | |
openstackgerrit | Merged openstack/ironic-python-agent: Update gitignore https://review.openstack.org/232345 | 00:04 |
---|---|---|
*** david-lyle has quit IRC | 00:05 | |
*** mbound has quit IRC | 00:06 | |
*** achanda has quit IRC | 00:06 | |
*** Sukhdev has joined #openstack-ironic | 00:07 | |
*** achanda_ has quit IRC | 00:07 | |
*** Sukhdev has quit IRC | 00:10 | |
*** VikasC has joined #openstack-ironic | 00:12 | |
*** morgan is now known as mordgan | 00:16 | |
*** mordgan is now known as morgan | 00:16 | |
*** mordred is now known as morgred | 00:16 | |
*** morgan is now known as mordan | 00:16 | |
*** mordan is now known as mordgan | 00:16 | |
*** shadower has joined #openstack-ironic | 00:23 | |
*** morgred is now known as morgblue | 00:34 | |
*** sdake has joined #openstack-ironic | 00:35 | |
*** morgblue is now known as morgred | 00:36 | |
*** olaph is now known as doseedo | 01:03 | |
*** VikasC has quit IRC | 01:08 | |
openstackgerrit | Merged openstack/ironic: Use power manager to reboot in agent deployments https://review.openstack.org/232157 | 01:12 |
* devananda realises it's after 6pm and he still hasn't gotten around to fixing up some patches | 01:13 | |
jroll | devananda: I *just* left a comment that was "we can do this as long as there isn't hardware that behaves like <that patch>" | 01:13 |
jroll | (on another thing) | 01:13 |
*** baoli has joined #openstack-ironic | 01:14 | |
devananda | jroll: huh? | 01:14 |
jroll | devananda: the thing that just merged. | 01:14 |
devananda | ahh | 01:14 |
jroll | not quite the same behavior, but that also breaks my assumption :P | 01:14 |
*** david-ly_ has quit IRC | 01:14 | |
jroll | devananda: also, if you don't get a chance to deal with that FakeBoot thing tonight, let me know and I'll handle it in the AM | 01:15 |
devananda | will do | 01:16 |
devananda | I have about 20 minutes right now - it's either try to book a hotel using IBM's travel thing, or do that | 01:16 |
jroll | is it carlson wagonlit? | 01:16 |
devananda | nope | 01:16 |
jroll | well, maybe that's doable | 01:16 |
devananda | hehe | 01:16 |
jroll | :P | 01:16 |
jroll | anyway, out for the night, have a good one :) | 01:17 |
devananda | ciao! | 01:18 |
*** dims has quit IRC | 01:18 | |
*** lazy_prince has joined #openstack-ironic | 01:18 | |
*** dims has joined #openstack-ironic | 01:18 | |
*** david-lyle has joined #openstack-ironic | 01:22 | |
*** dims has quit IRC | 01:23 | |
jlvillal | lifeless, Have you heard of any other bugs detected because of new Fixture release? | 01:24 |
*** baoli has quit IRC | 01:25 | |
*** doseedo is now known as olaph | 01:26 | |
*** david-lyle has quit IRC | 01:29 | |
lifeless | jlvillal: nothing else has reached my easrs | 01:29 |
*** lazy_prince has quit IRC | 01:40 | |
*** naohirot has joined #openstack-ironic | 01:40 | |
*** lazy_prince has joined #openstack-ironic | 01:40 | |
*** baoli has joined #openstack-ironic | 01:43 | |
*** sdake has quit IRC | 01:44 | |
openstackgerrit | shravya Gaddam proposed openstack/ironic: ucs_hostname changed to ucs_address https://review.openstack.org/232834 | 01:45 |
*** jamielennox is now known as jamielennox|away | 01:46 | |
*** bnemec has quit IRC | 01:48 | |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - RPC https://review.openstack.org/206243 | 01:50 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 01:50 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - net https://review.openstack.org/206245 | 01:50 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - DB https://review.openstack.org/206232 | 01:50 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - objs https://review.openstack.org/206238 | 01:50 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add network provider interface and implementations https://review.openstack.org/139687 | 01:50 |
jlvillal | lifeless, Thanks | 01:55 |
*** bnemec has joined #openstack-ironic | 01:59 | |
*** baoli has quit IRC | 01:59 | |
*** baoli has joined #openstack-ironic | 02:02 | |
*** ijw has quit IRC | 02:05 | |
*** lazy_prince has quit IRC | 02:05 | |
*** dims has joined #openstack-ironic | 02:19 | |
*** penick has quit IRC | 02:21 | |
*** frontrunner has quit IRC | 02:21 | |
*** dims has quit IRC | 02:24 | |
*** harshs has joined #openstack-ironic | 02:24 | |
*** achanda has joined #openstack-ironic | 02:31 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic: Add secret=True to password option https://review.openstack.org/232843 | 02:35 |
*** davideagnello has quit IRC | 02:47 | |
*** davideagnello has joined #openstack-ironic | 02:54 | |
*** harshs has quit IRC | 03:29 | |
*** Nisha has joined #openstack-ironic | 03:34 | |
*** david-lyle has joined #openstack-ironic | 03:47 | |
*** sdake has joined #openstack-ironic | 03:51 | |
*** Nisha has quit IRC | 03:52 | |
*** links has joined #openstack-ironic | 03:53 | |
*** achanda has quit IRC | 03:57 | |
*** baoli has quit IRC | 03:58 | |
*** VikasC has joined #openstack-ironic | 04:00 | |
*** chlong has joined #openstack-ironic | 04:00 | |
*** chlong has quit IRC | 04:05 | |
*** dims has joined #openstack-ironic | 04:08 | |
*** Nisha has joined #openstack-ironic | 04:08 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Fixes logging of failure in deletion of swift temporary object https://review.openstack.org/232440 | 04:17 |
*** chlong has joined #openstack-ironic | 04:18 | |
*** sdake_ has joined #openstack-ironic | 04:41 | |
*** achanda has joined #openstack-ironic | 04:41 | |
*** sdake has quit IRC | 04:44 | |
*** rameshg87 has joined #openstack-ironic | 04:49 | |
*** dims has quit IRC | 04:51 | |
openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: Add support for emulation with VirtualBox https://review.openstack.org/155647 | 04:52 |
*** harshs has joined #openstack-ironic | 04:59 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/232870 | 05:00 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-lib: Updated from global requirements https://review.openstack.org/232871 | 05:00 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/232872 | 05:00 |
openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: Reduce logical disk size returned by 1 https://review.openstack.org/232880 | 05:01 |
*** sdake_ has quit IRC | 05:01 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/232892 | 05:04 |
*** sdake has joined #openstack-ironic | 05:07 | |
*** Marga__ has joined #openstack-ironic | 05:29 | |
*** jefabo has quit IRC | 05:30 | |
*** chlong has quit IRC | 05:31 | |
rameshg87 | congrats jlvillal vdrok (if you guys are around) :) | 05:32 |
jlvillal | rameshg87, Thank you very much :) Was a very nice surprise for me this week. Great to be nominated! | 05:32 |
*** Marga_ has quit IRC | 05:33 | |
rameshg87 | jlvillal: enjoy your surprise with beer ;-) | 05:33 |
jlvillal | rameshg87, I'll have a drink tomorrow night. Friday! :) | 05:34 |
*** Marga__ has quit IRC | 05:34 | |
rameshg87 | jlvillal: oh yeah, so it's still Thursday for you. it's already Friday for me :) | 05:34 |
rameshg87 | jlvillal: but thanks a lot for all your work in Ironic once again | 05:35 |
jlvillal | rameshg87, Well enjoy your weekend, if I don't chat with you before you leave. | 05:35 |
jlvillal | rameshg87, My pleasure. The Ironic group of people is a great group of people to work with! | 05:35 |
rameshg87 | jlvillal: sure same to you too..good night and have a great weekend | 05:35 |
jlvillal | rameshg87, Thanks. Bye | 05:36 |
rameshg87 | jlvillal: bye..see you | 05:36 |
*** harshs has quit IRC | 05:39 | |
*** dims has joined #openstack-ironic | 05:52 | |
*** pas-ha has joined #openstack-ironic | 06:10 | |
*** Sukhdev has joined #openstack-ironic | 06:19 | |
*** Marga_ has joined #openstack-ironic | 06:22 | |
*** Marga__ has joined #openstack-ironic | 06:23 | |
*** Marga_ has quit IRC | 06:23 | |
*** dims has quit IRC | 06:25 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Imported Translations from Zanata https://review.openstack.org/232930 | 06:31 |
*** Sukhdev has quit IRC | 06:33 | |
*** sdake has quit IRC | 06:33 | |
*** chlong has joined #openstack-ironic | 06:39 | |
*** max_lobur1 has joined #openstack-ironic | 06:40 | |
*** max_lobur1 has left #openstack-ironic | 06:40 | |
*** mrmoje has joined #openstack-ironic | 06:47 | |
*** dims has joined #openstack-ironic | 06:49 | |
*** dims has quit IRC | 06:54 | |
*** yog_ has joined #openstack-ironic | 06:59 | |
*** romainh has joined #openstack-ironic | 07:01 | |
*** Marga__ has quit IRC | 07:03 | |
*** romainh has quit IRC | 07:16 | |
*** romainh has joined #openstack-ironic | 07:16 | |
*** romainh has left #openstack-ironic | 07:18 | |
*** mrmoje has quit IRC | 07:19 | |
*** mrmoje has joined #openstack-ironic | 07:21 | |
*** mkovacik has quit IRC | 07:24 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Fixes logging of failure in deletion of swift temporary object https://review.openstack.org/232440 | 07:28 |
*** ifarkas has joined #openstack-ironic | 07:41 | |
*** marzif has joined #openstack-ironic | 07:45 | |
*** dims has joined #openstack-ironic | 07:50 | |
*** dtantsur|afk is now known as dtantsur | 07:52 | |
dtantsur | Morning Ironic | 07:54 |
*** mbound has joined #openstack-ironic | 07:56 | |
*** dims has quit IRC | 07:56 | |
*** MattMan has quit IRC | 07:57 | |
*** MattMan has joined #openstack-ironic | 07:58 | |
*** athomas has joined #openstack-ironic | 07:59 | |
openstackgerrit | Anton Arefiev proposed openstack/python-ironicclient: Use requests lib in HTTPClient https://review.openstack.org/229286 | 07:59 |
*** openstackstatus has quit IRC | 08:02 | |
*** openstackstatus has joined #openstack-ironic | 08:03 | |
*** ChanServ sets mode: +v openstackstatus | 08:03 | |
sambetts | Congrats jlvillal and vdrok !!!! | 08:04 |
dtantsur | :) | 08:09 |
dtantsur | morning sambetts | 08:09 |
sambetts | o/ | 08:09 |
*** chlong has quit IRC | 08:09 | |
sambetts | dtantsur: I need to do more reviews :-P | 08:09 |
dtantsur | always good thing to do ;) | 08:10 |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Prevent iRMC unit test from potential failure at the gate https://review.openstack.org/232946 | 08:13 |
*** achanda has quit IRC | 08:18 | |
*** lucasagomes has joined #openstack-ironic | 08:19 | |
*** e0ne has joined #openstack-ironic | 08:20 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor IRMCVirtualMediaIscsiDeploy by applying new BootInterface https://review.openstack.org/221371 | 08:20 |
rameshg87 | Nisha: hello | 08:21 |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor IRMCVirtualMediaAgentDeploy by applying new BootInterface https://review.openstack.org/221577 | 08:21 |
rameshg87 | Nisha: why can't we just call ilo_common.cleanup_vmedia_boot() in https://review.openstack.org/#/c/232323/2/ironic/drivers/modules/ilo/deploy.py | 08:22 |
rameshg87 | Nisha: we do that today when ipa ramdisk is booted in iscsi_ilo driver. so why not in agent_ilo driver ? | 08:22 |
vdrok | morning ironic | 08:22 |
vdrok | and morning dtantsur, sambetts, rameshg87 :) | 08:23 |
rameshg87 | morning vdrok o/ | 08:23 |
rameshg87 | vdrok and congrats too :) | 08:23 |
dtantsur | vdrok, o/ congrats on being nominated | 08:23 |
vdrok | sambetts, rameshg87, JoshNang, thiagop - thanks :) | 08:23 |
vdrok | dtantsur, thank you too :) | 08:24 |
vdrok | and congrats to jlvillal | 08:24 |
sambetts | o/ vdrok | 08:24 |
openstackgerrit | Merged openstack/ironic-python-agent: Move oslotest to test-requirements https://review.openstack.org/231880 | 08:26 |
romcheg | Hi fellas! | 08:27 |
*** mdbooth has quit IRC | 08:27 | |
vdrok | morning romcheg | 08:27 |
romcheg | lucasagomes, dtantsur: have you seen we've found the cause of that issue in testr with lifeless :) | 08:28 |
lifeless | romcheg: oh ? | 08:28 |
romcheg | lifeless: the unicode :) | 08:28 |
lifeless | romcheg: oh, so I still don't know the root cause | 08:29 |
lifeless | romcheg: we know the proximate cause and a decent workaround | 08:29 |
romcheg | no no, it's not the root cause | 08:29 |
dtantsur | so what was it? :) | 08:29 |
romcheg | lifeless: correct, I've used wrong words | 08:29 |
romcheg | dtantsur: the workaround was to supress all output to stdout | 08:29 |
dtantsur | does not sound like a good fix to be honest | 08:30 |
romcheg | this is basically why we don't call it a fix but a workaround :) | 08:30 |
dtantsur | our default testing system is already pretty quiet when it comes to outputting stuff... | 08:30 |
romcheg | so if you have similar issues, try that | 08:31 |
dtantsur | (before we switched to os-testr, this one is pretty chatty) | 08:31 |
dtantsur | romcheg, thanks for information anyway, will keep in mind | 08:31 |
lucasagomes | cool | 08:32 |
*** mdbooth has joined #openstack-ironic | 08:32 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add hardware inspection module for iRMC driver https://review.openstack.org/196480 | 08:36 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add documentation for Ceph Object Gateway support https://review.openstack.org/232068 | 08:39 |
*** jistr has joined #openstack-ironic | 08:41 | |
*** Nisha has quit IRC | 08:43 | |
*** subscope has joined #openstack-ironic | 08:44 | |
*** derekh has joined #openstack-ironic | 08:51 | |
lifeless | dtantsur: its the same capture mechanism all the other openstack projects use; | 08:51 |
lifeless | dtantsur: you get the output associated with the test in failures etc | 08:51 |
dtantsur | ack | 08:52 |
*** dims has joined #openstack-ironic | 08:53 | |
*** mkovacik has joined #openstack-ironic | 08:54 | |
*** aswadr has joined #openstack-ironic | 08:55 | |
*** dims has quit IRC | 08:58 | |
*** dtantsur is now known as dtantsur|brb | 09:08 | |
*** pelix has joined #openstack-ironic | 09:08 | |
*** marzif has quit IRC | 09:10 | |
*** jcoufal has joined #openstack-ironic | 09:19 | |
*** subscope has quit IRC | 09:32 | |
*** subscope has joined #openstack-ironic | 09:33 | |
*** athomas has quit IRC | 09:42 | |
*** athomas has joined #openstack-ironic | 09:49 | |
*** subscope has quit IRC | 09:52 | |
*** naohirot has quit IRC | 09:56 | |
*** marzif has joined #openstack-ironic | 10:00 | |
*** ndipanov has joined #openstack-ironic | 10:01 | |
*** zhenguo has quit IRC | 10:03 | |
*** amotoki has joined #openstack-ironic | 10:05 | |
*** fandi has quit IRC | 10:05 | |
*** ionutbalutoiu has joined #openstack-ironic | 10:07 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for partition images in agent driver https://review.openstack.org/162008 | 10:11 |
*** chlong has joined #openstack-ironic | 10:25 | |
openstackgerrit | Merged openstack/ironic-lib: Updated from global requirements https://review.openstack.org/232871 | 10:26 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: iPXE: Fix assumption that ${mac} is the MAC of the NIC it's booting https://review.openstack.org/232998 | 10:30 |
lucasagomes | rare problem with iPXE, but we just saw it happening ^. Someone can take a look please? | 10:31 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: iPXE: Fix assumption that ${mac} is the MAC of the NIC it's booting https://review.openstack.org/232998 | 10:33 |
* lucasagomes fix error message typo | 10:33 | |
*** dtantsur|brb is now known as dtantsur | 10:37 | |
openstackgerrit | Tan Lin proposed openstack/ironic-specs: Add `memdisk_boot` and support boot Windows PreInstallation Environment(WinPE) https://review.openstack.org/220889 | 10:47 |
*** ionutbalutoiu has quit IRC | 10:52 | |
*** krtaylor has quit IRC | 10:53 | |
*** krtaylor has joined #openstack-ironic | 10:53 | |
rameshg87 | dtantsur: for gate job in https://review.openstack.org/#/c/184443/ to use ironic-lib>=0.3.0, the patch to bump up g-r version should merge. right ? | 10:59 |
*** subscope has joined #openstack-ironic | 10:59 | |
rameshg87 | but did we have big enough changes between 0.2.0 and 0.3.0 ? | 10:59 |
rameshg87 | wondering why it's failing :) | 10:59 |
dtantsur | rameshg87, we did, a couple of functions changed their signature | 11:00 |
rameshg87 | dtantsur: okay, then wait is the only option | 11:00 |
dtantsur | rameshg87, ironic-lib bump patch is waiting for reviews: https://review.openstack.org/#/c/231368/2 | 11:00 |
dtantsur | you can ping someone ;) | 11:00 |
*** faizan has joined #openstack-ironic | 11:00 | |
rameshg87 | yeah I saw .. :) | 11:00 |
rameshg87 | faizan is here. | 11:00 |
*** ionutbalutoiu has joined #openstack-ironic | 11:01 | |
rameshg87 | faizan: the g-r patch needs to merge as you told | 11:01 |
faizan | yes, that is correct, | 11:01 |
faizan | rameshg87: even with the earlier version it should have worked, but it seems that is_iscsi_device(dev) is not working as expected | 11:03 |
rameshg87 | faizan: yeah it's checking lun13 | 11:04 |
rameshg87 | something like that ? | 11:04 |
faizan | rameshg87: this code has been removed from ironic-lib in the latest version, which is required when we use ironic-lib in agent-context | 11:05 |
rameshg87 | okay | 11:06 |
*** e0ne has quit IRC | 11:08 | |
dtantsur | faizan, IIRC is_iscsi_device was removed because it wasn't present in ironic | 11:13 |
*** chlong has quit IRC | 11:15 | |
* rameshg87 goes home | 11:15 | |
rameshg87 | happy weekend everyone | 11:15 |
*** rameshg87 has quit IRC | 11:15 | |
sambetts | o/ rameshg87 | 11:16 |
*** subscope has quit IRC | 11:20 | |
*** subscope has joined #openstack-ironic | 11:22 | |
openstackgerrit | Merged openstack/ironic: Add secret=True to password option https://review.openstack.org/232843 | 11:22 |
*** e0ne has joined #openstack-ironic | 11:24 | |
*** subscope has quit IRC | 11:29 | |
*** lucasagomes is now known as lucas-hungry | 11:31 | |
*** subscope has joined #openstack-ironic | 11:34 | |
*** thrash|g0ne is now known as thrash | 11:37 | |
*** subscope has quit IRC | 11:44 | |
*** subscope has joined #openstack-ironic | 11:45 | |
*** lazy_prince has joined #openstack-ironic | 11:50 | |
*** yog_ has quit IRC | 11:54 | |
TheJulia | good morning | 11:54 |
sinval | TheJulia, good morning o/ | 11:56 |
openstackgerrit | Merged openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/232892 | 12:00 |
*** baoli has joined #openstack-ironic | 12:02 | |
*** ukalifon has joined #openstack-ironic | 12:02 | |
*** trown|outttypeww is now known as trown | 12:05 | |
*** dprince has joined #openstack-ironic | 12:07 | |
*** baoli_ has joined #openstack-ironic | 12:08 | |
* sambetts is trying to workout how its possible for IPA to attempt to iscsi to /dev/sdb when sdb doesn't exist when you run an lsblk on the ramdisk | 12:10 | |
*** derekh has quit IRC | 12:10 | |
dtantsur | Oo | 12:10 |
sambetts | its failing because its saying the blocks are out of range, and I'm like of course they are the device doesn't exist! | 12:10 |
*** baoli has quit IRC | 12:11 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: WIP: Additional inspection role and adding role to NV job script https://review.openstack.org/232746 | 12:12 |
sambetts | dtantsur: Any ideas? | 12:14 |
dtantsur | sambetts, no root device hints are used? | 12:15 |
openstackgerrit | Merged openstack/ironic: PXE: Support Extra DHCP Options for IPv6 https://review.openstack.org/231573 | 12:16 |
sambetts | dtantsur: nope | 12:16 |
*** dims has joined #openstack-ironic | 12:16 | |
openstackgerrit | Merged openstack/ironic: iPXE: Fix assumption that ${mac} is the MAC of the NIC it's booting https://review.openstack.org/232998 | 12:17 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/232870 | 12:19 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/232872 | 12:19 |
*** VikasC has quit IRC | 12:24 | |
TheJulia | ugh, ironic now overwrites the boot.ipxe file. | 12:25 |
TheJulia | not good | 12:25 |
*** chlong has joined #openstack-ironic | 12:26 | |
dtantsur | hmmm, I think we were overwriting it for some time already? | 12:27 |
TheJulia | well, definitly overwrites now | 12:27 |
TheJulia | I guess the solution for people with weird pxe environments would be overwrite the install's boot.ipxe file, or define it directly | 12:28 |
TheJulia | or, chmod -w | 12:28 |
dtantsur | I think someone had a patch to change it, but it was mostly rejected | 12:28 |
*** subscope has quit IRC | 12:29 | |
dtantsur | lucas-hungry might remember | 12:29 |
TheJulia | Yeah, I could have sworn there was a patch | 12:29 |
*** subscope has joined #openstack-ironic | 12:29 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic: Fix tests giving erroneous output during os-testr run https://review.openstack.org/233028 | 12:30 |
TheJulia | yup, got changed on 9/2 to just copy the file that is defined in the config | 12:30 |
TheJulia | :( | 12:30 |
TheJulia | at least, thats the last day the line was touched | 12:31 |
*** subscope has quit IRC | 12:31 | |
*** lucas-hungry is now known as lucasagomes | 12:32 | |
* lucasagomes reads | 12:32 | |
TheJulia | yeah | 12:33 |
*** subscope has joined #openstack-ironic | 12:33 | |
TheJulia | I must have had it chmod -w for a long time in bifrost | 12:33 |
TheJulia | since the initial behavior change and revert were end of august/early september | 12:34 |
TheJulia | were at | 12:34 |
lucasagomes | TheJulia, hi there, the patch to not overwrite the boot.ipxe was reverted | 12:34 |
*** lazy_prince has quit IRC | 12:35 | |
TheJulia | yeah, I was looking at it backward | 12:35 |
TheJulia | s | 12:35 |
TheJulia | this is weird | 12:35 |
lucasagomes | I think the reasoning was that, if we don't overwrite and we update boot.ipxe in ironic ot fix something | 12:35 |
lucasagomes | that fix won't be applied on someone's env when they update ironic | 12:35 |
TheJulia | no, I got it correctly | 12:35 |
*** Marga_ has joined #openstack-ironic | 12:35 | |
lucasagomes | TheJulia, did the last change to boot.ipxe break bifrost? | 12:35 |
TheJulia | kind of broke bifrost, initial deploys work, but we also updated our boot.ipxe little | 12:36 |
TheJulia | lucasagomes: we basically had a case where we were running into old ipxe firmwares that did not fully support the feature set, so we actually try to detect that and load ipxe from the server | 12:37 |
lucasagomes | TheJulia, ok, so it wasn't using the upstream version? What you can do is have a different boot script in bifrost (tho it would be good to unify) and change it in the config file | 12:38 |
lucasagomes | TheJulia, oh right, chainload a new version of it? | 12:38 |
TheJulia | yup | 12:38 |
lucasagomes | TheJulia, oh I see here https://github.com/openstack/bifrost/blob/6840c963a28472675a96eebe6394947874ad79f6/playbooks/roles/bifrost-ironic-install/files/boot.ipxe#L3-L4 | 12:39 |
TheJulia | we had to do the same with dhcp config for cards that claimed ipxe support but were really just gpxe | 12:39 |
TheJulia | I think it actually kind of helps the ironic community to have an alternative example, its not an easy thing to wrap one's head around since it is fairly complex. Just saw your other fix and pondering it for bifrost | 12:40 |
lucasagomes | TheJulia, right, yeah it can get really complex | 12:41 |
lucasagomes | specially when we have old and new versions of iPXE | 12:41 |
lucasagomes | for gPXE vs iPXE you can indentify it by looking at a tag in the DHCP request (will look up the name) | 12:42 |
TheJulia | lucasagomes: exactly what we did for bifrost | 12:42 |
*** dims is now known as dimsum__ | 12:42 | |
lucasagomes | TheJulia, but anyway I'm +1 to have other examples, ipxe script can get really complex and yeah it's hard to have a common ground sometimes | 12:43 |
lucasagomes | "user-class" is the DHCP tag to identify gPXE vs iPXE | 12:45 |
*** morgred is now known as mordred | 12:46 | |
TheJulia | lucasagomes: http://git.openstack.org/cgit/openstack/bifrost/tree/playbooks/roles/bifrost-ironic-install/templates/dnsmasq.conf.j2#n438 :) | 12:48 |
lucasagomes | yup | 12:48 |
*** ukalifon has quit IRC | 12:50 | |
*** ukalifon has joined #openstack-ironic | 12:50 | |
*** marzif has quit IRC | 12:52 | |
openstackgerrit | Syed Ismail Faizan Barmawer proposed openstack/ironic: Refactor disk partitioner code from ironic and use ironic-lib. https://review.openstack.org/184443 | 12:53 |
*** marzif has joined #openstack-ironic | 12:53 | |
jroll | good morning y'all :) | 12:53 |
dtantsur | morning jroll | 12:55 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Fix boot.ipxe config behavior to ensure proper use https://review.openstack.org/233039 | 12:57 |
TheJulia | Good morning jroll | 12:57 |
*** albertoffb has joined #openstack-ironic | 12:59 | |
*** frontrunner has joined #openstack-ironic | 13:00 | |
sambetts | o/ jroll | 13:00 |
jroll | morning sambetts, TheJulia, dtantsur, lucasagomes \o | 13:00 |
lucasagomes | jroll, yo! good morning | 13:01 |
lucasagomes | TheJulia, ++ yeah I think that's the right fix for those using a custom script | 13:02 |
lucasagomes | TheJulia, in any case, depending on how old is the iPXE ROMs you have you may want to copy that loop from upstream as well | 13:02 |
lucasagomes | TheJulia, see here too http://git.ipxe.org/ipxe.git/commitdiff/66ea458 | 13:02 |
TheJulia | lucasagomes: Thanks, I was already thinking that. | 13:03 |
sambetts | jroll: Is anyone working on lldp discovery stuff for neutron intergration? I don't see a spec for it in the ironic-neutron ether pad | 13:03 |
dtantsur | sambetts, nobody right now, I guess | 13:04 |
*** Marga_ has quit IRC | 13:05 | |
sambetts | I didn't know if anyone had create a spec for it yet? | 13:05 |
jroll | sambetts: yeah not that I know of | 13:05 |
dtantsur | also nope | 13:05 |
jroll | we already have lldp stuff in IPA, I believe? | 13:05 |
*** Marga_ has joined #openstack-ironic | 13:05 | |
jroll | yeah https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/netutils.py#L109 | 13:06 |
jroll | so just need an inspection hook | 13:06 |
jroll | dtantsur: do you do specs for inspector? | 13:06 |
dtantsur | jroll, not really, usually we write a blueprint and discuss it right here :) | 13:06 |
jroll | yeah | 13:07 |
jroll | this shouldn't be much code anyway | 13:07 |
sambetts | Yeah, we have BPs, but there wasn't one of those either :-P | 13:07 |
jroll | sambetts: you agree it should all be in inspector? | 13:07 |
*** faizan has quit IRC | 13:07 | |
sambetts | I guess it depends how dynamic we're expecting it to be | 13:08 |
jroll | I tend to think people shouldn't be re-cabling things constantly :) | 13:10 |
TheJulia | lucasagomes: feel free to +2 in bifrost if you want :) | 13:11 |
sambetts | :-P true and if they do they should re-inspect I guess | 13:11 |
lucasagomes | TheJulia, hah ok I would totally +2 that patch | 13:11 |
* lucasagomes changes votes | 13:11 | |
* TheJulia goes hunting for more coffee while test things do their things | 13:14 | |
*** ukalifon has quit IRC | 13:14 | |
Bob_Ball | How do I get a RSA private key that can be used in pxe_ssh? So far I've just been using the password, but I'd like to use a key now... apparently needs to be on one line, but every way I've tried to set it gets an error from paramiko saying it's not a valid RSA key | 13:17 |
jroll | Bob_Ball: ssh_keygen? | 13:17 |
jroll | Bob_Ball: the driver_info thing is the filename | 13:18 |
Bob_Ball | I've got a keypair which works for normal ssh - I just can't figure out how to get it into ironic's config | 13:18 |
*** boris-42 has quit IRC | 13:18 | |
Bob_Ball | Trying to set "ssh_key_contents" in driver_info which is meant to be a single-line RSA private key | 13:18 |
*** smoriya has quit IRC | 13:18 | |
*** links has quit IRC | 13:19 | |
jroll | interesting, not sure about that, I've always used ssh_key_filename like devstack does https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L588 | 13:20 |
Bob_Ball | ironic just confirms that "BEGIN RSA PRIVATE" is part of the string before passing to paramiko (which it is) - paramiko then seems to check that BEGIN RSA PRIVATE KEY is in the first line, then somehow has run out of lines (which I don't understand" | 13:20 |
jroll | maybe there's a bug there | 13:20 |
jroll | the "single line" thing seems weird to me | 13:21 |
Bob_Ball | I'm guessing it usually works - RHEL OSP requires you to use ssh_key_contents, so I'd like to think that RHEL have tested it ;) | 13:21 |
lucasagomes | Bob_Ball, yeah... pointing to the key file as jroll mentioned didn't work for you? | 13:21 |
jroll | heh | 13:21 |
jroll | must be someting with the format then | 13:21 |
lucasagomes | Bob_Ball, oh, that's a bug then... it should allow the filename as well | 13:21 |
jroll | thanks RHEL for requiring keys in the database :( | 13:22 |
*** puranamr has joined #openstack-ironic | 13:22 | |
Bob_Ball | RHEL have got a batch import for nodes from a json file - the only thing they translate from that file is the ssh_key_contents | 13:22 |
lucasagomes | Bob_Ball, we need to extend this https://github.com/openstack/os-cloud-config/blob/master/os_cloud_config/nodes.py#L48 | 13:22 |
lucasagomes | to support filename | 13:22 |
Bob_Ball | That's exactly the bit I was talking about :) | 13:22 |
lucasagomes | yeah this is os-cloud-config | 13:23 |
Bob_Ball | but key contents should work :( and I don't understand why it doesn't :( | 13:23 |
* lucasagomes no idea either | 13:23 | |
Bob_Ball | No worries... I was assuming it was something stupidly simple I don't understand about private keys. May still be. But I'll add some debug commands to try and work it out | 13:24 |
lucasagomes | Bob_Ball, cool, yeah feel free to open a bug against os-cloud-config for that as well | 13:25 |
Bob_Ball | https://bugs.launchpad.net/os-cloud-config/+bug/1504550 | 13:28 |
openstack | Launchpad bug 1504550 in os-cloud-config "import should support private key file" [Undecided,New] | 13:28 |
TheJulia | cinerama: NobodyCam: please take a look at https://review.openstack.org/#/c/233039/ once coffee has hit your systems :) | 13:28 |
lucasagomes | Bob_Ball, thanks much! | 13:31 |
Bob_Ball | Raising a bug is the easy bit :D | 13:31 |
* lucasagomes sees a trick error here https://review.openstack.org/#/c/232440/3/ironic/drivers/modules/ilo/common.py | 13:31 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add agent_amt driver https://review.openstack.org/231665 | 13:33 |
TheJulia | dtantsur: Out of curiosity, does http://paste.openstack.org/show/dmjLOb8I7t1oRrRjnCtH/ ring any bells for you? | 13:34 |
*** Nisha has joined #openstack-ironic | 13:37 | |
lucasagomes | jroll, that unit test tho... is that really needed? | 13:38 |
* lucasagomes finds it a bit inconsistent with other drivers | 13:38 | |
*** lazy_prince has joined #openstack-ironic | 13:39 | |
*** davideagnello has quit IRC | 13:40 | |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - RPC https://review.openstack.org/206243 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - net https://review.openstack.org/206245 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Update the deploy drivers with network flipping logic https://review.openstack.org/213262 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - DB https://review.openstack.org/206232 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add portgroups to support LAG interfaces - objs https://review.openstack.org/206238 | 13:40 |
openstackgerrit | Om Kumar proposed openstack/ironic: Add network provider interface and implementations https://review.openstack.org/139687 | 13:40 |
*** puranamr has quit IRC | 13:41 | |
jroll | lucasagomes: ramesh asked for it, also adding them for the other agent drivers because I don't think it hurts | 13:43 |
*** mbound has quit IRC | 13:43 | |
jroll | that's failing tests right now though | 13:43 |
lucasagomes | hmm ok.. I mean, mostly it tries to catch are syntax errors or import errors imo | 13:43 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add agent_amt driver https://review.openstack.org/231665 | 13:44 |
jroll | lucasagomes: he seemed to think it would have helped with the boot interface thing | 13:44 |
lucasagomes | ok | 13:44 |
jroll | idk, I'm not sure it's needed | 13:45 |
jroll | we do have it for irmc stuff | 13:45 |
jroll | but that one is more complex | 13:45 |
Bob_Ball | lucasagomes/jroll: FYI looks like I just needed to replace the newlines with embedded \n's | 13:46 |
lucasagomes | right, yeah I'm not convinced as well, but if it helps with the boot interface thing, why not | 13:46 |
lucasagomes | a-ha... Bob_Ball that's tricky :P | 13:47 |
*** mbound has joined #openstack-ironic | 13:47 | |
jroll | Bob_Ball: ha, cool | 13:47 |
Bob_Ball | Pretty obvious when you think about it that way round... but apparently there is no such thing as a one-line private key, so you just have to trick convince json that you want multiple lines | 13:48 |
*** davideagnello has joined #openstack-ironic | 13:48 | |
*** mehip has joined #openstack-ironic | 13:48 | |
NobodyCam | good moorning Ironicers and ..... | 13:49 |
NobodyCam | TGIF | 13:49 |
NobodyCam | :) | 13:49 |
lucasagomes | NobodyCam, morning | 13:51 |
NobodyCam | mornign lucasagomes :) | 13:52 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic-specs: Proxy configuration for agent images downloading https://review.openstack.org/225181 | 13:52 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic-specs: Proxy configuration for agent images downloading https://review.openstack.org/225181 | 13:55 |
vdrok | morning jroll NobodyCam Bob_Ball and lucasagomes | 13:57 |
lucasagomes | vdrok, morning! | 13:57 |
vdrok | and TheJulia | 13:57 |
NobodyCam | morning vdrok | 13:57 |
NobodyCam | morning jroll dtantsur sambetts romcheg | 13:58 |
*** dansmith is now known as superdan | 13:59 | |
TheJulia | NobodyCam: good morning | 14:07 |
jlvillal | sambetts, Thank you | 14:13 |
jlvillal | vdrok, Thank you and congrats to you. Well deserved :) | 14:13 |
jlvillal | Good morning Ironic | 14:13 |
sambetts | \o/ jlvillal | 14:13 |
vdrok | morning jlvillal, thanks for congrats, congrats! | 14:14 |
jlvillal | :) | 14:14 |
jroll | morning vdrok, NobodyCam :) | 14:15 |
jroll | tgif \o/ | 14:15 |
*** bnemec has quit IRC | 14:16 | |
* jlvillal waits for jroll to post a youtube link ;) | 14:16 | |
jroll | now I feel like I can't post it because you said that :( | 14:17 |
romcheg | Morning NobodyCam and the other guys :) | 14:19 |
dtantsur | TheJulia, eeek, WUT?? I'm pretty sure we're using mysql in the gate now.... | 14:19 |
dtantsur | morning NobodyCam, romcheg, jlvillal and everyone | 14:19 |
*** harshs has joined #openstack-ironic | 14:20 | |
dtantsur | trown, sambetts, does http://paste.openstack.org/show/dmjLOb8I7t1oRrRjnCtH/ look familiar to you? | 14:20 |
TheJulia | dtantsur: might be the version i'm getting in CI from pip, I'll switch it out for something not from pip a little later | 14:20 |
*** bnemec has joined #openstack-ironic | 14:20 | |
jlvillal | jroll, I like when you post it! :D | 14:21 |
dtantsur | version of inspector? I don't think we landed anything new since 2.2.1 | 14:21 |
trown | dtantsur: I have not seen that one | 14:21 |
jlvillal | dtantsur, Morning :) | 14:21 |
TheJulia | dtantsur: yeah, 2.2.1 is what it is getting from pip. Odd | 14:21 |
jroll | jlvillal: too late, you ruined it | 14:21 |
TheJulia | dtantsur: I'll dig into it some more locally, it worked earlier in the week just fine :) | 14:22 |
jlvillal | :( | 14:22 |
TheJulia | dtantsur: likely something I did switching my stuff over to mysql | 14:22 |
dtantsur | yeah, I remember you confirming it... | 14:22 |
jlvillal | jroll, Okay, old school then: https://www.youtube.com/watch?v=jl1mQASHc48 | 14:23 |
jroll | jlvillal: heh :D | 14:23 |
TheJulia | Yeah, that was me just doing a quick local pip install, running the dbsync command and the world being a happy place. This, is a from scratch deploy with changed playbooks. Maybe I've got something wrong someplace. Need to dig | 14:23 |
trown | ah mysql, I have not tried that yet | 14:24 |
trown | now that we have a "real" db, that is on my list | 14:25 |
dtantsur | TheJulia, if you take the query in the bottom, and just execute it in mysql console, will it work? | 14:25 |
TheJulia | dtantsur: can't, its on a CI job, but looks like it should be fine.... which is what has me going "what?!?" | 14:26 |
*** mehip has quit IRC | 14:26 | |
TheJulia | atleast can't in the ci context | 14:26 |
*** r-daneel has joined #openstack-ironic | 14:27 | |
NobodyCam | morning jroll jlvillal TheJulia trown ] | 14:29 |
NobodyCam | :) | 14:29 |
jlvillal | NobodyCam, morning | 14:29 |
NobodyCam | :) | 14:29 |
trown | morning NobodyCam | 14:29 |
NobodyCam | TGIF!!!! (again) | 14:29 |
*** harshs has quit IRC | 14:29 | |
TheJulia | dtantsur: doing it by hand after creating a local nodes table, in my mysql install, works like a champ. | 14:30 |
dtantsur | weird.... | 14:31 |
* TheJulia thinks beer may fix it | 14:32 | |
*** mtanino has joined #openstack-ironic | 14:35 | |
dtantsur | ++ | 14:35 |
*** davideag_ has joined #openstack-ironic | 14:36 | |
*** davideagnello has quit IRC | 14:38 | |
lucasagomes | jlvillal, morning | 14:39 |
jlvillal | lucasagomes, You too. Almost the weekend for you :) | 14:40 |
lucasagomes | :-) yup | 14:40 |
*** mehip has joined #openstack-ironic | 14:49 | |
*** harshs has joined #openstack-ironic | 14:51 | |
*** mehip has quit IRC | 14:52 | |
*** thrash is now known as thrash|biab | 14:53 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic: Fix tests giving erroneous output during os-testr run https://review.openstack.org/233028 | 14:55 |
jlvillal | yuriyz, I hope my comment about no_proxy makes sense. | 14:57 |
jlvillal | yuriyz, regarding https://review.openstack.org/#/c/225181/ | 14:58 |
*** ntpttr has joined #openstack-ironic | 15:01 | |
*** Marga_ has quit IRC | 15:01 | |
*** Marga_ has joined #openstack-ironic | 15:01 | |
*** Marga_ has quit IRC | 15:04 | |
*** Marga_ has joined #openstack-ironic | 15:05 | |
krotscheck | I've... got a problem with cors middleware that I'm trying to address. | 15:08 |
krotscheck | The issue is that keystone is often layered on top of existing services. | 15:08 |
krotscheck | keystonemiddleware I mean | 15:08 |
krotscheck | So the order of operations ends up being: keystone, then cors, then api | 15:09 |
krotscheck | This presents 2 problems. | 15:09 |
krotscheck | keystone intercepts all requests, so even an OPTIONS request will be rejected. | 15:09 |
krotscheck | Second- if your auth token is invalid, the browser will never be told, because keystone rejects the request before it can be annotated. | 15:10 |
TheJulia | ugh | 15:11 |
openstackgerrit | Merged openstack/ironic: Fix a bug error by passwords only includes numbers https://review.openstack.org/220385 | 15:12 |
jlvillal | vdrok, Responded to your comment. Hopefully it makes sense. | 15:12 |
*** sdake has joined #openstack-ironic | 15:13 | |
jlvillal | vdrok, I've been living in proxy madness for the last 15+ years at my work ;) | 15:13 |
krotscheck | TheJulia: Yeah. Which is why paste.ini is apparently the way most people do their thing | 15:14 |
*** mordgan is now known as morgan | 15:14 | |
openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/232870 | 15:14 |
krotscheck | Is there some way to tell the wsgi pipeline to "no no no, execute this first"? | 15:14 |
TheJulia | i dunno, i would think so, but the browser never being told a token is invalid seems like a fundimental flaw | 15:19 |
krotscheck | TheJulia: Right, but the CORS middleware would handle that, if it was first in the wsgi chain | 15:20 |
krotscheck | But it can't be first in the chain, because it's baked into ironic, and keystone is applied on top of ironic. | 15:20 |
krotscheck | Or is it. | 15:20 |
krotscheck | WAIT | 15:20 |
krotscheck | No it's not. | 15:20 |
* krotscheck doublechecks | 15:20 | |
*** thrash|biab is now known as thrash | 15:22 | |
*** mbound has quit IRC | 15:22 | |
sambetts | dtantsur: How do I turn on IPA debug logging?/ | 15:23 |
*** masterbound has joined #openstack-ironic | 15:23 | |
dtantsur | sambetts, looks like we don't have such an option right now, which surprises me a lot | 15:23 |
dtantsur | probably someone should just implement it, don't know.. | 15:23 |
sambetts | dtantsur: if you --help it says you can --debug :/ | 15:26 |
dtantsur | sambetts, so you probably can inject a custom config (if you're using coreos, that is) | 15:26 |
dtantsur | I'd really love to have a kernel option though | 15:27 |
sambetts | dtantsur: I'm actually logged in to the IPA box and am running commands on there so, I'll work it out | 15:28 |
*** subscope has quit IRC | 15:28 | |
devananda | krotscheck: keystone is a middleware (ie, a hook inside our API code) not an external thing layered on top | 15:28 |
devananda | also - good morning, all | 15:28 |
sambetts | o/ devananda | 15:28 |
krotscheck | devananda: Yeah, I just realized that. | 15:28 |
krotscheck | devananda: I'm just doublechecking the order of invocation before I get back to jtomasek | 15:29 |
devananda | ah | 15:29 |
*** jcoufal has quit IRC | 15:30 | |
lucasagomes | devananda, morning | 15:31 |
*** marzif has quit IRC | 15:35 | |
*** marzif has joined #openstack-ironic | 15:36 | |
*** e0ne has quit IRC | 15:37 | |
*** lazy_prince has left #openstack-ironic | 15:37 | |
krotscheck | devananda, TheJulia: Yep, looks like someone was trying to deploy iroinc with keystonemiddleware | 15:37 |
*** lazy_prince has joined #openstack-ironic | 15:37 | |
lazy_prince | jroll: When you have time, can you look at https://review.openstack.org/#/c/139687/29/ironic/networks/neutron_plugin.py | 15:38 |
*** e0ne has joined #openstack-ironic | 15:39 | |
dtantsur | morning devananda | 15:39 |
*** harshs has quit IRC | 15:39 | |
vdrok | jlvillal, agree, makes sense :) | 15:39 |
vdrok | morning devananda | 15:39 |
*** mjturek1 has quit IRC | 15:41 | |
dtantsur | folks, have you seen the thread "jobs that make break when we remove Devstack extras.d in 10 weeks"? | 15:41 |
dtantsur | our jobs are listed there, and I was unable to understand why | 15:42 |
openstackgerrit | Lilia Sampaio proposed openstack/ironic: OneView Driver for Ironic https://review.openstack.org/191822 | 15:42 |
*** e0ne has quit IRC | 15:43 | |
*** persia has quit IRC | 15:46 | |
*** persia has joined #openstack-ironic | 15:47 | |
*** mjturek1 has joined #openstack-ironic | 15:48 | |
jlvillal | vdrok, Thanks | 15:51 |
johnthetubaguy | jroll: would be good to quickly catch up with some summit sessions planning quickly, if you have a sec, let me know when works | 15:51 |
*** mjturek1 has quit IRC | 15:52 | |
*** dims_ has joined #openstack-ironic | 15:56 | |
*** dims_ has quit IRC | 15:56 | |
*** dimsum__ has quit IRC | 15:59 | |
jlvillal | dtantsur, Maybe you know who to ask to get the ironic-lib version patch approved. | 15:59 |
dtantsur | jlvillal, one of https://review.openstack.org/#/admin/groups/131,members | 15:59 |
*** ntpttr has quit IRC | 15:59 | |
dtantsur | (well, actually two of) | 16:00 |
jlvillal | :) | 16:00 |
jlvillal | I did do some begging on infra, but no luck so far | 16:00 |
dtantsur | calling it a day now, beer times approaches quickly :) | 16:03 |
dtantsur | have a nice weekend | 16:03 |
sambetts | dtantsur: Have a great weekend!! o/ | 16:03 |
*** Marga_ has quit IRC | 16:03 | |
*** dtantsur is now known as dtantsur|afk | 16:04 | |
lucasagomes | dtantsur|afk, g'night | 16:04 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic-specs: Proxy configuration for agent images downloading https://review.openstack.org/225181 | 16:04 |
*** mjturek1 has joined #openstack-ironic | 16:05 | |
*** sdake has quit IRC | 16:06 | |
*** sdake_ has joined #openstack-ironic | 16:06 | |
*** mjturek1 has quit IRC | 16:07 | |
*** mjturek1 has joined #openstack-ironic | 16:08 | |
*** jistr has quit IRC | 16:08 | |
*** dimsum__ has joined #openstack-ironic | 16:09 | |
*** dimsum__ has quit IRC | 16:09 | |
*** dimsum__ has joined #openstack-ironic | 16:10 | |
*** tsekiyama has joined #openstack-ironic | 16:10 | |
*** trown is now known as trown|lunch | 16:18 | |
*** lsmola_ has quit IRC | 16:18 | |
*** mkovacik has quit IRC | 16:23 | |
*** Sukhdev has joined #openstack-ironic | 16:25 | |
Sukhdev | lazy_prince: ping | 16:26 |
*** sdake_ is now known as sdake | 16:28 | |
*** Marga_ has joined #openstack-ironic | 16:29 | |
lazy_prince | Sukhdev: hi.. | 16:29 |
Sukhdev | lazy_prince: got few minutes to chat now? | 16:31 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Fix agent_ilo to remove temporary images https://review.openstack.org/232323 | 16:31 |
lazy_prince | Sukhdev: sure... Did you take a look at the patches that I pushed today..? | 16:32 |
lazy_prince | Sukhdev: Please review them when you get time. | 16:32 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Fix agent_ilo to remove temporary images https://review.openstack.org/232323 | 16:32 |
Sukhdev | lazy_prince: no. Not yet. But will look at them later | 16:34 |
*** Sukhdev_ has joined #openstack-ironic | 16:36 | |
lazy_prince | Sukhdev: okay.. so apart from issues pointed by you, I found another issue where dhcp options were being added to the tenant port instead of provisioning port. | 16:36 |
lazy_prince | Sukhdev: so wanted to check with you, if you saw the same in your environment too.. | 16:36 |
*** achanda has joined #openstack-ironic | 16:38 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Fix agent_ilo to remove temporary images https://review.openstack.org/232323 | 16:39 |
*** achanda has quit IRC | 16:39 | |
*** achanda has joined #openstack-ironic | 16:40 | |
*** masterbound has quit IRC | 16:40 | |
jroll | johnthetubaguy: free now if you are | 16:41 |
jroll | lazy_prince: I'll try to take a look soon | 16:42 |
lucasagomes | folks I'm calling it a day | 16:42 |
lucasagomes | have a great night everyone! | 16:42 |
lucasagomes | enjoy the weekend | 16:42 |
*** achanda has quit IRC | 16:42 | |
JayF | have a good weekend lucasagomes | 16:42 |
lucasagomes | you too! | 16:43 |
*** lucasagomes is now known as lucas-dinner | 16:43 | |
*** ifarkas has quit IRC | 16:47 | |
*** garthb has joined #openstack-ironic | 16:48 | |
*** trown|lunch is now known as trown | 16:49 | |
*** achanda has joined #openstack-ironic | 16:50 | |
*** mjturek1 has quit IRC | 16:52 | |
*** thiagop has joined #openstack-ironic | 16:53 | |
thiagop | Hey Ironic, good UGT morning | 16:53 |
thiagop | and TGIF! | 16:53 |
BadCub | howdy thiagop Happy Friday to ya too | 16:54 |
johnthetubaguy | jroll: hey, sorry about that, was talking quotas with some folks | 16:59 |
jroll | johnthetubaguy: no worries | 16:59 |
*** achanda has quit IRC | 16:59 | |
johnthetubaguy | jroll: I went through the Nova summit planning, are you happy with the ironic session in that thursday slow we said | 16:59 |
* johnthetubaguy goes to find etherpad tab | 17:00 | |
jroll | johnthetubaguy: which slot, and do you have the google docs link handy? | 17:00 |
johnthetubaguy | Thursday 11:50 I think | 17:00 |
jroll | actually, I think I have the docs thing | 17:00 |
johnthetubaguy | https://etherpad.openstack.org/p/mitaka-nova-summit-suggestions | 17:00 |
johnthetubaguy | that has all my notes, and a draft schedule down the bottom | 17:00 |
johnthetubaguy | should have the links in there to that spreadsheet actually | 17:00 |
johnthetubaguy | http://tinyurl.com/mitaka-design-summit | 17:01 |
* jroll looking | 17:01 | |
jroll | so that gives us an ironic fishbowl slot | 17:02 |
jroll | and you have a talk at that time, that's the only lame part | 17:02 |
jroll | but we can probably get things done without you, like you said before | 17:02 |
*** athomas has quit IRC | 17:02 | |
jroll | johnthetubaguy: that works for me, let's roll with it | 17:03 |
*** ijw has joined #openstack-ironic | 17:03 | |
johnthetubaguy | jroll: cool, that works, I would prefer to be there, but that should be OK | 17:03 |
jroll | johnthetubaguy: I'll expand on topics next week if that's good; so far I have in mind the multi-compute thing and some networking related things | 17:03 |
*** amotoki has quit IRC | 17:04 | |
jroll | going to take a look at other outstanding specs/etc that involve nova work, like boot from volume | 17:04 |
johnthetubaguy | jroll: so in cheddar, I think you can add the Nova track into your Ironic session, which should get folks turning up :) | 17:04 |
jroll | yeah, I'll do that now, and mark it on our etherpad | 17:04 |
johnthetubaguy | jroll: sweet, happy to go through those some time with you, if that helps | 17:04 |
jroll | we're going to finish up our schedule next week | 17:04 |
jroll | ++ | 17:04 |
jroll | I'll get a list together by monday or so, also need to write/update some of that stuff | 17:05 |
johnthetubaguy | sounds good, I mean its not crazy urgent, but yeah, lets try sync up next week sometime | 17:05 |
jroll | kinda stuck on the multi-compute thing | 17:05 |
jroll | yeah, not urgent but I want to get it out of the way | 17:05 |
johnthetubaguy | yeah, actually, I did have some crazy ideas to add in there | 17:06 |
jroll | orly, would love to hear | 17:06 |
*** ijw_ has joined #openstack-ironic | 17:06 | |
johnthetubaguy | I was thinking about treating all resources shared between the computes, like shared storage of multiple regular nova-compute nodes (like a nfs or ceph setup) | 17:06 |
johnthetubaguy | so the resource claim might then just live in the scheduler | 17:07 |
jroll | ah, right | 17:07 |
johnthetubaguy | which, I hope leaves very little host specific triggers at the bottom | 17:07 |
jroll | is that a thing that exists today, or is that the jay pipes spec you pointed me at? | 17:07 |
*** ijw has quit IRC | 17:07 | |
johnthetubaguy | thats the rub, its make believe | 17:07 |
jroll | heh | 17:07 |
*** mjturek1 has joined #openstack-ironic | 17:07 | |
jroll | yeah, that's what I figured | 17:08 |
johnthetubaguy | so my thinking is, if we could agree the shiny rainbow land place, we can figure how to get there more quickly | 17:08 |
jroll | I like the idea, though from an ops perspective this is the #1 biggest problem with ironic imo | 17:08 |
johnthetubaguy | yeah, agreed | 17:08 |
johnthetubaguy | well, ironic + nova I guess, but yeah | 17:08 |
jroll | so I'd like to not depend on something that might take a while | 17:08 |
jroll | right | 17:08 |
johnthetubaguy | so my though is really, find something that looks right, then work out what a good short term thing might be | 17:09 |
johnthetubaguy | thought^ | 17:09 |
*** harshs has joined #openstack-ironic | 17:09 | |
jroll | another random question I had about it: one of the longer goals is to eliminate node from being something we schedule against | 17:09 |
johnthetubaguy | unsure, honestly | 17:09 |
jroll | heh, did that change? | 17:10 |
johnthetubaguy | I think I would say, make it less different than libvirt nova-compute | 17:10 |
johnthetubaguy | s | 17:10 |
jroll | yeah | 17:10 |
johnthetubaguy | which is kinda the same thing, I guess | 17:10 |
jroll | so if you remember the nova midcycle, we talked about the scheduler just calling the ironic api to get a thing | 17:10 |
johnthetubaguy | yeah | 17:10 |
jroll | and ironic api doing the heavy lifting | 17:10 |
jroll | there's folks in ironic that don't like that, because we basically have to reinvent nova-sch | 17:10 |
jroll | which is totally fair | 17:11 |
johnthetubaguy | so we could split out that project and you could run a different instance of it, but thats not close yet either | 17:11 |
*** Sukhdev_ has quit IRC | 17:11 | |
johnthetubaguy | oh wait... | 17:11 |
johnthetubaguy | we could run one under the ironic driver, in nova | 17:11 |
jroll | so we're thinking a BaremetalScheduler subclass that could keep ironic info in memory and somehow munge things around so that the filters look at that data | 17:11 |
jroll | but the scheduler eventually just returns random.choice([all_hosts]) | 17:12 |
jroll | and we just tack the node id onto an instance | 17:12 |
jroll | is that completely insane? :) | 17:12 |
johnthetubaguy | a little bit I guess | 17:13 |
jroll | heh | 17:13 |
jroll | really I just need to play with the code instead of trying to think about it, I think | 17:13 |
johnthetubaguy | just trying to see it in my head I guess | 17:13 |
jroll | yeah, think of it as a pipeline, right, where the filter/weigher/selection is in the middle | 17:13 |
jroll | and we tack on a piece on each end to munge data | 17:13 |
johnthetubaguy | so the way the scheduler works right now is fun | 17:13 |
johnthetubaguy | we get all the data, and cache it | 17:14 |
johnthetubaguy | then we run filters and weights over that for each request | 17:14 |
jroll | (I foresee lots of conflicts with the resource tracker here, btw) | 17:14 |
johnthetubaguy | eventually update the cache so its less out of date | 17:14 |
johnthetubaguy | jroll: so you could argue you don't need a resource tracker, because its done by ironic | 17:14 |
jroll | yeah, I'm thinking of things like tracking orphans | 17:15 |
jroll | (something we just hit downstream that we need to fix) | 17:15 |
johnthetubaguy | so we do have orphan vm tracking | 17:15 |
johnthetubaguy | you could copy that | 17:15 |
jroll | right, ironic doesn't | 17:15 |
JayF | That orphan tracking bug: https://bugs.launchpad.net/nova/+bug/1503453 | 17:16 |
openstack | Launchpad bug 1503453 in OpenStack Compute (nova) "unavailable ironic nodes being scheduled to" [Medium,Triaged] | 17:16 |
jroll | I want to fix that, requires updating driver api to add node to the method that reports them | 17:16 |
jroll | but I'm trying to reason about how all of that code would work without a resource tracker, and if not working means it would blow up spectacularly or just kinda ignore it | 17:16 |
johnthetubaguy | jroll: true, its a bit too heavily used right now, thinking about it | 17:17 |
jroll | yeah, if nova doesn't trust the driver to report resources correctly, how will it trust ironic :) | 17:17 |
johnthetubaguy | jroll: did you see this bit: https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L5954 | 17:18 |
johnthetubaguy | jroll: true, we can't trust it, mostly because of ourselves screwing up | 17:19 |
jroll | heh | 17:19 |
johnthetubaguy | jroll: so I guess the action is to get something we both kinda like before the summit, to present in that session | 17:20 |
*** achanda has joined #openstack-ironic | 17:22 | |
jroll | johnthetubaguy: yeah, going to see what I can do. I still like the long term vision of ironic handling all of the resource tracking and claims | 17:22 |
jroll | johnthetubaguy: and from there, we could do really deep hardware inspection and schedule against all types of things, without storing all of that in nova as well | 17:22 |
johnthetubaguy | true | 17:22 |
johnthetubaguy | we can share code, and things, but thats not the point really | 17:22 |
jroll | a flavor could be version 123 bios, mellanox nic, 7200 RPM HDD | 17:23 |
jroll | or whatever. | 17:23 |
jroll | stuff nova gives zero cares about | 17:23 |
johnthetubaguy | so there is another idea I had, which involves building a layer above both Nova Ironic and Magnum, with its own API... | 17:23 |
JayF | jroll: ++ I agree with that b/c one of the "scariest" things about running a production Ironic cloud is the dubious state syncing that is implicit between Nova and Ironic | 17:23 |
johnthetubaguy | so I need to run, but we should totally talk more on this | 17:25 |
johnthetubaguy | my stomach is starting to control me now, so its time | 17:25 |
jroll | johnthetubaguy: yeah, that's super interesting, but that's so far off | 17:25 |
* jroll tosses a beer and a snack to johnthetubaguy | 17:25 | |
jroll | have a good weekend, this was helpful :) | 17:26 |
* johnthetubaguy catches | 17:26 | |
johnthetubaguy | and you, when it happens! | 17:26 |
johnthetubaguy | catch you later | 17:26 |
jroll | see ya | 17:26 |
jroll | devananda: ^ good conversation you care about above | 17:26 |
jlvillal | jroll: Priorities between functional testing and AMT? I'm thinking functional testing is what I should focus on, but thought I would double check. | 17:27 |
jroll | jlvillal: 1) you're leading the testing effort, so that should be your main goal. 2) I suspect nobody will ever run AMT in production, while everyone will run ironic in production, if that answers the question :) | 17:28 |
jlvillal | It does :) Thank you. | 17:28 |
* jlvillal goes back to reading other projects functional testing code... | 17:28 | |
devananda | johnthetubaguy: I've been thinking a bunch about a layer on top of nova/ironic/magnum as well | 17:28 |
devananda | johnthetubaguy: and how we need to be able to schedule different types of things -- not just VMs with volumes attached -- and that's really not being solved by the existing scheduler work | 17:29 |
johnthetubaguy | devananda: I should write that up really, I think its basically what we (as in Nova) always called heat, but they actually built something more interesting | 17:29 |
*** thiagop has quit IRC | 17:29 | |
jroll | devananda: think about deploying that to an existing cloud and how much of a very-long-term-thing that becomes | 17:29 |
jroll | we still need something in the meantime | 17:30 |
johnthetubaguy | jroll: +1 on both fronts | 17:30 |
jroll | (I know that probably isn't what you're saying) :) | 17:30 |
devananda | johnthetubaguy: please do write up your thoughts. | 17:30 |
johnthetubaguy | devananda: I will move that up the TODO list a little, its an interesting answer to a few problems we keep hitting | 17:31 |
johnthetubaguy | the problem is its another API | 17:31 |
johnthetubaguy | anyways, I should totally do that | 17:32 |
*** lazy_prince has quit IRC | 17:32 | |
devananda | yea, well, there's a couple APIs that tease out of it, which is not great for simplicity's sake, but it may actually make things easier to consume | 17:33 |
*** nicodemos has joined #openstack-ironic | 17:33 | |
devananda | end state should be a single API for describing and interacting with compute resources, regardless of VM / baremetal / container -- I think we all agree that that needs to exist | 17:35 |
devananda | but, like jroll said, the types of things one cares about when scheduling baremetal (or containers, for that matter) are somewhat different than what one cares about with VMs | 17:35 |
devananda | I don't know if the best way to solve that is that each service grows a unique scheduler, but given the non-starter that gannt was, I wonder if we just need to do go down that ^ road so we can make some progress on this | 17:36 |
devananda | anyway, good topics for the summit | 17:36 |
*** marzif has quit IRC | 17:37 | |
jroll | devananda: I'm going to add "ironic filter API / scheduler" to summit session list | 17:37 |
jroll | we definitely need the former, it tends to lead to the latter if we do need it | 17:37 |
devananda | yup | 17:37 |
jroll | scheduler would be a filter API -> random.choice(results) and lock | 17:37 |
jroll | lock/claim/reserve | 17:38 |
*** achanda has quit IRC | 17:38 | |
devananda | yea - we've talked about adding that facility to our API before. I'm all for it, but it starts to get tangled in with how nova does resource claims / tracking | 17:38 |
jroll | right, so we would use this instead of nova's bits for that | 17:39 |
devananda | ++ | 17:39 |
jroll | O_o | 17:39 |
jroll | you were so against this in seattle :D | 17:39 |
devananda | I'm allowed to change my mind on things :) | 17:40 |
*** thiagop has joined #openstack-ironic | 17:40 | |
jroll | devananda: reviews welcome here then https://review.openstack.org/#/c/194453/7/specs/liberty/approved/ironic-multiple-compute-hosts.rst | 17:41 |
jroll | :) | 17:41 |
NobodyCam | good morning devananda :-p | 17:42 |
*** achanda has joined #openstack-ironic | 17:43 | |
mordred | devananda: hi | 17:43 |
devananda | jroll: ack. also - I need to run now. need food, and need to run to a dr's appt | 17:44 |
jroll | devananda: yeah, no worries. want me to update the fakeboot thing today? | 17:44 |
* SpamapS makes a wookie noise | 17:44 | |
devananda | jroll: pls do | 17:44 |
jroll | k | 17:44 |
devananda | sorry I couldn't get to it | 17:44 |
jroll | no worries | 17:44 |
jroll | hopefully I can figure it all out :P | 17:44 |
jroll | devananda: so instantiate FakeBoot for intree drivers, do your thing otherwise? | 17:44 |
devananda | jroll: on the plus side, I got busy diving into softlayer's architecture | 17:45 |
jroll | devananda: when is diving into existing architecture a plus? :P | 17:45 |
NobodyCam | mornign mordred & SpamapS | 17:45 |
mordred | devananda, jroll, johnthetubaguy I would like to be involved in such a follow up conversation | 17:45 |
mordred | morning NobodyCam | 17:45 |
jroll | mordred: re api in front of nova/magnum/ironic? | 17:45 |
SpamapS | NobodyCam: hah, a Star Wars wookie, not a NobodyCam wookie. ;) | 17:45 |
NobodyCam | lol | 17:46 |
SpamapS | NobodyCam: but hi! | 17:46 |
mordred | jroll: yes | 17:46 |
BadCub | lol | 17:46 |
jroll | mordred: I fully suspect it will start either in the mailing list or in a bar in tokyo, if the latter happens we'll try to find you | 17:46 |
* mordred may have some recent experience in the area of making things that are not the same look the same | 17:46 | |
SpamapS | Regarding "lets make a scheduler" .. perhaps.. lets not? --> http://mesos.apache.org/ | 17:46 |
jroll | suspect/expect | 17:46 |
devananda | mordred: there were two threads there, I think. one was that ^. the other was doing actual scheduling and resource tracking in ironic, and allowing nova to just "hand off" that responsibility rather than duplicate and try to stay in sync | 17:46 |
mordred | jroll: let's start it at a bar | 17:46 |
jroll | ++ | 17:46 |
mordred | jroll: I think starting exploratory conversations on the mailing list is a waste of time | 17:46 |
devananda | SpamapS: ++ to not reinventing that and just using mesos .... | 17:46 |
jroll | yeah | 17:47 |
devananda | mordred: if this happens in a bar ,I will text you | 17:47 |
* jroll hms | 17:47 | |
jroll | hrms, too. | 17:47 |
mordred | SpamapS: it has been communicated to me that mesos does not work nearly so well in real life as the hype would have you believe | 17:47 |
*** achanda has quit IRC | 17:47 | |
mordred | "falls over a lot" was mentioned | 17:47 |
SpamapS | Seems like if you can teach Mesos how to find magnum, ironic, and nova things.. you get a high level interface to "computers of various types" | 17:47 |
SpamapS | mordred: perhaps we could fix it? | 17:47 |
devananda | mordred: <sadface> | 17:47 |
SpamapS | It's v0.24 | 17:48 |
SpamapS | Pretty sure everything falls over a lot at this age. ;) | 17:48 |
mordred | possibly - but they're chasing a goal I probably don't agree with when it comes down to it | 17:48 |
jroll | SpamapS: ok yeah, that's two different things, in the meantime we have to teach nova how to schedule to ironic, and as such will need to build some scheduler things :/ | 17:48 |
mordred | which is the theory that everything can be disassembled into a generic "workload" and then "magic" will manage it | 17:48 |
mordred | that works really well for wordpress blogs | 17:49 |
mordred | BUT - it's a great bar conversatoin | 17:49 |
mordred | I'm probably wrong | 17:49 |
SpamapS | The problem is | 17:49 |
SpamapS | we're probably all simultaneously right and wrong | 17:49 |
mordred | yup | 17:49 |
*** achanda has joined #openstack-ironic | 17:49 | |
SpamapS | Drinks make us more right though. | 17:50 |
mordred | to be clear ... | 17:50 |
mordred | I was volunteering to do heavy lifting on the API layer in front | 17:50 |
mordred | since I've been basically doing a form of that for the past year (although it doesn't consolidate ironic and nova) | 17:50 |
SpamapS | jroll: yeah I agree thats something else. "physical constraints" has always been something just out of reach for Ironic. | 17:51 |
mordred | so I've got a BUNCH of thoughts and anger there | 17:51 |
jroll | it's be really odd to switch the front line API out from under users | 17:51 |
mordred | for the resource scheduling, I have many different thoughts | 17:51 |
mordred | jroll: I would NEVER switch the frontline api out | 17:51 |
mordred | jroll: I would add a new optional api on beside | 17:51 |
mordred | if someone wants to use a new set of abstractions, neat | 17:51 |
mordred | if they want to continue to use nova and ironic and magnum more directly, awesome | 17:51 |
jroll | mordred: yeah, but the real goal is "doing all this stuff in nova gets really weird", right? | 17:52 |
mordred | nothing about a new api should mean the old ones need to go away | 17:52 |
jroll | our users use the nova api to get a baremetal | 17:52 |
thiagop | vdrok: ping | 17:52 |
jroll | s/real goal/real problem/ | 17:52 |
mordred | jroll: yes | 17:52 |
vdrok | thiagop, pong | 17:52 |
mordred | jroll: but imagine if the new API could start by doing what your users are doing | 17:52 |
mordred | jroll: get a VM from nova, get a baremetal from nova - but do so with semantics that make sense | 17:52 |
*** sdake has quit IRC | 17:53 | |
jroll | mordred: yeah, it's probably doable, just trying to work it all out in my head | 17:53 |
thiagop | vdrok: wrt your comment about not having vendor on the fake interface of the oneview driver, we need to test that interface | 17:53 |
mordred | jroll: that's what the alcohol at the bar is for | 17:53 |
thiagop | vdrok: you think that overwrite on the test setUp is sufficient? | 17:54 |
jroll | mordred: because "what users do today" is use the nova API, and presumably the goal of this new thing would be to not be the nova api | 17:54 |
dprince | jroll, devananda: hi, a minor DIB change here... I wanted to ask if this effects Ironic CI https://review.openstack.org/#/c/233088/ | 17:54 |
mordred | jroll: yes | 17:54 |
SpamapS | What would it be? | 17:54 |
vdrok | thiagop, oops, sorry, you overloaded it | 17:54 |
vdrok | please ignore that comment :) | 17:54 |
*** sdake has joined #openstack-ironic | 17:54 | |
mordred | jroll: do you want the users to say "hey, give me a vm" or "hey, give me a baremetal"? | 17:55 |
*** ntpttr has joined #openstack-ironic | 17:55 | |
sambetts | jroll, mordred isn't that what flavors are for? | 17:55 |
* thiagop lost something... | 17:55 | |
devananda | mordred: so to your point -- I think Ironic should also start presenting a more user-friendly API than it does today | 17:56 |
thiagop | vdrok: take a look on test_vendor, line 48 | 17:56 |
thiagop | vdrok: please | 17:56 |
mordred | sambetts: yes. but I believe there is an impedence mismatch in that nova's api was designed with VMs in mind, and we've backported non-VM into those constructs | 17:56 |
mordred | sambetts: which is great - because both a VM and a baremetal are a "computer" | 17:56 |
jroll | mordred: I got lost, I think | 17:57 |
jroll | my brain isn't necessarily working today | 17:57 |
mordred | sambetts: but there are also differences between them, and the nova API does not really have way to expose that to the user that is leaving people feel satisfied | 17:57 |
thiagop | vdrok: that's the same way BaseAgentVendor's test is implemented. I don't know if I should keep it there or point the vendor on FakeOneViewDriver | 17:57 |
mordred | jroll: that's fine :) | 17:57 |
devananda | I find this conversation *really* important, but I also do not want to be late for an appointment :-/ | 17:57 |
mordred | devananda: we'll come back to it | 17:57 |
jroll | yeah | 17:58 |
sambetts | ah... I see... :/ I also have to run, but am v.interested in this convo | 17:58 |
vdrok | thiagop, well this vendor interface in fake driver (if used by someone) won't work with fake boot and fake deploy | 17:58 |
sambetts | have a good weekend everyone o/ | 17:58 |
jlvillal | sambetts: Enjoy! | 17:58 |
NobodyCam | night sambetts | 17:58 |
jlvillal | sambetts: eavesdrop.openstack.org :) | 17:58 |
sambetts | jlvillal: :-P good point | 17:59 |
devananda | dprince: at a glance, i do not understand the change in question :( | 17:59 |
devananda | perhaps jroll does? | 17:59 |
dprince | devananda: maybe you don't care about this then :). Just wanted to mention it | 18:00 |
*** pelix has quit IRC | 18:00 | |
thiagop | vdrok: I'm putting it just for the sake of testing, don't know if I should put that on the fake driver or just let that self.passthru thing on the test | 18:00 |
*** thrash is now known as thrash|brb | 18:00 | |
dprince | devananda: basically just changing it so you'd set DIB_REPOREF_ironic_agent instead of DIB_REPOREF_agent, just makes things more consistent in DIB with regards to the element name | 18:00 |
jroll | dprince: not sure, I'll need to come back to this | 18:01 |
*** ionutbalutoiu has quit IRC | 18:02 | |
dprince | jroll: ack, no worries. I don't think it effects the CI in ironic | 18:02 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for partition images in agent driver https://review.openstack.org/162008 | 18:03 |
vdrok | thiagop, but with all this mocking you could use agent oneview in tests too? | 18:04 |
vdrok | not fake one | 18:04 |
thiagop | vdrok: let'me see what I've got running with agent... | 18:06 |
jroll | dprince: we'll find out eh | 18:06 |
*** sdake has quit IRC | 18:07 | |
*** cdearborn has joined #openstack-ironic | 18:10 | |
thiagop | vdrok: good catch, it works. Will remove that vendor from fake and run the tests with agent_pxe_oneview | 18:10 |
thiagop | TheJulia: ping | 18:10 |
vdrok | thiagop, so if my understanding is correct, you should add vendor interface to fake driver if there is some specific vendor functional that can be used without other interfaces (like e.g. ipmitool vendor passthru) | 18:11 |
vdrok | if it is only for deployment, and you use fake deploy it can be ommitted | 18:11 |
vdrok | good night everyone, have a good weekend! | 18:12 |
thiagop | vdrok: ok. thaks | 18:12 |
thiagop | vdrok: have a great weekend | 18:12 |
*** ntpttr has quit IRC | 18:12 | |
openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: OneView Driver for Ironic https://review.openstack.org/191822 | 18:15 |
TheJulia | thiagop: pong | 18:16 |
thiagop | TheJulia: Hi, I sent another patch addressing the comments of vdrok now but I'll wait your answer about the docs thing, right? | 18:18 |
thiagop | TheJulia: just for you to know that I'm not ignoring it :D | 18:18 |
TheJulia | hmm.. interresting discussions in bars in tokyo, this could be good. | 18:18 |
BadCub | hmmmm | 18:19 |
TheJulia | thiagop: Ok, at lunch, Opening review now | 18:20 |
thiagop | TheJulia: bon apetit | 18:20 |
*** achanda has quit IRC | 18:22 | |
*** Marga_ has quit IRC | 18:28 | |
*** Marga_ has joined #openstack-ironic | 18:28 | |
TheJulia | thiagop: responded on the review | 18:28 |
TheJulia | thiagop: hopefully that makes sense :) | 18:29 |
*** sdake has joined #openstack-ironic | 18:31 | |
*** thrash|brb is now known as thrash | 18:32 | |
*** sambetts has quit IRC | 18:34 | |
*** Marga_ has quit IRC | 18:35 | |
*** sambetts has joined #openstack-ironic | 18:38 | |
*** jistr has joined #openstack-ironic | 18:40 | |
*** ijw_ is now known as ijw | 18:41 | |
*** achanda has joined #openstack-ironic | 18:44 | |
liliars | TheJulia: hi, thiagop left for the day just now so I answered your comment on the review, thanks! Will be sending a patch with an initial documentation asap (: | 18:48 |
*** Sukhdev has quit IRC | 18:49 | |
TheJulia | liliars: Thank you! :) | 18:50 |
*** jistr is now known as jistr|afk | 18:50 | |
*** achanda has quit IRC | 18:53 | |
*** pas-ha has quit IRC | 18:54 | |
BadCub | liliars: awesome! Thank you :) | 18:59 |
*** albertoffb has quit IRC | 18:59 | |
*** ukalifon1 has joined #openstack-ironic | 19:00 | |
*** jistr|afk has quit IRC | 19:05 | |
*** zer0c00l has quit IRC | 19:10 | |
*** harshs has quit IRC | 19:11 | |
*** balajin_ has quit IRC | 19:12 | |
*** harlowja has quit IRC | 19:12 | |
*** [1]cdearborn has joined #openstack-ironic | 19:13 | |
*** ukalifon1 has quit IRC | 19:18 | |
*** balajin has joined #openstack-ironic | 19:19 | |
*** zer0c00l has joined #openstack-ironic | 19:19 | |
*** mrmoje has quit IRC | 19:19 | |
*** cdearborn has quit IRC | 19:21 | |
*** mrmoje has joined #openstack-ironic | 19:21 | |
*** harlowja has joined #openstack-ironic | 19:24 | |
*** achanda has joined #openstack-ironic | 19:31 | |
* jlvillal decides he is hungry | 19:36 | |
*** tsekiyam_ has joined #openstack-ironic | 19:44 | |
*** mtanino has quit IRC | 19:45 | |
*** tsekiyama has quit IRC | 19:47 | |
*** e0ne has joined #openstack-ironic | 19:49 | |
*** nicodemos has quit IRC | 19:50 | |
*** Nisha has quit IRC | 19:51 | |
*** achanda has quit IRC | 19:53 | |
*** Marga_ has joined #openstack-ironic | 19:54 | |
*** mtanino has joined #openstack-ironic | 20:03 | |
*** aswadr has quit IRC | 20:12 | |
*** mjturek1 has quit IRC | 20:15 | |
*** ntpttr has joined #openstack-ironic | 20:17 | |
*** baoli_ has quit IRC | 20:22 | |
*** baoli has joined #openstack-ironic | 20:22 | |
*** baoli has quit IRC | 20:22 | |
*** e0ne has quit IRC | 20:26 | |
*** achanda has joined #openstack-ironic | 20:34 | |
*** krtaylor has quit IRC | 20:48 | |
*** willsama has joined #openstack-ironic | 20:51 | |
*** Sukhdev has joined #openstack-ironic | 20:58 | |
*** trown is now known as trown|outttypeww | 20:58 | |
*** krtaylor has joined #openstack-ironic | 21:02 | |
*** marzif has joined #openstack-ironic | 21:02 | |
*** sdake_ has joined #openstack-ironic | 21:05 | |
*** sambetts has quit IRC | 21:06 | |
*** sdake has quit IRC | 21:06 | |
*** sambetts has joined #openstack-ironic | 21:11 | |
*** sdake_ has quit IRC | 21:14 | |
*** dprince has quit IRC | 21:16 | |
*** Sukhdev has quit IRC | 21:19 | |
*** sdake has joined #openstack-ironic | 21:19 | |
*** Sukhdev has joined #openstack-ironic | 21:20 | |
*** tsekiyama has joined #openstack-ironic | 21:26 | |
jlvillal | lekha: krtaylor: So I have investigating the functional testing. So the first thing we need is a way to bring up and tear down the Ironic services. ironic-conductor & ironic-api. | 21:28 |
jlvillal | So I'm starting to see how Nova is doing it with fixtures and see if that is what we want. | 21:29 |
jlvillal | But thought I'd see if either of you had feedback/ideas | 21:29 |
*** tsekiyam_ has quit IRC | 21:29 | |
*** mbound has joined #openstack-ironic | 21:45 | |
*** tsekiyama has quit IRC | 21:51 | |
*** ntpttr has quit IRC | 21:54 | |
*** willsama has quit IRC | 22:01 | |
*** davideag_ has quit IRC | 22:08 | |
*** davideagnello has joined #openstack-ironic | 22:09 | |
*** [1]cdearborn has quit IRC | 22:11 | |
*** mtanino has quit IRC | 22:13 | |
*** Sukhdev has quit IRC | 22:14 | |
*** ntpttr has joined #openstack-ironic | 22:31 | |
*** Sukhdev has joined #openstack-ironic | 22:38 | |
*** dims_ has joined #openstack-ironic | 22:50 | |
*** marzif has quit IRC | 22:51 | |
*** dimsum__ has quit IRC | 22:53 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix AgentDeploy compatibility with pre-Liberty drivers https://review.openstack.org/231215 | 22:53 |
JayF | jroll: does ^ need backport-potential-liberty ? | 22:59 |
JayF | jroll: or whatever the tag is nowadays | 22:59 |
jroll | JayF: the bug has it yeah | 22:59 |
JayF | I thought there was a commit tag for it? | 22:59 |
jroll | except the commit message has the wrong bug | 22:59 |
jroll | no | 22:59 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix AgentDeploy compatibility with pre-Liberty drivers https://review.openstack.org/231215 | 23:00 |
jroll | fixed commit msg | 23:00 |
JayF | If you don't have the appropriate permissions to nominate the bug, then tagging it with icehouse-backport-potential is also sufficient e.g. for nova. | 23:00 |
JayF | ^ from the docs, so apparently it's just a workaroud for those of us less powerful | 23:00 |
jroll | ah | 23:01 |
jroll | yeah never heard of that | 23:01 |
jroll | devananda: ^ that patch is up btw, it's passing unit tests which feels wrong :) | 23:05 |
*** r-daneel has quit IRC | 23:07 | |
*** dims_ has quit IRC | 23:07 | |
jlvillal | jroll: So task.driver.boot exists pre-liberty but is set to None? | 23:12 |
jroll | jlvillal: no, but if a driver doesn't set self.boot at instantiation, it will be None | 23:13 |
jroll | er wait | 23:13 |
jroll | no, it didn't exist in Kilo | 23:13 |
jlvillal | That is what I was wondering. Does it magically get created? | 23:14 |
jlvillal | I was thinking maybe I was being dumb. | 23:14 |
jroll | jlvillal: no, that's why things exploded with a kilo-level out-of-tree driver | 23:14 |
jlvillal | jroll: Then wouldn't it need a hasattr(driver, 'boot') ? | 23:15 |
jroll | jlvillal: we removed the direct references to driver.boot | 23:16 |
jroll | oh | 23:16 |
jroll | sec, will link | 23:16 |
jlvillal | jroll: https://review.openstack.org/#/c/231215/3/ironic/drivers/modules/agent.py L181 | 23:16 |
jroll | basically the base driver sets it to None | 23:16 |
jlvillal | Okay, good. | 23:16 |
jroll | http://git.openstack.org/cgit/openstack/ironic/tree/ironic/drivers/base.py#n96 | 23:17 |
jroll | github throwing unicorns :( | 23:17 |
jlvillal | jroll: Okay and thanks. | 23:17 |
jroll | np | 23:17 |
openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Ensure all methods in utils.py have docstrings https://review.openstack.org/220313 | 23:22 |
jroll | JayF: wow, that patch is still going? | 23:25 |
jroll | oh I thuoght it was an older one, ignore me | 23:25 |
JayF | jroll: it's had a mild change that's been needed for a long time | 23:25 |
JayF | jroll: so I jfdi just now | 23:25 |
jroll | ah | 23:25 |
*** sdake has quit IRC | 23:26 | |
JayF | jlvillal: vdrok: Congratulations | 23:26 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix AgentDeploy compatibility with pre-Liberty drivers https://review.openstack.org/231215 | 23:32 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix AgentDeploy compatibility with pre-Liberty drivers https://review.openstack.org/231215 | 23:32 |
jroll | damn that commit message | 23:32 |
* jroll fixes locally | 23:33 | |
jroll | devananda: ^ I think that's good to go now | 23:33 |
JayF | jroll: for verify -2, is a recheck sufficient? | 23:33 |
JayF | jroll: for https://review.openstack.org/#/c/232338/1 it doesn't seem to have budged jenkins | 23:33 |
jroll | JayF: yes, it's sufficient, give it a minute | 23:34 |
JayF | jroll: aweeks ^ We should also make this change in our downstream ramdisk image repo as well fwiw, that's a great find :) | 23:34 |
jroll | oh NICE | 23:34 |
jroll | I like | 23:35 |
JayF | actually, we don't have to change it downstream, b/c we use upstream dockerfile | 23:35 |
jroll | right | 23:35 |
* JayF prs a thing downstream anyway | 23:35 | |
*** mmedvede has joined #openstack-ironic | 23:36 | |
JayF | jroll: fwiw 232338 is still stuck, and I even workflow -1 / +1'd it | 23:55 |
jroll | JayF: I don't have any special knowledge that you don't :) | 23:55 |
jroll | hit up infra if you continue having problems | 23:55 |
JayF | already there | 23:56 |
JayF | just letting you know that waiting didn't work :) | 23:56 |
jroll | cool | 23:57 |
*** harshs has joined #openstack-ironic | 23:57 | |
*** garthb has quit IRC | 23:58 | |
*** achanda has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!