*** harshs has joined #openstack-ironic | 00:00 | |
*** Haomeng has quit IRC | 00:00 | |
*** [1]cdearborn has quit IRC | 00:00 | |
*** phschwartz has quit IRC | 00:01 | |
*** phschwartz has joined #openstack-ironic | 00:01 | |
*** praneshp_ has joined #openstack-ironic | 00:02 | |
*** praneshp has quit IRC | 00:04 | |
*** praneshp_ has quit IRC | 00:06 | |
*** Marga_ has quit IRC | 00:11 | |
*** praneshp has joined #openstack-ironic | 00:12 | |
*** kkoski has quit IRC | 00:13 | |
*** Marga_ has joined #openstack-ironic | 00:14 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add note about driver API breakage https://review.openstack.org/235026 | 00:16 |
---|---|---|
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add release notes for 4.2.1 https://review.openstack.org/235027 | 00:20 |
jroll | a couple release notes that I'd love eyes on | 00:20 |
*** achanda has quit IRC | 00:21 | |
*** shadower has quit IRC | 00:23 | |
*** shadower has joined #openstack-ironic | 00:23 | |
*** achanda has joined #openstack-ironic | 00:25 | |
*** Marga_ has quit IRC | 00:28 | |
*** mrmoje has quit IRC | 00:30 | |
*** garthb has quit IRC | 00:32 | |
*** Sukhdev has quit IRC | 00:34 | |
*** Sukhdev has joined #openstack-ironic | 00:38 | |
*** cemason has joined #openstack-ironic | 00:41 | |
lintan | morning, Ironic | 00:44 |
*** rloo has quit IRC | 00:53 | |
openstackgerrit | Yuiko Takada proposed openstack/ironic-inspector: Support IPA for devstack-plugin https://review.openstack.org/223563 | 00:56 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add note about driver API breakage https://review.openstack.org/235026 | 00:58 |
jroll | g'morning lintan :) | 00:58 |
jroll | I'm out for the night, have a good one y'all | 01:02 |
lintan | morning, jroll | 01:02 |
lintan | jroll, I am reading your spec about rolling upgrade, it looks good to me and I am willing to take the work | 01:04 |
jroll | lintan: excellent | 01:05 |
*** meghal1 has quit IRC | 01:05 | |
jroll | will you be in tokyo? | 01:05 |
lintan | jroll, yes, I will be there | 01:05 |
jroll | awesome, we should work some of it out then :) | 01:05 |
jroll | I'm heading out for real now, have a good day :) | 01:06 |
lintan | jroll, OK, have a good night | 01:06 |
*** penick has quit IRC | 01:10 | |
Haomeng|2 | morning lintan | 01:14 |
Haomeng|2 | jroll: good night:) | 01:14 |
*** Haomeng|2 is now known as Haomeng | 01:14 | |
*** zhenguo has joined #openstack-ironic | 01:14 | |
lintan | morning, Haomeng | 01:15 |
*** naohirot has joined #openstack-ironic | 01:16 | |
Haomeng | lintan: :) | 01:23 |
openstackgerrit | Zhenguo Niu proposed openstack/ironic: Convert set() to list in ListType https://review.openstack.org/234697 | 01:36 |
*** huangkai has joined #openstack-ironic | 01:39 | |
huangkai | hello | 01:40 |
huangkai | just want to drop a message here for the reviewers if you could get some time rate the spec https://review.openstack.org/#/c/208319/, and I'll appreciate it very much. | 01:43 |
*** spandhe has quit IRC | 01:43 | |
*** links has joined #openstack-ironic | 01:46 | |
*** baoli has joined #openstack-ironic | 01:56 | |
*** Haomeng|2 has joined #openstack-ironic | 01:59 | |
*** Haomeng has quit IRC | 02:02 | |
*** r-daneel has quit IRC | 02:08 | |
*** boris-42 has quit IRC | 02:08 | |
*** baoli has quit IRC | 02:13 | |
*** kozhukalov_ has joined #openstack-ironic | 02:18 | |
*** kan_ has joined #openstack-ironic | 02:18 | |
*** naohirot has quit IRC | 02:22 | |
huangkai | just want to drop a message here for the reviewers if you could get some time rate the spec https://review.openstack.org/#/c/208319/, and I'll appreciate it very much. | 02:23 |
*** harshs has quit IRC | 02:33 | |
openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add db migration and model for tags table https://review.openstack.org/192576 | 02:34 |
*** boris-42 has joined #openstack-ironic | 02:58 | |
openstackgerrit | Merged openstack/ironic: remove lxml requirement https://review.openstack.org/234981 | 03:03 |
*** chenglch has joined #openstack-ironic | 03:15 | |
*** dims_ has quit IRC | 03:18 | |
*** notq has quit IRC | 03:21 | |
*** Sukhdev has quit IRC | 03:22 | |
*** Sukhdev has joined #openstack-ironic | 03:22 | |
*** puranamr has joined #openstack-ironic | 03:26 | |
*** rameshg87 has joined #openstack-ironic | 03:35 | |
*** hblixt has joined #openstack-ironic | 03:40 | |
*** VikasC_ has joined #openstack-ironic | 03:43 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO drivers to use new boot interface https://review.openstack.org/229274 | 03:48 |
*** puranamr has quit IRC | 03:58 | |
*** Sukhdev has quit IRC | 03:58 | |
*** puranamr has joined #openstack-ironic | 03:58 | |
*** puranamr has quit IRC | 04:00 | |
*** Nisha has joined #openstack-ironic | 04:02 | |
*** puranamr has joined #openstack-ironic | 04:02 | |
*** puranamr has quit IRC | 04:05 | |
*** puranamr has joined #openstack-ironic | 04:06 | |
*** puranamr has quit IRC | 04:08 | |
*** MattMan has quit IRC | 04:10 | |
*** MattMan has joined #openstack-ironic | 04:10 | |
*** puranamr has joined #openstack-ironic | 04:12 | |
*** puranamr has quit IRC | 04:18 | |
*** dims has joined #openstack-ironic | 04:18 | |
*** puranamr has joined #openstack-ironic | 04:19 | |
*** puranamr has quit IRC | 04:22 | |
*** puranamr has joined #openstack-ironic | 04:23 | |
*** dims has quit IRC | 04:24 | |
*** puranamr has quit IRC | 04:30 | |
*** puranamr has joined #openstack-ironic | 04:32 | |
*** achanda has quit IRC | 04:32 | |
*** puranamr has quit IRC | 04:34 | |
*** puranamr_ has joined #openstack-ironic | 04:34 | |
*** spandhe has joined #openstack-ironic | 04:37 | |
*** Nisha has quit IRC | 04:37 | |
*** puranamr_ has quit IRC | 04:39 | |
*** Nisha has joined #openstack-ironic | 04:40 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 04:41 |
*** puranamr has joined #openstack-ironic | 04:41 | |
*** Sukhdev has joined #openstack-ironic | 04:54 | |
*** amotoki has quit IRC | 04:57 | |
*** amotoki_ has joined #openstack-ironic | 04:57 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 05:09 |
*** Nisha has quit IRC | 05:12 | |
*** puranamr has quit IRC | 05:16 | |
*** Nisha has joined #openstack-ironic | 05:18 | |
*** notq has joined #openstack-ironic | 05:18 | |
*** harshs has joined #openstack-ironic | 05:20 | |
*** dims has joined #openstack-ironic | 05:21 | |
*** Marga_ has joined #openstack-ironic | 05:24 | |
*** Marga_ has quit IRC | 05:24 | |
*** Marga_ has joined #openstack-ironic | 05:25 | |
*** achanda has joined #openstack-ironic | 05:25 | |
*** dims has quit IRC | 05:26 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Documentation for Swift-less iLO Virtual Media deploy https://review.openstack.org/228020 | 05:29 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO driver console interface into new module https://review.openstack.org/229284 | 05:34 |
*** Marga_ has quit IRC | 05:38 | |
openstackgerrit | Merged openstack/ironic: Convert set() to list in ListType https://review.openstack.org/234697 | 05:40 |
*** harshs_ has joined #openstack-ironic | 05:41 | |
Haomeng|2 | huangkai: welcome | 05:41 |
*** harshs has quit IRC | 05:43 | |
*** harshs_ is now known as harshs | 05:43 | |
*** rameshg871 has joined #openstack-ironic | 05:43 | |
*** lazy_prince has joined #openstack-ironic | 05:44 | |
*** rameshg87 has quit IRC | 05:44 | |
*** puranamr has joined #openstack-ironic | 05:48 | |
*** puranamr has quit IRC | 05:49 | |
*** puranamr has joined #openstack-ironic | 05:51 | |
*** priteau has joined #openstack-ironic | 05:51 | |
*** harshs has quit IRC | 05:59 | |
*** harshs has joined #openstack-ironic | 06:00 | |
*** kozhukalov_ has quit IRC | 06:05 | |
*** priteau has quit IRC | 06:11 | |
*** BharatK_ has joined #openstack-ironic | 06:14 | |
*** Marga_ has joined #openstack-ironic | 06:14 | |
*** spandhe has quit IRC | 06:18 | |
*** Sukhdev has quit IRC | 06:19 | |
*** meghal has joined #openstack-ironic | 06:19 | |
*** Sukhdev has joined #openstack-ironic | 06:19 | |
*** Marga_ has quit IRC | 06:21 | |
*** dims has joined #openstack-ironic | 06:22 | |
*** puranamr has quit IRC | 06:26 | |
*** dims has quit IRC | 06:27 | |
*** amotoki has joined #openstack-ironic | 06:32 | |
*** amotoki has quit IRC | 06:32 | |
*** yuikotakada has joined #openstack-ironic | 06:33 | |
*** kozhukalov_ has joined #openstack-ironic | 06:37 | |
*** romainh has joined #openstack-ironic | 06:38 | |
*** jamielennox is now known as jamielennox|away | 06:39 | |
*** pas-ha has joined #openstack-ironic | 06:40 | |
*** getvasanth has joined #openstack-ironic | 06:46 | |
*** jcoufal has joined #openstack-ironic | 06:49 | |
openstackgerrit | Merged openstack/ironic: Update python-oneviewclient version https://review.openstack.org/234428 | 06:51 |
*** Sukhdev has quit IRC | 06:56 | |
*** meghal1 has joined #openstack-ironic | 07:04 | |
*** meghal has quit IRC | 07:06 | |
openstackgerrit | Haomeng,Wang proposed openstack/bifrost: Initial addition of an inspection module https://review.openstack.org/232745 | 07:12 |
*** dtantsur|afk is now known as dtantsur | 07:13 | |
dtantsur | Morning Ironic | 07:13 |
*** meghal1 has quit IRC | 07:13 | |
yuikotakada | dtantsur, o/ | 07:14 |
dtantsur | yuikotakada, o/ how are you doing? | 07:14 |
yuikotakada | dtantsur, I've been back to community from today finally | 07:15 |
*** harshs has quit IRC | 07:15 | |
dtantsur | that's great | 07:15 |
yuikotakada | dtantsur, yeah, I'm checking our etherpad... thank you for putting it on ML! | 07:16 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor agent_ilo driver to use new boot interface https://review.openstack.org/217102 | 07:16 |
*** cemason has quit IRC | 07:20 | |
*** cemason has joined #openstack-ironic | 07:20 | |
*** achanda has quit IRC | 07:27 | |
*** ifarkas has joined #openstack-ironic | 07:34 | |
*** achanda has joined #openstack-ironic | 07:36 | |
*** achanda has quit IRC | 07:37 | |
*** achanda has joined #openstack-ironic | 07:39 | |
*** mrmoje has joined #openstack-ironic | 07:41 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 07:46 |
*** amotoki_ has quit IRC | 07:55 | |
*** priteau has joined #openstack-ironic | 07:57 | |
*** yog_ has joined #openstack-ironic | 07:59 | |
*** Haomeng has joined #openstack-ironic | 08:00 | |
*** pradipta has joined #openstack-ironic | 08:02 | |
*** Haomeng|2 has quit IRC | 08:03 | |
*** achanda has quit IRC | 08:03 | |
*** amotoki has joined #openstack-ironic | 08:06 | |
*** achanda has joined #openstack-ironic | 08:12 | |
*** marzif has joined #openstack-ironic | 08:12 | |
*** mkovacik has joined #openstack-ironic | 08:14 | |
*** lucasagomes has joined #openstack-ironic | 08:17 | |
*** huangkai has quit IRC | 08:17 | |
*** marzif has quit IRC | 08:19 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO drivers to use new boot interface https://review.openstack.org/229274 | 08:19 |
*** marzif has joined #openstack-ironic | 08:20 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 08:21 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO driver console interface into new module https://review.openstack.org/229284 | 08:22 |
*** derekh has joined #openstack-ironic | 08:22 | |
*** praneshp has quit IRC | 08:24 | |
*** dims has joined #openstack-ironic | 08:25 | |
*** kozhukalov_ has quit IRC | 08:26 | |
*** cemason has quit IRC | 08:27 | |
*** cemason has joined #openstack-ironic | 08:27 | |
openstackgerrit | chenglch proposed openstack/ironic: set-console-mode should validate terminal command https://review.openstack.org/235159 | 08:28 |
*** jistr has joined #openstack-ironic | 08:30 | |
*** dims has quit IRC | 08:30 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor agent_ilo driver to use new boot interface https://review.openstack.org/217102 | 08:36 |
*** chenke has joined #openstack-ironic | 08:39 | |
chenke | Haomeng, how are you | 08:40 |
Haomeng | chenke: fine, and you? | 08:42 |
Haomeng | chenke: :) | 08:42 |
*** amotoki has quit IRC | 08:43 | |
chenke | i have node with wrong ip addr, ironic takes long time to change this node state to unuseable, anyboy know how to short this time. | 08:46 |
*** amotoki has joined #openstack-ironic | 08:53 | |
*** mbound has joined #openstack-ironic | 08:55 | |
chenke | Haomeng, fine too. do you encounter the question i asked above | 08:56 |
*** shuquan has joined #openstack-ironic | 08:56 | |
*** BharatK_ has quit IRC | 08:58 | |
*** subscope has joined #openstack-ironic | 08:59 | |
chenke | Haomeng, if i set a wrong ip addr, it takes long time for ironic to change the Maintenance to True | 08:59 |
Haomeng | chenke: maybe I lost connection during your question raised,can you paste again? | 09:00 |
chenke | i have node with wrong ip addr, ironic takes long time to change the Maintenance attribute of this node to True | 09:01 |
chenke | Haomeng, i know the ironic will sync power state 3 times, but seems the interval time is too long. it last 14 min among them. | 09:02 |
Haomeng | chenke: yes, we can check cond's log, maybe it try several times to get ipmi connection | 09:02 |
Haomeng | chenke: this can be improved I thinnk | 09:03 |
Haomeng | chenke: let me double check latest code and confirm the behavior | 09:03 |
*** amotoki has quit IRC | 09:03 | |
*** marzif_ has joined #openstack-ironic | 09:04 | |
dtantsur | chenke, if on Liberty, you can start using ENROLL state | 09:04 |
chenke | in the ironic.conf, the value "sync_power_state_interval=60s", seems it dosen't work as config set. | 09:04 |
*** marzif has quit IRC | 09:04 | |
dtantsur | chenke, http://docs.openstack.org/developer/ironic/deploy/install-guide.html#api-version-1-11-and-above | 09:04 |
dtantsur | this code path will verify your power credentials when moving enroll->manageable | 09:05 |
Haomeng | chenke: and we have option CONF.conductor.power_state_sync_max_retries which can setthe max retries | 09:07 |
chenke | dtantsur, tku, let me check that | 09:07 |
dtantsur | Haomeng, I don't think it's good to change these values to anything substantially lower | 09:08 |
chenke | Haomeng, yes, i know this max retries. but the issue is the interval time is too long. | 09:08 |
Haomeng | dtantsur: yes, default value 3 should good | 09:08 |
dtantsur | chenke, the lower this interval, the higher are chances of false negative, e.g. your nodes going to maintenance mode due to temporary network problem | 09:09 |
Haomeng | chenke: let me check ipmi operation tiimeout settings | 09:09 |
Haomeng | totaltime=retries * ipmi timeout,I think | 09:10 |
dtantsur | mkovacik, hi there! how is your Mac patch doing? | 09:10 |
lucasagomes | morning all | 09:12 |
lucasagomes | :-) | 09:12 |
Haomeng | lucasagomes: morning:) | 09:12 |
*** BharatK_ has joined #openstack-ironic | 09:12 | |
*** shuquan has quit IRC | 09:13 | |
Haomeng | chenke: default retry_timeout =60 | 09:13 |
yuikotakada | lucasagomes, morning :) | 09:13 |
rameshg871 | vdrok: can you please upgrade your vote and +A this - https://review.openstack.org/#/c/222264/ | 09:13 |
rameshg871 | :) | 09:13 |
vdrok | morning ironic | 09:13 |
Haomeng | chenke: so total time = 60 * 3 | 09:13 |
Haomeng | chenke: as design | 09:13 |
vdrok | and morning Haomeng chenke dtantsur rameshg871 yuikotakada | 09:14 |
rameshg871 | vdrok: o/ | 09:14 |
vdrok | rameshg871, looking | 09:14 |
Haomeng | vdrok: morning:) | 09:14 |
dtantsur | morning vdrok, lucasagomes, hi, rameshg871 | 09:14 |
chenke | Haomeng, if like what you said, the time should be 60. but now interval are >10 min | 09:14 |
yuikotakada | vdrok, rameshg871, Haomeng, morning :) | 09:14 |
*** shuquan has joined #openstack-ironic | 09:14 | |
rameshg871 | yuikotakada: dtantsur: o/ | 09:14 |
Haomeng | chenke: can you show your log to see more details where is the time consumed | 09:14 |
chenke | yes. | 09:14 |
Haomeng | chenke: yes, so this should be issue | 09:14 |
*** achanda has quit IRC | 09:15 | |
vdrok | rameshg871, done :) | 09:15 |
chenke | Haomeng, i treat it as a not exprience for enduser. | 09:15 |
rameshg871 | vdrok: thanks :) | 09:15 |
chenke | Haomeng, i mean it is not good | 09:16 |
Haomeng | chenke: yes,should not consume more time then 3*60 based on default configuration | 09:16 |
Haomeng | chenke: did you change default configurations? | 09:16 |
chenke | Haomeng, no, i don't. how can i upload the log to u | 09:17 |
Haomeng | chenke: ok,thank you | 09:17 |
Haomeng | chenke: we can grep "During sync_power_state, node" | 09:17 |
Haomeng | chenke: to see the details | 09:18 |
Haomeng | chenke: and "During sync_power_state, max retries exceeded..." | 09:18 |
chenke | Haomeng, i have done it. the interval is not 60s | 09:20 |
Haomeng | chenke: maybewe have other ipmi call during sync_power, so they are conflict each other, to consumemore time | 09:20 |
Haomeng | chenke: ok | 09:20 |
chenke | Haomeng, i can see the ipmi cmd get err soon because of wrong ip addr. | 09:21 |
chenke | 2015-10-15 16:45:12.528 8481 WARNING ironic.drivers.modules.ipmitool [-] IPMI power status failed for node 69ba31df-bf0a-4953-9004-8e7a605b2005 with error: Failed to create the password file. Unexpected error while running command | 09:23 |
chenke | 2015-10-15 16:58:49.575 8481 WARNING ironic.drivers.modules.ipmitool [-] IPMI power status failed for node 69ba31df-bf0a-4953-9004-8e7a605b2005 with error: Failed to create the password file. Unexpected error while running command | 09:23 |
chenke | 2015-10-15 17:12:26.768 8481 WARNING ironic.drivers.modules.ipmitool [-] IPMI power status failed for node 69ba31df-bf0a-4953-9004-8e7a605b2005 with error: Failed to create the password file. Unexpected error while running command | 09:24 |
chenke | 2015-10-15 17:12:26.783 8481 ERROR ironic.conductor.manager [-] During sync_power_state, max retries exceeded for node 69ba31df-bf0a-4953-9004-8e7a605b2005, node state None does not match expected state 'None'. Updating DB state to 'None' Switching node to maintenance mode. | 09:24 |
chenke | Haomeng, you can check above log | 09:25 |
Haomeng | chenke: so itis from 16:45 to 17:12? | 09:26 |
*** e0ne has joined #openstack-ironic | 09:26 | |
*** dims has joined #openstack-ironic | 09:27 | |
*** mrmoje has quit IRC | 09:27 | |
chenke | Haomeng, yes. you can see the interval is 60s, but >10 min. this is what i mind | 09:27 |
chenke | Haomeng, the interval is not 60s, sorry for type wrong | 09:28 |
Haomeng | chenke: yes, between two ipmi power operations, the consumed time is long - around 7 mins | 09:28 |
*** mrmoje has joined #openstack-ironic | 09:29 | |
chenke | Haomeng, i want to short this time. do you have any idea? | 09:30 |
*** dims has quit IRC | 09:32 | |
*** shuquan has quit IRC | 09:34 | |
*** achanda has joined #openstack-ironic | 09:34 | |
*** shuquan has joined #openstack-ironic | 09:36 | |
*** chenke has quit IRC | 09:38 | |
*** pradipta has quit IRC | 09:38 | |
*** chenke has joined #openstack-ironic | 09:38 | |
*** shuquan has quit IRC | 09:40 | |
openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add db api layer for CRUD operations on node tags https://review.openstack.org/232598 | 09:41 |
*** shuquan has joined #openstack-ironic | 09:41 | |
Haomeng | chenke: will check code behavior,so far no idea | 09:43 |
Haomeng | chenke: should have bug I think | 09:43 |
Haomeng | chenke: willhelp to investigate | 09:44 |
*** _degorenko is now known as degorenko | 09:45 | |
*** getvasanth_ has joined #openstack-ironic | 09:45 | |
*** pas-ha has quit IRC | 09:46 | |
*** links has quit IRC | 09:47 | |
*** getvasanth has quit IRC | 09:48 | |
*** dims has joined #openstack-ironic | 09:51 | |
*** achanda has quit IRC | 09:55 | |
openstackgerrit | Merged openstack/ironic: Refactor deploy_utils methods https://review.openstack.org/222264 | 10:00 |
*** chenglch has quit IRC | 10:03 | |
Haomeng | chenke: looks like it is not caused by wrong ip, but the CONF.tempdir folder no space or permision,so log indicate: Failed to create the password file. | 10:03 |
Haomeng | chenke: did you get these log continus lines in log? we need to check with timestamps | 10:04 |
*** shuquan has quit IRC | 10:05 | |
mkovacik | dtantsur, hi, just returned from lunch (forgotten to set AFK) got it working, now about to write the commit message | 10:07 |
mkovacik | any suggestions? | 10:07 |
dtantsur | lemme find a link | 10:07 |
dtantsur | mkovacik, https://wiki.openstack.org/wiki/GitCommitMessages | 10:08 |
mkovacik | dtantsur, precisely what I had in mind ;) thanks | 10:08 |
mkovacik | btw I'm running pep8 on the test case, any switches I should apply? | 10:09 |
* mkovacik concerned with e.g. line too long | 10:09 | |
dtantsur | mkovacik, if `tox -epep8` works, you're fine :) | 10:09 |
*** zhenguo has quit IRC | 10:09 | |
mkovacik | dtantsur, ok great :) | 10:10 |
openstackgerrit | Rico Lin proposed openstack/python-ironicclient: improve readme contents https://review.openstack.org/235228 | 10:15 |
*** subscope has quit IRC | 10:16 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 10:26 |
*** BharatK_ has quit IRC | 10:30 | |
sambetts | Morning all o/ | 10:36 |
dtantsur | morning sambetts | 10:37 |
*** Nisha has quit IRC | 10:37 | |
yuikotakada | sambetts, o/ | 10:38 |
sambetts | Hey yuikotakada, dtantsur :D | 10:39 |
*** romainh has quit IRC | 10:39 | |
vdrok | morning sambetts | 10:40 |
*** alexpilotti has joined #openstack-ironic | 10:41 | |
sambetts | o/ vdrok | 10:43 |
*** BharatK_ has joined #openstack-ironic | 10:43 | |
*** pelix has joined #openstack-ironic | 10:44 | |
sambetts | dtantsur: Why the need for the addition of kmod in that patch? | 10:48 |
dtantsur | sambetts, rtslib uses modprobe internally. I'll add it to the commit message | 10:49 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Never run Flask application with debug mode https://review.openstack.org/235258 | 10:53 |
sambetts | dtantsur: Ah :) ok, I thought modprobe would be built in :/ | 10:53 |
dtantsur | so did I, but seems like it's not | 10:53 |
dtantsur | sambetts, and believe or not, we have one more mysql-related issue >_< https://bugs.launchpad.net/ironic-inspector/+bug/1506160 | 10:54 |
openstack | Launchpad bug 1506160 in Ironic Inspector mitaka "Database schema requires InnoDB engine with MySQL" [High,Triaged] - Assigned to Dmitry Tantsur (divius) | 10:54 |
dtantsur | we have 2 bugs targeted to stable/liberty now (this + https://bugs.launchpad.net/ironic-inspector/+bug/1506419) | 10:54 |
openstack | Launchpad bug 1506419 in Ironic Inspector mitaka "Running Flask server in debug mode may be a security issue" [High,In progress] - Assigned to Dmitry Tantsur (divius) | 10:54 |
sambetts | ;( | 10:55 |
*** subscope has joined #openstack-ironic | 10:56 | |
sambetts | dtantsur: With the debug mode thing, can't we just leave it as a configuration option? and make the default=False with a massive warning for those that turn it to True? | 11:02 |
dtantsur | sambetts, then we need a new configuration option with a not-too-clear purpose.. why would a user set it to true? | 11:02 |
sambetts | dtantsur: ah so the problem is the overlap between the oslo log debug option and the flask one, because sometimes you want oslo log debug but not flask debug ? | 11:04 |
dtantsur | sambetts, well, to be honest I don't think anyone needs flask debug at all... | 11:04 |
* rameshg871 goes home | 11:08 | |
*** rameshg871 has quit IRC | 11:08 | |
sambetts | dtantsur: hmm, just looking into it more on the flask site, it doesn't seem to add much other than enabling the debugger, I was concerned that if we removed flask debug mode it wouldn't display any flask middleware debug logging | 11:08 |
dtantsur | ok, now our gate is broken by a devstack change, awesome | 11:10 |
sambetts | great :( | 11:13 |
*** [1]cdearborn has joined #openstack-ironic | 11:14 | |
*** yuikotakada has quit IRC | 11:16 | |
sambetts | dtantsur: looking in the mySQL issue now, why pick a relational database and then kill the relationships by using an engine that doesn't support foreign keys ;'( might as well use mongo | 11:16 |
dtantsur | yeah... | 11:16 |
dtantsur | crazy bugs today, I'm already pissed off... | 11:16 |
sambetts | XD | 11:17 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix gate broken by the devstack trueorfalse change https://review.openstack.org/235265 | 11:20 |
dtantsur | sambetts, ^^ | 11:20 |
sambetts | dtantsur: What do you think about the mySQL one, do we remove the foreign key relationship and then let the sqlalchemy/manually deal with the relationships? | 11:25 |
sambetts | or force the engine type | 11:25 |
dtantsur | sambetts, I'll force the engine type | 11:27 |
*** romainh has joined #openstack-ironic | 11:27 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Never run Flask application with debug mode https://review.openstack.org/235258 | 11:28 |
dtantsur | rebased on top of gate fix ^^ | 11:28 |
sambetts | dtantsur, does that work per table? is it a global setting in mySQL? I'm just wondering if forcing the engine type will affect other projects tables | 11:29 |
dtantsur | sambetts, it's per table, I'm mostly copying what ironic did | 11:29 |
sambetts | dtantsur: Ah ok :) | 11:29 |
*** alexpilotti has quit IRC | 11:33 | |
*** [1]cdearborn has quit IRC | 11:35 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Always default to InnoDB for MySQL https://review.openstack.org/235274 | 11:37 |
dtantsur | TheJulia, ^^ | 11:37 |
*** BharatK_ has quit IRC | 11:38 | |
*** BharatK has joined #openstack-ironic | 11:39 | |
TheJulia | \o/ | 11:39 |
*** thrash|g0ne is now known as thrash | 11:39 | |
TheJulia | dtantsur: thank you! | 11:39 |
dtantsur | TheJulia, mind testing it? | 11:40 |
sambetts | dtantsur: We don't need to make any changes to our migrations do we? | 11:40 |
dtantsur | oh, true, good catch | 11:40 |
TheJulia | dtantsur: not at all, just helping an internal person out with some IPA issues first, so give me an hour or so | 11:40 |
dtantsur | TheJulia, wait a bit, I didn't update the migrations | 11:40 |
sambetts | :-P | 11:40 |
TheJulia | dtantsur: 10-4 | 11:41 |
openstackgerrit | Sam Betts proposed openstack/ironic-python-agent: Add tinyipa to IPA imagebuild directory https://review.openstack.org/234902 | 11:42 |
*** BharatK has quit IRC | 11:43 | |
jroll | morning everyone :) | 11:44 |
sambetts | o/ jroll | 11:45 |
vdrok | morning jroll and TheJulia | 11:45 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Always default to InnoDB for MySQL https://review.openstack.org/235274 | 11:46 |
dtantsur | must be good now ^^ | 11:46 |
dtantsur | morning jroll | 11:46 |
*** trown|outttypeww is now known as trown | 11:47 | |
sambetts | dtantsur: :'( having to had 2 lines to every table create sucks, at least we don't have many tables yet | 11:48 |
jroll | heya sambetts, vdrok, dtantsur \o | 11:49 |
openstackgerrit | Milan Kováčik proposed openstack/ironic: mocking sendfile to make the test run on OSX https://review.openstack.org/235277 | 11:49 |
dtantsur | gate gods are not friendly to me today :( | 11:50 |
sambetts | :( | 11:50 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 11:52 |
dtantsur | dropped tgt from dependencies to figure out which version is actually running in gate ^^ | 11:53 |
*** BharatK has joined #openstack-ironic | 11:53 | |
mkovacik | dtantsur, https://review.openstack.org/#/c/235277/ | 11:54 |
*** baoli has joined #openstack-ironic | 11:55 | |
*** getvasanth__ has joined #openstack-ironic | 11:55 | |
* mkovacik is away: coffee | 11:56 | |
*** marzif_ has quit IRC | 11:57 | |
*** marzif_ has joined #openstack-ironic | 11:58 | |
*** baoli has quit IRC | 11:58 | |
*** getvasanth_ has quit IRC | 11:58 | |
*** mrmoje has quit IRC | 12:00 | |
*** mrmoje has joined #openstack-ironic | 12:01 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 12:02 |
*** caiobo has joined #openstack-ironic | 12:03 | |
*** baoli has joined #openstack-ironic | 12:03 | |
*** baoli has quit IRC | 12:04 | |
dtantsur | damn, we get one more gate failure, now seems related to ironic itself: http://logs.openstack.org/58/235258/2/check/gate-ironic-inspector-dsvm/03a80f4/logs/devstacklog.txt.gz | 12:04 |
dtantsur | is ironic gate even ok? | 12:04 |
*** sambetts has quit IRC | 12:04 | |
*** sambetts has joined #openstack-ironic | 12:05 | |
*** e0ne has quit IRC | 12:06 | |
*** baoli has joined #openstack-ironic | 12:08 | |
TheJulia | dtantsur: PRIMARY KEY (`uuid`) | 12:08 |
TheJulia | ) ENGINE=InnoDB DEFAULT CHARSET=utf8; | 12:08 |
TheJulia | I think thats a sign that it works! | 12:08 |
TheJulia | :) | 12:08 |
*** baoli has quit IRC | 12:08 | |
dtantsur | good to know :) now to fix 2 gate breakages....... | 12:09 |
TheJulia | ugh | 12:09 |
*** baoli has joined #openstack-ironic | 12:09 | |
*** JoshNang has quit IRC | 12:09 | |
*** JoshNang has joined #openstack-ironic | 12:09 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add release notes for 4.2.1 https://review.openstack.org/235027 | 12:11 |
*** thiagop has joined #openstack-ironic | 12:13 | |
thiagop | Good morning Ironicers | 12:13 |
thiagop | jroll: 4.2.1 is just bugfix? The oneview driver will be on this one? | 12:14 |
jroll | thiagop: just bugfixes on stable/liberty | 12:15 |
jroll | oneview will be in 4.3.0 | 12:15 |
thiagop | jroll: right, tks | 12:15 |
* jroll wonders how much stuff would be in 4.3 if we released before summit | 12:15 | |
jroll | np | 12:15 |
*** bradjones has joined #openstack-ironic | 12:16 | |
*** bradjones has quit IRC | 12:16 | |
*** bradjones has joined #openstack-ironic | 12:16 | |
thiagop | jroll: lp has something about 4.3 milestone? | 12:16 |
* thiagop know people hater lp, but... | 12:16 | |
thiagop | knows* | 12:16 |
dtantsur | so, looks like https://review.openstack.org/#/c/231814/ broke ironic-inspector gate, will figure out after lunch if nobody does it before.. | 12:17 |
jroll | thiagop: https://launchpad.net/ironic/+milestone/4.3.0 | 12:17 |
jroll | I haven't targeted anything else there yet | 12:17 |
openstackgerrit | Allan Vidal proposed stackforge/pyghmi: Fix TS inventory parsing bugs (DIMM, PSU) https://review.openstack.org/234953 | 12:18 |
jroll | once the docs land I'll mark oneview complete | 12:18 |
thiagop | jroll: awesome. We'll work on that again tomorrow. Some HP folks asked for CLI examples for node enrollment and so on... | 12:19 |
thiagop | or today, if the day allows it... | 12:19 |
thiagop | jroll: did you see my question yesterday about boot interface validation always failing? | 12:20 |
jroll | thiagop: no? | 12:20 |
thiagop | jroll: when we use "deploy_[kernel, ramdisk]" it fails asking for kernel, ramdisk and image_source | 12:21 |
jroll | thiagop: got a bug filed? | 12:21 |
jroll | and does that affect every agent driver or just oneview? | 12:22 |
thiagop | not yet, just probing for that to be the expected behavior, though it's very weird | 12:22 |
* mkovacik is back (gone 00:26:40) | 12:22 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Update troubleshooting text for custom IPA images https://review.openstack.org/235291 | 12:23 |
thiagop | jroll: I'll test it on a devstack before file a bug to be sure | 12:23 |
TheJulia | thiagop: there are variances based on the driver, example, the ilo driver with virtual media | 12:25 |
*** amotoki has joined #openstack-ironic | 12:26 | |
jroll | thiagop: ok, thanks | 12:26 |
thiagop | TheJulia: it was pxe boot | 12:26 |
jroll | mkovacik: mind turning off the away/back messages to the channel? :) | 12:26 |
jroll | mkovacik: also, hello, I don't believe we've met! | 12:26 |
TheJulia | thiagop: :( | 12:26 |
mkovacik | jroll, hi, yeah, will change the setting | 12:27 |
thiagop | But I'll look into that further. | 12:27 |
jroll | thanks! | 12:27 |
*** marzif_ has quit IRC | 12:28 | |
lucasagomes | jroll, zanata! damn I missed that | 12:28 |
*** marzif_ has joined #openstack-ironic | 12:28 | |
*** phschwartz has quit IRC | 12:29 | |
*** phschwartz has joined #openstack-ironic | 12:29 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 12:29 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add agent_amt docs https://review.openstack.org/235293 | 12:30 |
jroll | lucasagomes: heh, saw you were still working on transifex stuff :( | 12:30 |
jroll | ^ that docs patch should be an easy one | 12:30 |
lucasagomes | jroll, lol yeah | 12:30 |
lucasagomes | from time to time I translated stuff there... they should delete that tho | 12:31 |
jroll | heh yeah | 12:31 |
jroll | or have a big warning message | 12:31 |
lucasagomes | yeah | 12:31 |
lucasagomes | cause I don't follow the stuff going on in the i18n, I'm just a casual translator | 12:31 |
lucasagomes | anyway... will talk to them | 12:31 |
jroll | cool | 12:32 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Fix gate broken by the devstack trueorfalse change https://review.openstack.org/235265 | 12:33 |
*** puranamr has joined #openstack-ironic | 12:33 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add OneView driver documentation https://review.openstack.org/234396 | 12:33 |
jroll | just changed commit message there | 12:34 |
thiagop | righ, tks | 12:35 |
thiagop | right* | 12:35 |
*** BharatK has quit IRC | 12:36 | |
*** kozhukalov_ has joined #openstack-ironic | 12:38 | |
*** phschwartz has quit IRC | 12:41 | |
*** VikasC_ has quit IRC | 12:42 | |
*** phschwartz has joined #openstack-ironic | 12:42 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: ironic-inspector install doc polish https://review.openstack.org/234810 | 12:43 |
*** puranamr has quit IRC | 12:43 | |
sambetts | lucasagomes: https://review.openstack.org/#/c/234902 if you want to try it out :) | 12:44 |
lucasagomes | sambetts, hi there, awesome stuff! Will add to the TODO list here | 12:44 |
*** puranamr has joined #openstack-ironic | 12:45 | |
sambetts | lucasagomes: :D | 12:45 |
lucasagomes | sambetts, btw since it's a bunch of bash scripts, are you thinking about creating a DIB element for it too? | 12:45 |
*** phschwartz has quit IRC | 12:45 | |
sambetts | lucasagomes: i'll need to look into it, but I don't know why it couldn't be | 12:46 |
lucasagomes | yeah it totally can be | 12:46 |
lucasagomes | but maybe you just don't want to work on it, so you can leave for someone else to do | 12:46 |
lucasagomes | but would be cool | 12:46 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 12:49 |
*** phschwartz has joined #openstack-ironic | 12:49 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 12:52 |
openstackgerrit | Milan Kováčik proposed openstack/ironic: mocking sendfile to make the test run on OSX https://review.openstack.org/235277 | 12:55 |
*** phschwartz has quit IRC | 12:58 | |
*** phschwartz has joined #openstack-ironic | 12:58 | |
*** rloo has joined #openstack-ironic | 13:00 | |
*** amotoki has quit IRC | 13:00 | |
dtantsur | mkovacik, "abandon" button is used when you're no longer planning on working on a patch. not sure it's what you wanted | 13:00 |
mkovacik | dtantsur, ah sh*t wanted just to cancel the last patchset | 13:01 |
dtantsur | mkovacik, you can't cancel the patchset, but you can mark it as work-in-progress (workflow -1 button in "review") | 13:01 |
dtantsur | mkovacik, also of course you can restore an abandoned patchset | 13:01 |
mkovacik | dtantsur, thanx | 13:02 |
* mkovacik will try to remember that | 13:02 | |
*** marzif_ has quit IRC | 13:02 | |
*** marzif_ has joined #openstack-ironic | 13:03 | |
openstackgerrit | Milan Kováčik proposed openstack/ironic: mocking sendfile to make the test run on OSX https://review.openstack.org/235277 | 13:05 |
*** cemason has quit IRC | 13:06 | |
*** marzif_ has quit IRC | 13:08 | |
openstackgerrit | Tan Lin proposed openstack/ironic: Validate the input of properties of nodes. https://review.openstack.org/215505 | 13:09 |
jroll | dtantsur: verified bugs go to confirmed or triaged, generally? | 13:11 |
dtantsur | jroll, generally triaged, unless we have no idea what's going on | 13:11 |
jroll | cool, thanks | 13:11 |
* jroll moving a thing out of in progress | 13:12 | |
*** amotoki has joined #openstack-ironic | 13:12 | |
mkovacik | dtantsur, w/r the mock.patch('<namespace>.open') i'm not quite sure how to do it properly, since i need to patch it on 2 spots: the test case namespace and the ironic module namespace and of course there's no "open" in either of those (the patch fails unless create=True specified)... what do you think? | 13:12 |
* mkovacik not sure how to patch it globally and not to break stuff | 13:12 | |
dtantsur | mkovacik, hmmm, gotcha... I don't know what to do best, let's leave it as it is. anyway, it's better to move all mocks to be @decorators | 13:14 |
dtantsur | and let's see what other folks say | 13:14 |
mkovacik | dtantsur, there's an issue with that | 13:15 |
jroll | jlvillal: you say you've submitted a patch for this, but the patch isn't linked. is that an ironic patch or something else? should we leave the bug open? https://bugs.launchpad.net/ironic/+bug/1503086 | 13:15 |
openstack | Launchpad bug 1503086 in Ironic "Mis-formatted calls to log.* are not detected in unit tests" [Medium,In progress] - Assigned to John L. Villalovos (happycamp) | 13:15 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO drivers to use new boot interface https://review.openstack.org/229274 | 13:15 |
mkovacik | dtantsur, the openstack.module patch is better kept as with statement , as then you're able to work with the ctx manager returned | 13:15 |
mkovacik | if you make it a decorator, the reference is lost | 13:16 |
dtantsur | mkovacik, no, it's provided to you as a decorated function argument | 13:16 |
mkovacik | dtantsur, nope, because you specify new=<stuff> | 13:16 |
mkovacik | which changes the patch behaviour | 13:17 |
* mkovacik had a face palm because of that | 13:17 | |
dtantsur | mkovacik, oh yeah, I forgot that you use new= | 13:17 |
mkovacik | dtantsur, right ;) | 13:18 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 13:18 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iLO driver console interface into new module https://review.openstack.org/229284 | 13:19 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Always default to InnoDB for MySQL https://review.openstack.org/235274 | 13:25 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Never run Flask application with debug mode https://review.openstack.org/235258 | 13:26 |
* mkovacik is away: I'm busy | 13:27 | |
dtantsur | sambetts, so seems like https://review.openstack.org/#/c/235294 and the previous patch fix gate for us | 13:27 |
dtantsur | now we only need to make it land... | 13:27 |
*** puranamr has quit IRC | 13:27 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 13:29 |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic-specs: Ironic Group Management https://review.openstack.org/235344 | 13:30 |
*** ionutbalutoiu has joined #openstack-ironic | 13:31 | |
*** puranamr has joined #openstack-ironic | 13:32 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor agent_ilo driver to use new boot interface https://review.openstack.org/217102 | 13:33 |
*** kozhukalov has quit IRC | 13:34 | |
*** kozhukalov has joined #openstack-ironic | 13:34 | |
*** mrmoje has quit IRC | 13:36 | |
*** mehip has joined #openstack-ironic | 13:38 | |
lucasagomes | jroll, JayF when you guys have some time, mind taking a look at https://review.openstack.org/#/c/234339/ ? | 13:41 |
jroll | lucasagomes: sure thing, I'll trade you for https://review.openstack.org/#/c/220898/ | 13:43 |
jroll | :P | 13:43 |
lucasagomes | jroll, fair | 13:43 |
lucasagomes | lol | 13:43 |
* lucasagomes looks | 13:43 | |
openstackgerrit | Merged stackforge/pyghmi: Fix TS inventory parsing bugs (DIMM, PSU) https://review.openstack.org/234953 | 13:43 |
lucasagomes | oh yeah that's important, I reviewed it before... /me re-reviews | 13:43 |
jroll | yeah, it lgtm | 13:43 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 13:44 |
lucasagomes | ramsh<tab> | 13:48 |
lucasagomes | jroll, with that patch, the bash ramdisk will boot but the get_clean_steps() will return [] because the agent_url is not set | 13:50 |
lucasagomes | and clean is finished (machine powered down etc...) | 13:50 |
lucasagomes | that's what it does right? (checking if I understood it correctly) | 13:50 |
sambetts | lucasagomes: Thats how I understood it | 13:51 |
lucasagomes | sambetts, thanks! Right, it's tricky but I don't have a better solution if we want to add it prior to removing the bash ramdisk support | 13:51 |
rloo | hi everyone, lucasagomes, sambetts, jroll, dtantsur | 13:53 |
sambetts | lucasagomes: it seems like a logical way to do it from my perspective, of course we can remove that check once bash is fully deprecated :) | 13:53 |
sambetts | o/ rloo | 13:53 |
rloo | jroll: wrt 4.2.1, did you see this patch (and would it affect release notes if it went in): https://review.openstack.org/#/c/235109/ | 13:53 |
jroll | lucasagomes: correct, and it still leaves room for OOB clean steps to be mixed in | 13:54 |
lucasagomes | sambetts, yeah, the only thing that sucks is that it's going to boot a deploy ramdisk, which will expose the hardware as an iscsi target with the previous tenant data on it | 13:54 |
lucasagomes | for a window of time | 13:54 |
lucasagomes | but ... well | 13:54 |
dtantsur | morning rloo | 13:54 |
lucasagomes | I don't have a better idea to do that if we want to add the support for that now (with the bash ramdisk still being supported) | 13:54 |
lucasagomes | rloo, morning | 13:54 |
jroll | rloo: morning, no, nobody has told me about that until now. | 13:55 |
rloo | jroll: I just saw it. am about to +A your release notes, so thought I'd check first. | 13:55 |
jroll | rloo: it doesn't seem high severity so I'm kind of inclined to wait. +A is fine, we can update later if we decide to get it in | 13:56 |
rloo | jroll: ok | 13:56 |
* jroll will think about it | 13:56 | |
sambetts | lucasagomes: :( ah, yes, won't be a problem once we manage to get the network stuff in because provisioning would have been in a separate network | 13:56 |
*** Nisha has joined #openstack-ironic | 13:56 | |
* rloo won't think about it | 13:56 | |
jroll | hah | 13:56 |
lucasagomes | sambetts, yeah, cleaning has it's own network as well so should be fine | 13:57 |
jroll | lucasagomes: now you have me wondering if the bash ramdisk will be shut back down... also something to think about, if people are using the bash ramdisk, the tenant data never gets cleaned up anyway :/ | 13:57 |
lucasagomes | +a | 13:57 |
jroll | \o/ \o/ \o/ | 13:57 |
lucasagomes | jroll, at the end of the cleaning it shutsdown | 13:57 |
jroll | all the +A going on in here | 13:57 |
jroll | lucasagomes: yeah, that's what I figured | 13:58 |
NobodyCam | good morning ironicers | 13:58 |
jroll | hey hey NobodyCam | 13:58 |
sambetts | o/ NobodyCam | 13:58 |
lucasagomes | jroll, one thing we could do, which may help a bit with it is to destroy the disks metadata when the bash ramdisk boot | 13:58 |
dtantsur | morning NobodyCam | 13:58 |
lucasagomes | even for deployment, because we will destroy it via iscsi anyway | 13:58 |
lucasagomes | but anyway doesn't help much | 13:58 |
NobodyCam | morning lucasagomes dtantsur jroll rloo sambetts vdrok jlvillal devananda and anyone else not listed here | 13:59 |
* lucasagomes will check with dan prince what's the progress on the tripleo ci with IPA | 13:59 | |
lucasagomes | NobodyCam, morning | 13:59 |
NobodyCam | :) | 13:59 |
jroll | lucasagomes: yeah, more like security by obscurity there | 13:59 |
lucasagomes | jroll, totally | 13:59 |
jroll | lucasagomes: though we have a clean step for that downstream that we should totally upstream | 13:59 |
jroll | (for a variety of reasons) | 13:59 |
lucasagomes | jroll, to clean disk metadata? | 13:59 |
lucasagomes | cool yeah | 13:59 |
rloo | morning NobodyCam! | 14:00 |
jroll | lucasagomes: yeah, so basically as long as the agent ramdisk boots once, we nuke the bootloader so any further boots won't boot from disk | 14:00 |
jroll | (if pxe fails) | 14:00 |
*** kan_ has quit IRC | 14:00 | |
lucasagomes | jroll, yes! exactly that | 14:00 |
sambetts | jroll: ooo, I like that idea! | 14:00 |
lucasagomes | and this is pretty easy to implement, even for the bash ramdisk | 14:00 |
lucasagomes | 1 command | 14:00 |
lucasagomes | gdisk --zap should do it | 14:01 |
jroll | lucasagomes: https://github.com/rackerlabs/onmetal-ironic-hardware-manager/blob/master/onmetal_ironic_hardware_manager/__init__.py#L85 and https://github.com/rackerlabs/onmetal-ironic-hardware-manager/blob/master/onmetal_ironic_hardware_manager/__init__.py#L164 | 14:01 |
lucasagomes | jroll, right, that only clears MBR tho | 14:01 |
jroll | right, so we did it just for bootloader, but cleaning all metadata is a good idea | 14:01 |
NobodyCam | :) | 14:01 |
jroll | probably for every disk | 14:01 |
lucasagomes | jroll, you may want to clean up the GPT as well... For GPT there's a copy the partition table at the end of the disk | 14:01 |
jroll | that way you don't get raid superblocks messing with things etc | 14:01 |
lucasagomes | jroll, ++ | 14:02 |
lucasagomes | cool thing to do indeed | 14:02 |
vdrok | morning rloo and NobodyCam :) | 14:02 |
rloo | hi vdrok | 14:03 |
lucasagomes | jroll, http://www.rodsbooks.com/gdisk/sgdisk.html (see the --zap-all, we could use something like that, it's quick) | 14:03 |
*** ndipanov has quit IRC | 14:04 | |
thiagop | morning NobodyCam | 14:06 |
jlvillal | NobodyCam, Good morning | 14:06 |
jroll | lucasagomes: yeah +1 | 14:06 |
jlvillal | Good morning Ironic :) | 14:06 |
thiagop | morning rloo vdrok and whoever else have arrived after me | 14:06 |
dtantsur | morning jlvillal | 14:06 |
thiagop | morning jlvillal | 14:06 |
rloo | morning thiagop, jlvillal | 14:06 |
vdrok | morning thiagop jlvillal | 14:06 |
jlvillal | :) | 14:06 |
sambetts | o/ jlvillal thiagop | 14:06 |
jroll | morning jlvillal :) | 14:07 |
*** mehip has quit IRC | 14:07 | |
TheJulia | dtantsur: ever seen anything like http://logs.openstack.org/30/229930/17/check/gate-bifrost-integration-coreosipa-cirros/4f6a816/logs/ironic-conductor.log even though the module is (apparently) installed? | 14:07 |
*** mehip has joined #openstack-ironic | 14:08 | |
dtantsur | TheJulia, no never. if you definitely installed python-ironic-inspector-client, then I have no other ideas... | 14:09 |
dtantsur | this is CI machine again, right? | 14:09 |
TheJulia | dtantsur: yup, the ansible logs show it being installed | 14:09 |
jlvillal | jroll, Regarding https://bugs.launchpad.net/ironic/+bug/1503086 I updated bug with patch info | 14:09 |
openstack | Launchpad bug 1503086 in Ironic "Mis-formatted calls to log.* are not detected in unit tests" [Medium,In progress] - Assigned to John L. Villalovos (happycamp) | 14:10 |
TheJulia | Guess it is time to try a fresh local machine | 14:10 |
jroll | jlvillal: thank you | 14:10 |
*** r-daneel has joined #openstack-ironic | 14:10 | |
jroll | unreliable bug updater hook :( | 14:10 |
dtantsur | TheJulia, yeah... | 14:12 |
*** coolsvap has quit IRC | 14:12 | |
lucasagomes | TheJulia, really odd | 14:14 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-python-agent: Refactor the image download and checksum computation bits https://review.openstack.org/234338 | 14:14 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-python-agent: Add support for streaming raw images directly onto the disk https://review.openstack.org/234339 | 14:14 |
*** coolsvap has joined #openstack-ironic | 14:14 | |
jlvillal | jroll, No that was my fault. If I forget to put 'Closes-Bug' on the very first patch, it will not update the bug. | 14:14 |
jroll | jlvillal: O_o I don't think that's true, nor correct | 14:15 |
jroll | correct behavior* | 14:15 |
jlvillal | jroll, That is what they told me in infra | 14:15 |
jroll | x.x | 14:15 |
jroll | TIL | 14:15 |
jlvillal | So it doesn't spam the bug report. | 14:16 |
jroll | I swear that was not the case at one point in time | 14:16 |
rloo | jlvillal: that explains why i see missing patch info in bugs. | 14:17 |
sambetts | jlvillal, jroll: Its been like that for a little while, its quite frustating if you accidently forget one time | 14:17 |
* jlvillal often forgets to put Closes-Bug in his first revision | 14:17 | |
jroll | urgh :( | 14:17 |
dtantsur | yeah, just got hit by the same | 14:17 |
rloo | seems like spamming the bug report is better than it not 'working' as expected. | 14:18 |
*** puranamr has quit IRC | 14:21 | |
lucasagomes | ouch that bug thing really sucks | 14:28 |
*** mgoddard__ has joined #openstack-ironic | 14:29 | |
openstackgerrit | Stephanie Miller proposed openstack/bifrost: Migrate playbooks to use 'become' https://review.openstack.org/235388 | 14:32 |
*** mehip has quit IRC | 14:32 | |
*** mehip has joined #openstack-ironic | 14:32 | |
*** mgoddard_ has quit IRC | 14:33 | |
*** mehip has quit IRC | 14:34 | |
jroll | lucasagomes: do you know if we still want https://blueprints.launchpad.net/ironic/+spec/drac-hw-discovery | 14:37 |
jroll | ditto for https://blueprints.launchpad.net/ironic/+spec/drac-vendor-raid-mgmt | 14:37 |
jroll | ifarkas: ^ | 14:37 |
dtantsur | jroll, my guess is that we only want the latter | 14:37 |
jroll | that's my guess as well | 14:37 |
jroll | I see two abandoned specs for discovery :P | 14:37 |
dtantsur | we have ironic-inspector \o/ | 14:37 |
lucasagomes | jroll, not sure, but both still valid stuff | 14:38 |
jroll | \o/ | 14:38 |
jlvillal | Did we ever nail down options for locations for the mid-cycle? | 14:38 |
jroll | lucasagomes: what would need to be done for DRAC? | 14:38 |
lucasagomes | I mean doesn't harm to leave the blueprint as ideas, if someone wants to implement | 14:38 |
jroll | drac inspection class? | 14:38 |
sambetts | dtantsur: Your not going to do oob inspection? | 14:38 |
jroll | sure | 14:38 |
lucasagomes | jroll, for inspector? I think this is about OOB inspection | 14:38 |
jroll | jlvillal: I don't think we started talking about it :) | 14:38 |
*** garthb has joined #openstack-ironic | 14:38 | |
lucasagomes | jroll, so you don't need to boot a ramdisk on it | 14:38 |
jroll | lucasagomes: ok, I'll leave it | 14:38 |
lucasagomes | which != inspector | 14:38 |
jroll | yeah | 14:39 |
lucasagomes | iLO does it too, so it's aligned IMHO | 14:39 |
jroll | jlvillal: let's have deva host at IBM in minnesota, in february :) | 14:39 |
jlvillal | jroll, I thought we started... Maybe that was all. I remember talk of Dublin :) | 14:39 |
dtantsur | sambetts, me? I am definitely not :) I guess RH folks lost interest in OOB inspecton | 14:39 |
sambetts | :-P | 14:39 |
jlvillal | Eeek!! | 14:39 |
*** rameshg87 has joined #openstack-ironic | 14:39 | |
jroll | jlvillal: oh, we were just throwing ideas around, nothing solid | 14:39 |
* jroll would vote texas but N summit will be there already, so meh | 14:39 | |
jlvillal | Minnesota is nice in the summer. The only time I went there. | 14:40 |
jroll | ditto | 14:40 |
lucasagomes | Dublin !!! :-) | 14:40 |
dtantsur | weren't we going to have this midcycle around Europe? ;) | 14:40 |
lucasagomes | jk | 14:40 |
sambetts | dtantsur: I think OOB is something we hope to add to the Cisco drivers | 14:40 |
* lucasagomes would like to visit minnesota as well | 14:40 | |
sambetts | Europe would be awesome ;) | 14:40 |
dtantsur | sambetts, I think the spec is up for it even | 14:41 |
jroll | lucasagomes: ++, I'm a huge fan of the midwest in general | 14:41 |
* jlvillal suspects lucasgomes will not attend in Minessota... | 14:41 | |
sambetts | dtantsur: for one of them, yh :) | 14:41 |
* dtantsur does not want to spend winter holidays making a visa to the US.. | 14:41 | |
jroll | lucasagomes: you don't want to go to minnesota in february though :) | 14:41 |
lucasagomes | jroll, nice, I have never been to be honest. Would love to | 14:41 |
lucasagomes | jroll, too cold? | 14:41 |
* lucasagomes likes cold | 14:41 | |
ifarkas | jroll, dtantsur, lucasagomes, both drac-vendor-raid-mgmt and drac-hw-discovery are obsolete | 14:41 |
* lucasagomes way more than hot places | 14:41 | |
jroll | lucasagomes: think -15 C | 14:41 |
jlvillal | lucasagomes, It is another level of cold.... | 14:41 |
*** dims has quit IRC | 14:41 | |
jroll | ifarkas: oh, excellent. thanks! | 14:42 |
lucasagomes | heh I'm cool with that, we can go snowboard! | 14:42 |
dtantsur | ifarkas, yeah, and I think we should restart planning on OOB raid, right? | 14:42 |
jlvillal | Dublin is not cold compared to Minnesota | 14:42 |
dtantsur | now that we have an interface for that | 14:42 |
jroll | lucasagomes: minnesota is totally flat :D | 14:42 |
ifarkas | jroll, raid will use the raid interface instead of vendor_passthru | 14:42 |
jroll | OH | 14:42 |
lucasagomes | jlvillal, yeah Dublin's not cold at all, like real cold | 14:42 |
jroll | let's do colorado. | 14:42 |
*** dims has joined #openstack-ironic | 14:42 | |
ifarkas | dtantsur, yes | 14:42 |
* jroll begins figuring out who has offices in colorado | 14:42 | |
* jlvillal likes the idea of Dublin or some low cost Europe location. | 14:42 | |
jlvillal | But US is good too | 14:43 |
jroll | I'm open to europe if we can get lots of people there | 14:43 |
lucasagomes | ++ perhaps having a pool to see who would attend each | 14:43 |
lucasagomes | just to see the quorum | 14:43 |
* jroll wonders what JayF thinks of colorado | 14:43 | |
jroll | er. | 14:43 |
* jroll wonders what JayF thinks of europe | 14:43 | |
jlvillal | lucasagomes, Another vote, like for dinner? | 14:43 |
ifarkas | dtantsur, actually, that will my first thing to do after finishing up with dracclient | 14:43 |
lucasagomes | the good thing about doing somewhere else is you can get diff types of people, which is good | 14:43 |
jlvillal | lucasagomes, Is Dublin affordable? | 14:44 |
dtantsur | Brno is VERY affordable ;) | 14:44 |
lucasagomes | jlvillal, yeah another vote heh too much votings | 14:44 |
lucasagomes | jlvillal, it's quite pricey here, not crazy as london or paris | 14:44 |
lucasagomes | but compared to other parts of europe it can be expensive | 14:44 |
lucasagomes | ~5 euros for a pint of beer | 14:44 |
lucasagomes | (not sure if beer is a good parameter tho) | 14:45 |
dtantsur | now to something completely different. jroll, do you have an idea why the coreos ramdisk does not have /lib/modules/VER? | 14:45 |
jlvillal | I wonder about Gallway and Bristol for cost. | 14:45 |
jroll | dtantsur: no clue, I don't know what that is :) | 14:45 |
sambetts | dtantsur: subtle monty python reference ;) | 14:45 |
dtantsur | ;) | 14:45 |
lucasagomes | jlvillal, right, it's not crazy expensive. It's ok here | 14:45 |
jroll | dtantsur: you act like I know how linux works or something | 14:45 |
*** dims_ has joined #openstack-ironic | 14:45 | |
lucasagomes | but Galway yeah, it's cheaper | 14:45 |
dtantsur | jroll, lol, ok :) | 14:45 |
jroll | :) | 14:45 |
jlvillal | lucasagomes, Thanks. | 14:46 |
jroll | dtantsur: inside or outside the container? | 14:46 |
dtantsur | jroll, good question. inside, I guess | 14:46 |
jroll | dtantsur: hm, maybe that's why, dunno though | 14:46 |
openstackgerrit | Merged openstack/ironic: Add release notes for 4.2.1 https://review.openstack.org/235027 | 14:46 |
dtantsur | yeah, that's the possibility. then modprobe won't work for the coreos ramdisk (meaning inspection on bare metal probably won't work as well) | 14:47 |
lucasagomes | dtantsur, we could probably bind mount that | 14:48 |
lucasagomes | since we share the same kernel as the host | 14:48 |
dtantsur | into the container? dunno, maybe. I don't know coreos that well :) | 14:48 |
*** dims has quit IRC | 14:48 | |
lucasagomes | yeah into the container | 14:48 |
jroll | dtantsur: I think the way to do that is modprobe in coreos, maybe? | 14:49 |
jroll | dtantsur: we were also talking downstream about moving from container to chroot | 14:49 |
jroll | for a variety of reasons, mostly related to software raid | 14:49 |
lucasagomes | dtantsur, https://github.com/openstack/ironic-python-agent/blob/0144e79df5dab0837a55a981fa23298908af14cf/imagebuild/coreos/oem/cloud-config.yml#L59-L63 | 14:49 |
dtantsur | or to sambetts's tinyipa? | 14:49 |
jroll | aweeks: mariojv ^^ do you remember how we modprobe stuff for the agent container? | 14:50 |
dtantsur | lucasagomes, good catch man, I need to try it | 14:50 |
lucasagomes | cool :-) | 14:50 |
lucasagomes | you would need to do that in a chroot too | 14:50 |
jroll | dtantsur: if you're going to edit cloud-config, I'd just modprobe in there | 14:50 |
jroll | containers share the kernel with the host, so that should work out | 14:50 |
*** mgoddard__ has quit IRC | 14:51 | |
dtantsur | jroll, well, I'm not modprobing, rtslib does | 14:51 |
*** mgoddard has joined #openstack-ironic | 14:51 | |
jroll | urgh | 14:51 |
lucasagomes | x.x | 14:51 |
lucasagomes | *sighs* | 14:51 |
jroll | this makes me wonder how iscsi works at all in coreos image, actually | 14:51 |
sambetts | dtantsur: tinyipa has it ;) | 14:51 |
dtantsur | (I do modprobe in inspection, but it's easier to fix) | 14:51 |
jroll | ... or does it | 14:51 |
dtantsur | jroll, well, it does, judging by pxe_ipa gate ;) | 14:51 |
jroll | dtantsur: ikr, but I have no idea how, everyone has always said iscsi + container doesn't work | 14:52 |
lucasagomes | it does, well /dev is there | 14:52 |
lucasagomes | (bind mounted as well) | 14:52 |
jroll | yeah, maybe bind mounting /lib will just work, dunno | 14:52 |
lucasagomes | well... /lib/modules | 14:53 |
lucasagomes | cause only /lib will mess things around | 14:53 |
jroll | er, yeah | 14:53 |
* lucasagomes will add support for ironic in plan9 | 14:54 | |
* lucasagomes is jk | 14:54 | |
*** mehip has joined #openstack-ironic | 14:55 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [WIP] Support Linux-IO in addition to tgtd https://review.openstack.org/234861 | 14:55 |
jroll | y'all good with me approving this? https://blueprints.launchpad.net/ironic/+spec/multi-nodes-operations | 14:55 |
* dtantsur never thought about it | 14:55 | |
jroll | not sure it needs a spec | 14:55 |
dtantsur | jroll, I can't imagine how multi operation will work for get-*. also not sure why this commands, but not e.g. node-update | 14:56 |
*** mehip has quit IRC | 14:56 | |
dtantsur | jroll, that said, I definitely like the idea | 14:56 |
dtantsur | but I would probably want something written on it.. | 14:56 |
jroll | dtantsur: yeah, I'll ask haomeng for more info first | 14:57 |
Madasi | I know nova refused to add the functionality to api, insisted it was done client side instead | 14:57 |
Madasi | but not sure that means we want to be like nova | 14:57 |
jroll | Madasi: yeah, this is just client | 14:57 |
*** hblixt has quit IRC | 14:57 | |
jroll | dtantsur: maybe just a list of commands and an example of each? | 14:57 |
dtantsur | ++ for doing it in client | 14:57 |
jroll | in the BP | 14:57 |
*** mehip has joined #openstack-ironic | 14:57 | |
Madasi | jroll: ok, cool. Looked but couldn't tell from the description | 14:57 |
dtantsur | jroll, yes, something really simple, just so that we all can see what we agree on :) | 14:57 |
jroll | yep | 14:57 |
lucasagomes | jroll, that's only for the client right? | 14:58 |
lucasagomes | delete is not multi-node in the API | 14:58 |
jroll | lucasagomes: correct | 14:58 |
lucasagomes | jroll, yeah I'm good with that :-) | 14:58 |
Madasi | +1 for doing it, and for doing it client side | 14:58 |
*** mehip has quit IRC | 14:59 | |
jroll | alright, I asked for more details before approving | 14:59 |
* jroll might add a specless blueprint etherpad or meeting section or something | 14:59 | |
dtantsur | jroll, +1 to both etherpad and meeting section | 14:59 |
* jroll also wonders if that blueprint should be in the client launchpad instead :/ | 14:59 | |
jroll | dtantsur: alright, I was going to write an email about spec-less blueprints today, I'll add that | 15:00 |
dtantsur | jroll, if we have a place to track reviews/priorities, we can add specless bp's there | 15:00 |
jroll | what's a priority? | 15:00 |
jroll | :) | 15:00 |
jroll | yeah, working on that stuff before summit | 15:01 |
dtantsur | a thing people are constantly asking about, assuming that we have it :D | 15:01 |
*** Nisha has quit IRC | 15:01 | |
*** mehip has joined #openstack-ironic | 15:01 | |
*** Nisha has joined #openstack-ironic | 15:01 | |
jroll | trying to clean up blueprints first | 15:01 |
jroll | holy cow so many blueprints | 15:01 |
*** mkovacik has quit IRC | 15:02 | |
dtantsur | many blueprints | 15:02 |
dtantsur | so features | 15:02 |
dtantsur | wow code | 15:02 |
lucasagomes | heh we have the spreadsheet | 15:02 |
lucasagomes | but not sure if it's the perfect place tho | 15:03 |
*** amotoki has quit IRC | 15:03 | |
* dtantsur has a crazy idea to use gerrit for it | 15:03 | |
dtantsur | jroll, we can have an rst in our specs repo with a list of specless blueprints :) | 15:03 |
* dtantsur is not sure whether he's kidding or not | 15:04 | |
jroll | dtantsur: oh lord, code review for a list of things to review | 15:04 |
rameshg87 | devananda: hi | 15:04 |
sambetts | or just a rst with the text from the BP in it ? | 15:04 |
dtantsur | jroll, we need to go deeper! | 15:05 |
*** Nisha has quit IRC | 15:05 | |
*** Nisha has joined #openstack-ironic | 15:05 | |
dtantsur | first we need a spec about specless blueprints | 15:05 |
* jroll ragequits | 15:05 | |
* sambetts faints | 15:05 | |
* dtantsur hides | 15:05 | |
*** caiobo has quit IRC | 15:06 | |
lucasagomes | heh gosh | 15:06 |
sambetts | I think a specless BP could be represented in our specs repo by a rst with just the title and a link to the BP ? | 15:07 |
dtantsur | ... and the we don't need blueprints at all. bye-bye launchpad | 15:07 |
sambetts | at least then they can still be in the backlog or approved | 15:07 |
* lucasagomes renames ironic to vogon | 15:07 | |
* dtantsur googles | 15:07 | |
jroll | sambetts: meh | 15:08 |
lucasagomes | heh yeah google for vogons | 15:08 |
jroll | lol | 15:08 |
lucasagomes | they bureaucrats | 15:08 |
lucasagomes | the* | 15:08 |
sambetts | lol :-p all the references today | 15:08 |
sambetts | the LP BPs would still have the whiteboard tracking for the patches | 15:09 |
* dtantsur didn't recognize the English spelling, apparently | 15:10 | |
sambetts | related to it, but with an rst in the specs repo the nice index page that was being reviewed last week will actually include them | 15:10 |
mariojv | good morning ironic | 15:11 |
jroll | I'm not strictly opposed to it, but I don't think that's necessary | 15:11 |
jroll | would need reviewer time for a copy/paste :/ | 15:11 |
mariojv | jroll: just saw your message from yesterday - i thought we didn't modprobe anything, just a bunch of bind mounts | 15:11 |
jroll | morning mariojv | 15:11 |
jroll | mariojv: yeah, I'm uninformed, that answers my question :P | 15:12 |
mariojv | cool | 15:12 |
jroll | "yesterday" was like, an hour ago btw | 15:12 |
jroll | :P | 15:12 |
lucasagomes | heh | 15:14 |
lucasagomes | mariojv, morning | 15:14 |
sambetts | jroll: title and link is all I think needs to be included, the full description could stay in the BP, it could even be automated quite easily like once after a certain period something takes all the approved specless BPs and syncs the specs repo with the files, then you'd just have 1 review per period we decide to approve, similar to the requirements sync ones | 15:14 |
mariojv | Ah, i'm bad at reading utc times | 15:14 |
mariojv | :P | 15:14 |
jroll | sambetts: meh, still a waste of time IMO :) | 15:14 |
*** thiagop is now known as thiagop-afk | 15:14 | |
jroll | if someone wants to see what features are inbound, they can look at launchpad | 15:15 |
rloo | sambetts, jroll: I'm against adding spec-less information to the specs repo, unless we rename the specs repo to features repo. | 15:15 |
jroll | +1 | 15:15 |
dtantsur | morning mariojv! so, new way to access iSCSI seems to require modprobing. so now I have a lot of fun :) | 15:15 |
jroll | dtantsur: should we move this BP to inspector BPs or? https://blueprints.launchpad.net/ironic/+spec/access-port-discovery | 15:16 |
jroll | (wow, much old) | 15:16 |
mariojv | dtantsur: maybe modprobing in the base system through some sort of config and then bind mounting some dir would work? | 15:16 |
dtantsur | jroll, at first glance, yes | 15:16 |
sambetts | rloo: thats a good point, I'm just brain vommiting any way, btw how are you? | 15:16 |
rloo | wrt midcycles, i liked deva's idea about having the midcyle located close to where the summit is. but having said that, it didn't work the last time and i suspect asia won't work | 15:17 |
rloo | sambetts: trying not to vomit from all your suggestions :) | 15:17 |
openstackgerrit | Merged openstack/ironic: Add support for in-band cleaning in ISCSIDeploy https://review.openstack.org/220898 | 15:17 |
sambetts | rloo: haha | 15:17 |
rloo | sambetts: otherwise fine. although i feel like i need a break. maybe after the summit. | 15:17 |
dtantsur | mariojv, yeah, maybe.. not sure how do it with the coreos ramdisk though. I'd really want us to be able to run modprobe from IPA itself, in case plugins want to do it | 15:17 |
dtantsur | rloo, ++ for after summit break, my pto was just approved | 15:18 |
rloo | dtantsur: nice! | 15:18 |
* sambetts wonders why IPA needs to run in a container | 15:18 | |
dtantsur | rloo, asia won't work? we can try Moscow :D | 15:18 |
dtantsur | (bad idea, don't listen to me) | 15:18 |
jroll | sambetts: because easy mode | 15:18 |
jroll | (it doesn't need to, that's just what we did for ease of build) | 15:18 |
*** cemason has joined #openstack-ironic | 15:18 | |
rloo | dtantsur: yeah, i knew you were joking. and if you weren't...!! | 15:19 |
jroll | sambetts: like I said before, we've talked about doing a chroot instead | 15:19 |
*** mehip has quit IRC | 15:20 | |
sambetts | ah, makes sense | 15:20 |
*** mehip has joined #openstack-ironic | 15:20 | |
rameshg87 | hello all, a couple of things related to documentation | 15:21 |
dtantsur | o/ | 15:21 |
rameshg87 | wanted to check with devananda but looks like he is not around | 15:21 |
rameshg87 | https://review.openstack.org/#/c/219603/ | 15:21 |
rameshg87 | any thoughts on documentation fixes for stable branches ? | 15:21 |
jroll | lintan: left a question here, this seems like a dupe https://blueprints.launchpad.net/ironic/+spec/add-clean-tasks-field-to-node | 15:22 |
*** puranamr has joined #openstack-ironic | 15:22 | |
rameshg87 | accepted/not accepted/suggestions ? | 15:22 |
jroll | rameshg87: I think that's probably a question for #openstack-stable, but I think in general docs patches are not backported | 15:22 |
rameshg87 | jroll: yeah, but it's not a backport either. | 15:23 |
rameshg87 | jroll: it's a fix that is applicable only in that stable branch | 15:23 |
jroll | rameshg87: I mean, applied to stable branches | 15:23 |
*** cemason has quit IRC | 15:23 | |
* rameshg87 moves to openstack-stable | 15:23 | |
jroll | rameshg87: beyond that, I tend to think deployers should know to set up https if they want secure communication, but that's just me :) | 15:23 |
* jroll watches | 15:23 | |
*** e0ne has joined #openstack-ironic | 15:25 | |
rameshg87 | jroll: yeah, but we just want to be sure .. | 15:25 |
jroll | I suppose | 15:25 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 15:26 |
rameshg87 | jroll: and one more thing. we talked about this before. ability to submit driver documentation without using 2x+2 rules and | 15:26 |
rameshg87 | jroll: do you think it's a good idea to pursue ? | 15:26 |
jroll | rameshg87: that sounds like something we should talk about on the mailing list, I think that's what I said last time | 15:27 |
*** puranamr has quit IRC | 15:27 | |
rameshg87 | jroll: okay, I missed that part. probably I will shoot a mail, others can pitch in. thanks, will do. | 15:27 |
jroll | cool, thanks :) | 15:27 |
*** openstackgerrit has quit IRC | 15:31 | |
*** openstackgerrit has joined #openstack-ironic | 15:32 | |
dtantsur | going a bit earlier today, see you | 15:32 |
sambetts | o/ dtantsur | 15:32 |
jroll | later dtantsur \o | 15:33 |
lucasagomes | dtantsur, night | 15:33 |
dtantsur | sambetts, if you stay for some more time, could you push #openstack-qa folks for reviewing https://review.openstack.org/#/c/235294/ ? | 15:33 |
sambetts | Yup I'll see what I can do :) | 15:33 |
*** dtantsur is now known as dtantsur|afk | 15:34 | |
jroll | going to step out for a while, bbl | 15:34 |
*** dprince has joined #openstack-ironic | 15:44 | |
*** mgoddard_ has joined #openstack-ironic | 15:45 | |
*** linggao has joined #openstack-ironic | 15:47 | |
*** mgoddard has quit IRC | 15:48 | |
*** spandhe has joined #openstack-ironic | 15:49 | |
NobodyCam | night dtantsur|afk | 15:53 |
*** spandhe has quit IRC | 15:54 | |
* rameshg87 is done for the day | 15:54 | |
rameshg87 | good night folks | 15:54 |
*** rameshg87 has quit IRC | 15:54 | |
*** hblixt has joined #openstack-ironic | 15:54 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic: Enable radosgw support in ironic https://review.openstack.org/229389 | 15:58 |
*** mgoddard_ has quit IRC | 15:58 | |
*** mgoddard has joined #openstack-ironic | 15:58 | |
*** lazy_prince has quit IRC | 15:59 | |
*** thrash is now known as thrash|f00dz | 16:00 | |
openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Fix a missing detail in install guide. https://review.openstack.org/235463 | 16:00 |
*** trown is now known as trown|lunch | 16:02 | |
*** puranamr has joined #openstack-ironic | 16:02 | |
* sambetts is also heading off | 16:03 | |
sambetts | o/ night everyone | 16:03 |
NobodyCam | night sambetts | 16:03 |
*** ndipanov has joined #openstack-ironic | 16:03 | |
lucasagomes | folks I'm calling it a day too | 16:09 |
lucasagomes | have a great night everyone | 16:09 |
lucasagomes | talk soon :-) | 16:09 |
*** puranamr has quit IRC | 16:10 | |
*** puranamr has joined #openstack-ironic | 16:11 | |
* lucasagomes goes catch the train | 16:11 | |
NobodyCam | night lucasagomes | 16:11 |
*** derekh has quit IRC | 16:12 | |
*** lucasagomes has quit IRC | 16:16 | |
*** puranamr has quit IRC | 16:18 | |
*** puranamr has joined #openstack-ironic | 16:19 | |
*** yog_ has quit IRC | 16:20 | |
*** puranamr has quit IRC | 16:21 | |
*** mehip has quit IRC | 16:21 | |
*** penick has joined #openstack-ironic | 16:23 | |
*** Nisha has quit IRC | 16:24 | |
*** puranamr has joined #openstack-ironic | 16:25 | |
*** Nisha has joined #openstack-ironic | 16:25 | |
*** ifarkas has quit IRC | 16:26 | |
*** puranamr has quit IRC | 16:27 | |
*** Nisha has quit IRC | 16:28 | |
*** Nisha has joined #openstack-ironic | 16:28 | |
*** mbound has quit IRC | 16:28 | |
*** puranamr has joined #openstack-ironic | 16:28 | |
*** mehip has joined #openstack-ironic | 16:29 | |
*** e0ne has quit IRC | 16:30 | |
baoli | jroll: ping | 16:30 |
*** jistr has quit IRC | 16:33 | |
*** jcoufal has quit IRC | 16:34 | |
jroll | baoli: hi | 16:34 |
devananda | o/ | 16:37 |
NobodyCam | morning devananda | 16:37 |
jroll | devananda: morning, feeling any better? | 16:38 |
baoli | jroll: I have a quick question about the ironic-python-agent | 16:38 |
jroll | baoli: "don't ask to ask, just ask" :) | 16:38 |
NobodyCam | :p | 16:39 |
baoli | jroll: I saw a module netutils.py that you wrote as part of it. Is it being used? | 16:39 |
jroll | baoli: it's being used downstream, and I suspect it will be used for inspection things soon | 16:40 |
jroll | why do you ask? | 16:40 |
baoli | jroll: I saw that it's using lldp to discover the switchport info, and that's what I'm interested in | 16:41 |
jroll | baoli: oh, did you just volunteer to write the lldp inspection module? :) | 16:41 |
baoli | jroll: not really | 16:41 |
jroll | heh | 16:41 |
jroll | but yes, that lldp stuff is being used downstream a bit | 16:42 |
baoli | jroll: I was trying to find out its current status | 16:42 |
jroll | example: https://github.com/rackerlabs/onmetal-ironic-hardware-manager/blob/master/onmetal_ironic_hardware_manager/__init__.py#L425 | 16:42 |
baoli | jroll: what do you mean by being used downstream? | 16:42 |
jroll | oh, it totally works, rackspace has the code I just linked in production | 16:42 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 16:44 |
jroll | baoli: so yes, that code works. no, it isn't being used for inspection yet | 16:45 |
openstackgerrit | Ruby Loo proposed openstack/ironic-lib: Set a default for 'rootwrap_helper_cmd' https://review.openstack.org/235489 | 16:45 |
jroll | baoli: does that answer your question? | 16:45 |
baoli | jroll: yes, thanks! | 16:45 |
*** Sukhdev has joined #openstack-ironic | 16:46 | |
*** lazy_prince has joined #openstack-ironic | 16:46 | |
jroll | you're welcome :) | 16:46 |
baoli | jroll: I also saw reference to it from the network flip and ironic-network integration BP specs. | 16:46 |
jroll | baoli: yes, so those features rely on switchport info in the db; I'm hoping someone writes an inspector module to get lldp info and put it in the db | 16:47 |
baoli | jroll: so someone already volunteered to write the lldp inspection module | 16:47 |
jroll | baoli: not sure :) | 16:47 |
devananda | jroll: a bit, yea. thanks. still going to take it easy today | 16:47 |
jroll | \o/ | 16:48 |
*** puranamr has quit IRC | 16:50 | |
baoli | jroll: I read a couple of wikis on ironic-python-agent. What are the commands that the user can issue now? | 16:53 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add note about driver API breakage https://review.openstack.org/235026 | 16:53 |
jroll | rloo: NobodyCam ^ does that resolve your concerns? | 16:53 |
*** puranamr has joined #openstack-ironic | 16:53 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic-python-agent: Support proxies for image download https://review.openstack.org/235492 | 16:54 |
openstackgerrit | Vladyslav Drok proposed openstack/ironic: Add support for downloading images via proxy in agent driver https://review.openstack.org/235493 | 16:54 |
jroll | baoli: so the user doesn't typically interact with IPA directly; ironic does all of the interaction. some examples of commands are "write image to disk" or "run this clean step (e.g. erase disks)" | 16:54 |
*** Marga_ has joined #openstack-ironic | 16:54 | |
jroll | baoli: have you seen the docs? they may help you make sense of the code http://docs.openstack.org/developer/ironic-python-agent | 16:55 |
jroll | devananda: you'll also be interested in 235026 if you're up for relnotes reviews | 16:55 |
*** puranamr has quit IRC | 16:55 | |
*** subscope has quit IRC | 16:55 | |
*** Marga_ has quit IRC | 16:56 | |
*** Marga_ has joined #openstack-ironic | 16:56 | |
baoli | jroll: I browsed over it, and that's the question I had after it. | 16:57 |
jroll | aha | 16:57 |
baoli | jroll, so ironic needs to be configured to work with the agent, I guess | 16:57 |
jroll | oh, I guess that doesn't tell you much about commands, eh | 16:57 |
jroll | baoli: correct, though it actually works out of the box with every driver | 16:57 |
jroll | baoli: here's how to set it up: http://docs.openstack.org/developer/ironic/deploy/install-guide.html#building-or-downloading-a-deploy-ramdisk-image | 16:58 |
vdrok | could anyone please take a look at this spec, if you have a minute - https://review.openstack.org/#/c/209612/, it's been sitting without reviews for quite some time :) | 16:59 |
devananda | jroll: ack. queueing up now | 16:59 |
jroll | sweet, thanks | 16:59 |
*** mkovacik has joined #openstack-ironic | 16:59 | |
*** praneshp has joined #openstack-ironic | 17:00 | |
baoli | jroll: ok, I think I was confused with the title CoreOS tools, but it actually has instructions on both ubuntu and fedora | 17:00 |
jroll | baoli: yes, there are various builders. | 17:01 |
*** garthb_ has joined #openstack-ironic | 17:01 | |
baoli | jroll: thanks! | 17:01 |
jroll | baoli: coreos was the original builder (and is just a debian container in a coreos system), but people seem stuck on disk-image-builder for various reasons :) | 17:02 |
*** praneshp_ has joined #openstack-ironic | 17:02 | |
vdrok | good night everyone | 17:02 |
*** puranamr has joined #openstack-ironic | 17:02 | |
*** garthb has quit IRC | 17:03 | |
jroll | #success the final liberty release was just announced and includes ironic 4.2.0 \o/ | 17:03 |
*** trown|lunch is now known as trown | 17:03 | |
openstackstatus | jroll: Added success to Success page | 17:03 |
*** mehip has quit IRC | 17:04 | |
*** praneshp has quit IRC | 17:04 | |
*** puranamr has quit IRC | 17:04 | |
*** praneshp_ is now known as praneshp | 17:04 | |
*** puranamr has joined #openstack-ironic | 17:05 | |
*** Nisha has quit IRC | 17:05 | |
*** thrash|f00dz is now known as thrash | 17:06 | |
*** amotoki has joined #openstack-ironic | 17:06 | |
*** Nisha has joined #openstack-ironic | 17:06 | |
*** kozhukalov_ has quit IRC | 17:07 | |
*** afaranha has joined #openstack-ironic | 17:08 | |
*** albertoffb has joined #openstack-ironic | 17:09 | |
afaranha | Hello guys. I'm trying to make a deploy using the agent, but the images that I download http://tarballs.openstack.org/ironic-python-agent/ironic-python-agent-stable-liberty.tar.gz , comes with coreos_production_pxe.vmlinuz and a coreos_production_pxe_image-oem.cpio , is this cpio a ramdisk image? Can I just use it normally as a ramdisk? | 17:10 |
*** puranamr has quit IRC | 17:10 | |
*** spandhe has joined #openstack-ironic | 17:11 | |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add documentation for Ceph Object Gateway support https://review.openstack.org/232068 | 17:11 |
jroll | afaranha: yes | 17:11 |
*** mehip has joined #openstack-ironic | 17:11 | |
afaranha | jroll, thanks | 17:12 |
*** harshs has joined #openstack-ironic | 17:13 | |
jroll | np | 17:13 |
*** bitblt has joined #openstack-ironic | 17:13 | |
*** Marga_ has quit IRC | 17:14 | |
*** sambetts has quit IRC | 17:15 | |
*** Marga_ has joined #openstack-ironic | 17:15 | |
*** sambetts has joined #openstack-ironic | 17:17 | |
*** phschwartz is now known as phschwartz_aw | 17:17 | |
*** achanda has joined #openstack-ironic | 17:18 | |
*** Sukhdev has quit IRC | 17:19 | |
*** kozhukalov_ has joined #openstack-ironic | 17:21 | |
*** jobewan has joined #openstack-ironic | 17:25 | |
NobodyCam | night vdrok | 17:33 |
*** Marga_ has quit IRC | 17:33 | |
*** mehip has quit IRC | 17:37 | |
*** ionutbalutoiu has quit IRC | 17:39 | |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add note about driver API breakage https://review.openstack.org/235026 | 17:40 |
jroll | you guys and your awesome reviews | 17:40 |
*** mehip has joined #openstack-ironic | 17:41 | |
*** lekha is now known as lekha||away | 17:41 | |
*** romainh has quit IRC | 17:44 | |
NobodyCam | jroll: reads much better now :) | 17:47 |
jroll | \o/ | 17:47 |
*** mkovacik has quit IRC | 17:48 | |
*** pelix has quit IRC | 17:48 | |
*** lucasagomes__ has joined #openstack-ironic | 17:48 | |
*** degorenko is now known as _degorenko | 17:48 | |
rloo | jroll: can't believe you buckled under the 'double space after dot'! | 17:49 |
jroll | rloo: hahaha, I can't stand two spaces after a period | 17:50 |
*** mkovacik has joined #openstack-ironic | 17:50 | |
rloo | jroll: oh, you're one of *those* people :) | 17:50 |
*** mkovacik has quit IRC | 17:50 | |
* jroll wonders if this should have a version bump :/ https://review.openstack.org/#/c/215505/21/ironic/objects/node.py | 17:50 | |
NobodyCam | +2 | 17:51 |
jroll | NobodyCam: on version bump or relnotes update? :P | 17:51 |
jroll | oh wait, I'm an idiot | 17:52 |
* jroll scrolls up and sees version change | 17:52 | |
*** mehip has quit IRC | 17:52 | |
jroll | er | 17:52 |
jroll | ugh, objects version bump, not api version bump | 17:52 |
* jroll HRM | 17:52 | |
*** mehip has joined #openstack-ironic | 17:53 | |
jroll | dansmith: hate to bug you... should objects versions only be bumped for things that need to be munged for compatibility, or all the things? looking at https://review.openstack.org/#/c/215505/21/ironic/objects/node.py | 17:55 |
jroll | I feel like that doesn't need a bump, curious what nova rules on this is | 17:55 |
rloo | jroll: i would agree that it doesn't need a bump, unless our code works with non-int values | 17:56 |
jlvillal | dansmith: Why is this message in Spam? It has a from address in danplanet.com but has failed danplanet.com's required tests for authentication. | 17:56 |
rloo | jroll: if our code works with non-int values, then i dunno, hypothetically, someone may still want it to work with non-int values, and set the version accordingly. | 17:56 |
jlvillal | dansmith: Maybe you already know that Gmail treats your emails like spam when they go through openstack-dev | 17:56 |
jroll | jlvillal: sounds like dkim fail :( | 17:57 |
dansmith | jroll: anything remotable, so that means @remotable methods or anything changed in fields | 17:57 |
dansmith | jroll: with the exception of some changes where you're making a change, but it definitely doesn't affect the wire | 17:57 |
jroll | dansmith: so validating a value is an int seems like it doesn't need a bump | 17:57 |
jroll | right, the wire is the main thing yeah? | 17:57 |
jlvillal | dansmith: I should have quoted: "Why is this message in Spam? It has a from address in danplanet.com but has failed danplanet.com's required tests for authentication." | 17:58 |
dansmith | jroll: depends on what you're validating and how that might affect a caller that used to be able to provide some value there or something | 17:58 |
jroll | dansmith: https://review.openstack.org/#/c/215505/21/ironic/objects/node.py | 17:58 |
jroll | my initial thought is no | 17:58 |
dansmith | jlvillal: I have DMARC +q enabled on my domain, and mailman breaks that | 17:58 |
jroll | no, being don't bump it | 17:58 |
dansmith | jroll: will look in a sec | 17:59 |
jroll | cool, thanks | 17:59 |
* jroll adds a -1 for other things | 17:59 | |
dansmith | jlvillal: it's a change I made in the last 30 days and I'm thinking I probably will need to drop it | 17:59 |
jlvillal | dansmith: I did update my filter in Gmail to say, never send to Spam for openstack-dev email | 18:00 |
jlvillal | But others might not do that :( | 18:00 |
*** david-lyle has quit IRC | 18:01 | |
*** david-lyle has joined #openstack-ironic | 18:02 | |
*** amotoki has quit IRC | 18:03 | |
*** Sukhdev has joined #openstack-ironic | 18:07 | |
*** e0ne has joined #openstack-ironic | 18:09 | |
dansmith | jlvillal: I know | 18:11 |
*** lucasagomes has joined #openstack-ironic | 18:15 | |
*** lucasagomes is now known as lucas-afk | 18:16 | |
*** bitblt has quit IRC | 18:16 | |
*** Nisha has quit IRC | 18:18 | |
*** lucas-afk has quit IRC | 18:20 | |
jlvillal | dansmith: jroll: https://www.youtube.com/watch?v=_mlwE_Ha7Ws&list=TLBwJAMrqJwKcxNTEwMjAxNQ | 18:20 |
jlvillal | OpenStack video on how to get to hotel from Narita | 18:21 |
dansmith | jlvillal: I saw, that's good stuff | 18:21 |
jlvillal | :) | 18:21 |
*** garthb_ has quit IRC | 18:21 | |
*** mkovacik has joined #openstack-ironic | 18:27 | |
*** lucasagomes has joined #openstack-ironic | 18:28 | |
*** lucasagomes__ has quit IRC | 18:28 | |
*** lucasagomes is now known as lucas-afk | 18:28 | |
lucas-afk | sorry testing the IRC bouncer | 18:28 |
jlvillal | NobodyCam: lucas-afk dtantsur|afk and anybody else might like the video on how to get to Summit from airports: https://www.youtube.com/watch?v=_mlwE_Ha7Ws&index=1&list=TLBwJAMrqJwKcxNTEwMjAxNQ | 18:32 |
lucas-afk | jlvillal, oh that's useful! | 18:32 |
jlvillal | Has different videos for Narita and Haneda. | 18:32 |
jlvillal | rloo: ^^^ | 18:32 |
* lucas-afk checks which one he will land | 18:32 | |
rloo | jlvillal: thx. | 18:32 |
thiagop-afk | jlvillal: that's very useful, I'm spreading for the folks here who will go | 18:32 |
thiagop-afk | thanks | 18:33 |
jlvillal | thiagop-afk: Me too. Sending it out internally :) | 18:33 |
*** thiagop-afk is now known as thiagop | 18:33 | |
* lucas-afk will arrive at haneda | 18:33 | |
jlvillal | lucas-afk: There is a play list. 2nd one is Haneda | 18:34 |
rloo | jlvillal etc, there are also downloadable instructions: https://www.openstack.org/assets/tokyo-summit/directions/OSTokyo-Narita-Directions-PROOF-v2.pdf | 18:34 |
rloo | jlvillal: not sure why no one sent out email about them. | 18:34 |
lucas-afk | jlvillal, yup just found it, thanks much! It's really useful | 18:34 |
*** cemason has joined #openstack-ironic | 18:34 | |
jlvillal | rloo: If they did, I missed it | 18:34 |
rloo | jlvillal: i've got some other links too (that i haven't looked at yet) | 18:35 |
rloo | jlvillal: i don't think i saw any email except to women-of-openstack | 18:35 |
*** lsmola_ has quit IRC | 18:35 | |
*** marzif has joined #openstack-ironic | 18:35 | |
*** achanda has quit IRC | 18:36 | |
jlvillal | rloo: Thanks for the PDF! | 18:36 |
*** marzif has quit IRC | 18:36 | |
*** marzif has joined #openstack-ironic | 18:36 | |
rloo | jlvillal: i forwarded the email to you | 18:37 |
jlvillal | rloo: Thank you :) | 18:37 |
*** cemason has quit IRC | 18:38 | |
jlvillal | rloo: Got it | 18:39 |
jlvillal | https://www.openstack.org/summit/tokyo-2015/tokyo-and-travel/ | 18:39 |
rloo | jlvillal: ah | 18:40 |
jlvillal | rloo: I think the video and directions are relatively new. | 18:40 |
jlvillal | rloo: Video was published two days ago | 18:40 |
rloo | jlvillal: i find it amusing that we're going somewhere that requires a video etc. | 18:40 |
jlvillal | :) | 18:40 |
*** dims has joined #openstack-ironic | 18:43 | |
* lucas-afk gone... | 18:44 | |
*** achanda has joined #openstack-ironic | 18:44 | |
*** dims_ has quit IRC | 18:45 | |
*** achanda has quit IRC | 18:46 | |
NobodyCam | night lucas-afk | 18:47 |
*** dims has quit IRC | 18:47 | |
*** garthb_ has joined #openstack-ironic | 18:51 | |
*** ijw has joined #openstack-ironic | 19:02 | |
devananda | TheJulia: question on bifrost - when building the configdrive, it can inject an ipv4 address -- if using the dynamic-inventory module, where does it get that address from? | 19:05 |
devananda | i've been trying to trace that and have not figured it out yet :( | 19:06 |
*** kozhukalov_ has quit IRC | 19:06 | |
*** Lydia_K has quit IRC | 19:07 | |
openstackgerrit | Ruby Loo proposed openstack/ironic: Move pxe._parse_instance_info() to deploy_utils https://review.openstack.org/229274 | 19:08 |
*** dims has joined #openstack-ironic | 19:11 | |
*** mehip has quit IRC | 19:11 | |
*** Lydia_K has joined #openstack-ironic | 19:12 | |
*** mehip has joined #openstack-ironic | 19:13 | |
TheJulia | devananda: ipv4_address as part of a variable defined under the node it's self | 19:14 |
TheJulia | devananda: I think I have examples | 19:15 |
devananda | TheJulia: node.properties? node.instance_info? | 19:15 |
TheJulia | https://github.com/openstack/bifrost/blob/bef47c3e21e9c4fe707d0c2f660d18a64518cb68/playbooks/inventory/bifrost_inventory.py#L84 | 19:16 |
*** mgoddard_ has joined #openstack-ironic | 19:16 | |
TheJulia | I could have sworn I had ipmi/ssh examples defined in the tree | 19:16 |
devananda | TheJulia: yea - that's for static inventory from a csv/json file. what about dynamic? | 19:17 |
*** mgoddard__ has joined #openstack-ironic | 19:17 | |
TheJulia | https://github.com/openstack/bifrost/blob/bef47c3e21e9c4fe707d0c2f660d18a64518cb68/playbooks/inventory/baremetal.json.example | 19:17 |
devananda | there is no top-level "ipv4_address" resource in ironic | 19:17 |
TheJulia | which dynamic inventory data source? | 19:18 |
devananda | ironic | 19:18 |
TheJulia | we have no way to support that out of the box | 19:18 |
TheJulia | because we have nothing in ironic's data model defining ip addresses | 19:18 |
devananda | node.instance_info['ipv4_address'] ?? | 19:18 |
*** mgoddard has quit IRC | 19:19 | |
*** kozhukalov_ has joined #openstack-ironic | 19:19 | |
TheJulia | devananda: if it is completely free-form, I guess that would be easy | 19:19 |
devananda | might be a terrible idea ... | 19:19 |
TheJulia | or you could join data together as a step in a playbook | 19:19 |
TheJulia | and make it a fact | 19:19 |
devananda | and yes, instance_info is, in fact, freeform | 19:19 |
dansmith | jroll: sorry just getting back to this.. those values it's checking aren't fields on the object | 19:20 |
devananda | TheJulia: anyhow, thanks. that explains why I couldn't find the code path for it -- it's not there :) | 19:20 |
*** mgoddard_ has quit IRC | 19:20 | |
dansmith | jroll: oh, they're in the dict of properties ... | 19:20 |
TheJulia | devananda: heh, yeah, the playbooks just expect it to be a host level variable | 19:21 |
TheJulia | devananda: horrible idea... run the inventory script, dump the list, inject addresses, save as file, and point at the file as the inventory source | 19:22 |
TheJulia | devananda: absoloutely horrible idea | 19:22 |
devananda | TheJulia: any idea why I'm getting "One or more undefined variables: 'ipv4_address' is undefined" error then, when using INVENTORY_SOURCE=ironic and ansible-playbook -vvvv -i inventory/bifrost_inventory.py deploy-dynamic.yaml ? | 19:22 |
devananda | yea, i dont want to resort to what you suggested ... because this was working yesterday | 19:22 |
devananda | and today (on a different host) I'm getting this error | 19:23 |
*** phschwartz_aw is now known as phschwartz | 19:23 | |
devananda | like its decided not to use dhcp configs now .. | 19:23 |
TheJulia | I think some network config stuff landed on master in the last few days, going to look | 19:23 |
*** mgoddard_ has joined #openstack-ironic | 19:24 | |
*** mgoddard__ has quit IRC | 19:25 | |
*** e0ne has quit IRC | 19:25 | |
*** mgoddard__ has joined #openstack-ironic | 19:25 | |
*** lazy_prince has quit IRC | 19:28 | |
*** mgoddard_ has quit IRC | 19:29 | |
TheJulia | devananda: ugh, yeah, it was a change that landed yesterday. https://review.openstack.org/#/c/216372/ Filing a bug, you'll also want to revert dependent revs | 19:29 |
*** dims has quit IRC | 19:30 | |
*** mehip1 has joined #openstack-ironic | 19:30 | |
*** mehip has quit IRC | 19:32 | |
TheJulia | devananda: I'm in the middle of dumping brain into another file right now, I'll start on fixing that bug, 15066606 in a hour or two | 19:32 |
*** achanda has joined #openstack-ironic | 19:32 | |
devananda | TheJulia: oooh. thanks. | 19:38 |
*** mgoddard_ has joined #openstack-ironic | 19:43 | |
*** alexpilotti has joined #openstack-ironic | 19:43 | |
*** mgoddard__ has quit IRC | 19:44 | |
*** mgoddard has joined #openstack-ironic | 19:45 | |
*** Sukhdev has quit IRC | 19:47 | |
*** mgoddard_ has quit IRC | 19:48 | |
*** marzif has quit IRC | 19:50 | |
*** Marga_ has joined #openstack-ironic | 19:52 | |
*** mestery_ has joined #openstack-ironic | 19:53 | |
*** mgoddard_ has joined #openstack-ironic | 19:53 | |
*** mgoddard has quit IRC | 19:54 | |
*** mgoddard__ has joined #openstack-ironic | 19:54 | |
*** mgoddard_ has quit IRC | 19:56 | |
*** mgoddard_ has joined #openstack-ironic | 19:56 | |
*** mgoddard__ has quit IRC | 19:57 | |
*** mgoddard__ has joined #openstack-ironic | 19:57 | |
*** ijw has quit IRC | 19:58 | |
*** achanda has quit IRC | 19:58 | |
*** albertoffb has quit IRC | 20:00 | |
*** mgoddard_ has quit IRC | 20:01 | |
*** Marga_ has quit IRC | 20:02 | |
*** Marga_ has joined #openstack-ironic | 20:03 | |
jroll | dansmith: yeah, it's lame | 20:04 |
jroll | TheJulia: devananda: instance_info gets nuked on teardown fwiw | 20:04 |
dansmith | jroll: well, basically you should really only use dicts like this for unformatted strings (i.e. metadata) | 20:05 |
devananda | jroll: yah. and IPV4 should get nuked then too | 20:05 |
jroll | dansmith: don't get me started on our dict fields :) | 20:05 |
dansmith | jroll: so whether you can really do this and whether you need to bump the version is probably a judgment call based on what you know about the layers above | 20:05 |
dansmith | jroll: heh, okay :D | 20:05 |
jroll | dansmith: working on nuking those this cycle | 20:05 |
jroll | cool, that's a good answer | 20:05 |
jroll | or rather, the answer I expected | 20:06 |
*** Marga_ has quit IRC | 20:08 | |
mrda | Morning Ironic | 20:13 |
TheJulia | jroll: yeah, I'm not exactly liking the idea of using instance_info for storage of such details, although, getting nuked at teardown seems fine to me, but I may also be crazy | 20:14 |
thiagop | morning mrda | 20:14 |
mrda | o/ | 20:14 |
jroll | TheJulia: yeah, I don't know the details of what you're doing, just wanted to point that out | 20:16 |
TheJulia | jroll: taking over the world? | 20:17 |
jroll | TheJulia: I mean besides that, are you taking over the world persistently, etc :D | 20:17 |
*** boris-42 has quit IRC | 20:18 | |
jlvillal | TheJulia: Are you thinking what I'm thinking? | 20:18 |
*** mestery_ has quit IRC | 20:19 | |
* jlvillal wondering if this is Pinky & The Brain reference time ;) | 20:19 | |
*** achanda has joined #openstack-ironic | 20:19 | |
TheJulia | jlvillal: I think so, but where will we find enough storm trooper armor sets to make it happen? | 20:20 |
jlvillal | :D | 20:20 |
Madasi | jlvillal: http://media.giphy.com/media/12UWMsiE9gTuCI/giphy.gif | 20:21 |
openstackgerrit | Merged openstack/ironic: Fix a missing detail in install guide. https://review.openstack.org/235463 | 20:21 |
jlvillal | heh | 20:21 |
jroll | Madasi: that reminds me, one of my priorities for this cycle is to have more gifs in this channel | 20:22 |
Madasi | jroll: I'll add it to my goals for the sprint | 20:23 |
jroll | +1 ty | 20:23 |
*** sambetts has quit IRC | 20:24 | |
jroll | Madasi: I'll have JayF get it on your performance review goals as well | 20:24 |
Madasi | I mean, I had a 10 minute conversation with the ops team in all gifs today. | 20:26 |
*** sambetts has joined #openstack-ironic | 20:26 | |
*** achanda has quit IRC | 20:28 | |
*** mehip1 has quit IRC | 20:30 | |
*** mehip has joined #openstack-ironic | 20:31 | |
*** mestery_ has joined #openstack-ironic | 20:32 | |
openstackgerrit | Lilia Sampaio proposed openstack/ironic: Add OneView driver documentation https://review.openstack.org/234396 | 20:32 |
*** mestery_ has quit IRC | 20:32 | |
*** baoli has quit IRC | 20:32 | |
*** Guest71541 has joined #openstack-ironic | 20:32 | |
*** lifeless has quit IRC | 20:34 | |
*** thrash is now known as thrash|bbl | 20:42 | |
*** Sukhdev has joined #openstack-ironic | 20:46 | |
*** Guest71541 is now known as dims_ | 20:46 | |
*** hblixt_ has joined #openstack-ironic | 20:48 | |
*** hblixt has quit IRC | 20:52 | |
*** mehip has joined #openstack-ironic | 20:56 | |
*** achanda has joined #openstack-ironic | 20:56 | |
*** cemason has joined #openstack-ironic | 21:01 | |
*** cemason has quit IRC | 21:05 | |
*** harshs has quit IRC | 21:06 | |
Sukhdev | jroll: Ping | 21:09 |
Sukhdev | devananda: ping | 21:09 |
openstackgerrit | Merged openstack/ironic: Add note about driver API breakage https://review.openstack.org/235026 | 21:10 |
*** trown is now known as trown|outttypeww | 21:13 | |
*** lucas-afk has quit IRC | 21:14 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for installation of ironic-inspector https://review.openstack.org/229930 | 21:27 |
*** achanda has quit IRC | 21:32 | |
*** Marga_ has joined #openstack-ironic | 21:33 | |
*** mehip has quit IRC | 21:37 | |
*** _hblixt_ has joined #openstack-ironic | 21:40 | |
thiagop | jroll: that thing earlier: bug filed https://bugs.launchpad.net/ironic/+bug/1506636 | 21:41 |
openstack | Launchpad bug 1506636 in Ironic "node-validate fails for boot and deploy on agent_* drivers" [Undecided,New] | 21:41 |
*** hblixt_ has quit IRC | 21:43 | |
*** dprince has quit IRC | 21:45 | |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial addition of an inspection module https://review.openstack.org/232745 | 21:49 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: WIP: Additional inspection role and adding role to NV job script https://review.openstack.org/232746 | 21:49 |
*** cemason has joined #openstack-ironic | 21:52 | |
*** Sukhdev has quit IRC | 21:57 | |
jroll | thiagop: nice, ty | 21:57 |
openstackgerrit | Merged openstack/ironic: Add agent_amt docs https://review.openstack.org/235293 | 21:59 |
*** priteau has quit IRC | 22:06 | |
jroll | woo, agent_amt driver complete | 22:08 |
* jroll looks at onevbiew docs again | 22:08 | |
*** Marga_ has quit IRC | 22:10 | |
*** achanda has joined #openstack-ironic | 22:12 | |
*** harshs has joined #openstack-ironic | 22:12 | |
*** achanda has quit IRC | 22:13 | |
*** achanda has joined #openstack-ironic | 22:14 | |
*** jamielennox|away is now known as jamielennox | 22:16 | |
*** meghal has joined #openstack-ironic | 22:19 | |
meghal | jroll: filed a bug here - https://bugs.launchpad.net/ironic/+bug/1506657 | 22:19 |
openstack | Launchpad bug 1506657 in Ironic "Ironic conductor Hash ring reset needs to be independent of sync_local_state periodic task" [Undecided,New] | 22:19 |
jroll | meghal: cool, thanks | 22:19 |
jroll | I was thinking about this more earlier | 22:19 |
jroll | we should totally have an rpc call that hits every conductor and tells them to reset the ring | 22:20 |
meghal | what would be the criteria for the trigger of rpc call ? | 22:20 |
openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/235434 | 22:20 |
jroll | meghal: add/delete node, conductor joins/leaves the ring | 22:21 |
thiagop | jroll: we would totally appreciate an input on how to format these CLI commands on the docs... | 22:22 |
*** penick has quit IRC | 22:23 | |
jroll | thiagop: I got totally distracted heh | 22:23 |
jroll | gimme 2 minutes | 22:23 |
*** Sukhdev has joined #openstack-ironic | 22:23 | |
thiagop | jroll: no hurry | 22:23 |
jroll | Sukhdev: pong | 22:23 |
meghal | jroll: ok…I am not able to understand how a particular conductor leaving the ring is being detected right now ? maybe I asked this yesterday…but am still confused | 22:23 |
thiagop | It's about 20hs here. I'm getting home very soon.. | 22:23 |
jroll | thiagop: I'll leave a note on the patch | 22:24 |
thiagop | jroll: ty | 22:24 |
Sukhdev | jroll: Hey Jim - have a question for you - | 22:24 |
meghal | jroll: for example conductor 2 needs to know conductor 1 left and reset the ring | 22:24 |
jroll | meghal: it's only picked up in the periodic task | 22:24 |
jroll | meghal: so I'm thinking, in addition, on conductor shutdown it could spam a reset message to the others | 22:24 |
meghal | jroll: is this the sync_local_state or something else ? | 22:24 |
jroll | meghal: yeah | 22:24 |
jroll | Sukhdev: no need to ping me or tell me you have a question, you can just ask immediately next time | 22:25 |
jroll | meghal: sync_local_state forces a reset every time, so it will calculate based on what conductors are detected as active | 22:25 |
Sukhdev | jroll: OK | 22:26 |
Sukhdev | jroll: So, I am able to launch a VM(BM) on provisioning network - it is able to get the DHCP address correctly - but, it is timing out on tftp | 22:27 |
meghal | jroll: right…that is how it is done right now…I would suggest the mapping call to check when was last reset run and trigger if it is more than configured interval | 22:27 |
meghal | jroll: because you would still need some poll to figure out that conductors left unless as you said conductor is able to send a leave signal to others | 22:27 |
meghal | how would that work if the host just dies | 22:27 |
Sukhdev | jroll: if I go from the hypervisor (where the VM/BM is launched), I can reach the tftp server, but not from the VM - | 22:28 |
Sukhdev | jroll: any idea? or any tips? | 22:28 |
jroll | meghal: there's a periodic task to heartbeat to the DB, and dbapi methods use that to determine what's active. or something like that. | 22:28 |
jroll | meghal: I also like the 'check age' idea, but there's no ring age stored right now | 22:29 |
jroll | meghal: we should totally write something up here, these are just some thoughts I've had on it | 22:29 |
jroll | Sukhdev: tcpdump to find out where it's dropping? check iptables? | 22:30 |
meghal | jroll: I can update the bug with more ideas | 22:30 |
jroll | Sukhdev: honestly the best thing to do would be to write up instructions on how to test this so that other people (like me) can try to reproduce this and help out | 22:30 |
meghal | jroll: how are bugs handled during a release cycle ? | 22:30 |
*** _hblixt_ has quit IRC | 22:30 | |
jroll | meghal: in terms of what? | 22:31 |
jroll | meghal: we're always fixing bugs, I can tell you that much :P | 22:31 |
Sukhdev | jroll: OK - let me make some more discovery | 22:31 |
jroll | Sukhdev: really, instructions would be the most valuable thing here | 22:32 |
devananda | Sukhdev: ohhai | 22:32 |
* devananda reads backscroll | 22:32 | |
Sukhdev | jroll: understood - I will try to post something on the etherpad for others to see the steps | 22:33 |
jroll | devananda: same problem as the other day iirc | 22:33 |
devananda | jroll: heh, re: rpc call to initiate a hash ring reset -- I started work on that a year ago. | 22:33 |
jroll | devananda: well wtf :) | 22:33 |
devananda | jroll: I think I even proposed it ... :) | 22:33 |
jroll | devananda: generally, the hash ring thing is https://bugs.launchpad.net/ironic/+bug/1506657 | 22:34 |
openstack | Launchpad bug 1506657 in Ironic "Ironic conductor Hash ring reset needs to be independent of sync_local_state periodic task" [Undecided,New] | 22:34 |
jroll | rpc call solves this for normal shutdown | 22:34 |
devananda | meghal: the conductor heartbeat record in DB -- all conductors read that table and can deterine when any other conductor has not updated it | 22:35 |
devananda | meghal: this could lead to problems if you have master<->master replicated database, the replication fails, and conductors are split between them | 22:35 |
jroll | devananda: it's also an issue if you turn off the periodic task, which we document as supported O_o | 22:35 |
devananda | uh what? | 22:36 |
jroll | yeah | 22:36 |
devananda | turning off that periodic task is not supported | 22:36 |
devananda | no way | 22:36 |
devananda | when did that get documented as an OK thing to do? | 22:36 |
jroll | https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L158 | 22:36 |
jroll | meghal: so going back a bit - in current master, is there any reason to turn this off? just performance things? | 22:37 |
devananda | jroll: right. so if there is no local state to sync (eg, static DHCP, homogeneous hardware, etc) then yea, it's fine. | 22:38 |
meghal | jroll: I explained in the bug, we are not interested in takeover and so no need to run the periodic task | 22:38 |
meghal | which does db queries | 22:38 |
jroll | devananda: except then the hash ring never gets updated :) | 22:38 |
jroll | meghal: right, cool | 22:38 |
devananda | jroll: the API nodes will route requests as necessary, and they detect ring changes on the fly | 22:38 |
jroll | devananda: how about sync_power_state :) | 22:38 |
devananda | jroll: the hash ring as it is cached on the conductor -- which is meaningless if its not syncing local state | 22:38 |
devananda | oh | 22:38 |
devananda | well | 22:38 |
jroll | heh | 22:39 |
jroll | when you add a node, it'll never get put in the ring | 22:39 |
devananda | ok, that's a problem | 22:39 |
jroll | obligatory https://github.com/openstack/ironic/blob/master/ironic/conductor/rpcapi.py#L99 | 22:39 |
devananda | the conductor's ring isn't getting reset from any other periodic task | 22:39 |
jroll | :D | 22:39 |
*** Sukhdev has quit IRC | 22:39 | |
meghal | devananda: I do want the hash ring to be updated…but do not want it tied with sync local state | 22:39 |
devananda | so yea, if that periodic task is disabled -- I see the problem | 22:39 |
devananda | meghal: agreed | 22:39 |
devananda | jroll: so that's a different bug :) | 22:40 |
jroll | devananda: heh, I know, just sayin' | 22:40 |
jroll | I set the bug to critical | 22:40 |
* jroll adds backport potential for extra sadness | 22:40 | |
devananda | jroll: the API tier has recreates the hash ring for each request. or it used to, but I think there was some work done on caching it better | 22:40 |
jroll | devananda: yeah, api is still fine I believe | 22:41 |
devananda | in a very large cluster, it may be slow. but it should always be accurate | 22:41 |
meghal | so what options exist right now to detect join/leave of conductors ? periodic poll only…right ? | 22:42 |
jroll | meghal: right now, yes. or restart all the conductors :| | 22:42 |
meghal | jroll: so that bottoms down to resetting the hash ring periodically | 22:43 |
meghal | now we need to figure out where | 22:43 |
jroll | meghal: yeah, I totally agree | 22:43 |
jroll | I mean... maybe a different periodic task. or maybe an option to turn off the takeover parts of that one | 22:43 |
meghal | sync local state and power sync are the ones calling the mapping call where hash ring needs to return relatively fresh state | 22:44 |
jroll | meghal: maybe negative value to that config becomes "don't do takeover but still update the ring" | 22:44 |
jroll | or maybe refresh based on age | 22:44 |
meghal | yeah because mapping call is specifically for keyring…or do this in the hash ring get call itself | 22:44 |
meghal | where it is checked if hash ring is none | 22:44 |
meghal | inside the hash ring implementation | 22:45 |
jroll | yep | 22:45 |
*** Sukhdev has joined #openstack-ironic | 22:45 | |
meghal | jroll: https://github.com/openstack/ironic/blob/master/ironic/common/hash_ring.py#L177 | 22:45 |
jroll | if self.ring is None or self.age > 600: | 22:45 |
jroll | something like that | 22:45 |
meghal | yep | 22:45 |
jroll | that may be the best fix | 22:45 |
jroll | probably make the age configurable | 22:45 |
meghal | jroll: yeah | 22:46 |
meghal | and sync_local_state would be strictly about doing takeovers | 22:46 |
jroll | +1 | 22:46 |
jroll | meghal: what's the chance you want to write this code? | 22:46 |
jroll | :) | 22:47 |
jroll | we can take it on if you don't have time or whatever | 22:47 |
meghal | jroll: sure…I would be interested in it…would just be wondering on the timelines ? by what date do we need to get it in ? | 22:47 |
devananda | jroll: https://review.openstack.org/#/c/79194/ | 22:47 |
jroll | meghal: whenever is good | 22:48 |
jroll | meghal: I'd like to do a release shortly after summit; I'd probably also backport this | 22:48 |
jroll | meghal: left a game plan on the bug as a comment, btw | 22:50 |
jroll | I need to step away for a bit, brb | 22:50 |
meghal | jroll: sounds good…so I can begin with the implementation based on those details…right ? | 22:51 |
devananda | meghal: that review ^ was an early stab I took at this a year ago | 22:51 |
meghal | jroll: will try to get something done by first week of November | 22:51 |
meghal | devananda: going through it | 22:51 |
*** thrash|bbl is now known as thrash|g0ne | 22:52 | |
jroll | devananda: offhand, do you know if adding a config will make backporting this problematic? | 22:53 |
meghal | devananda: for the rpc cast call, how would it work if each conductor in the cluster kept on sending these messages ? | 22:55 |
*** Sukhdev has quit IRC | 22:58 | |
meghal | ok went through the comments…as you have mentioned in the last comment the initiation logic for cast was missing in the spec | 22:58 |
*** kozhukalov_ has quit IRC | 22:59 | |
devananda | jroll: yes - backporting a new CONF is not allowed | 22:59 |
jroll | urgh, hm | 23:00 |
devananda | meghal: "kept on sending" -- that might be bad. sending them how often? | 23:00 |
meghal | devananda: but then each conductor would send it to each other…right ? | 23:00 |
devananda | meghal: yah. IIRC, I had a problem getting the conductor to send messages to each other | 23:00 |
devananda | right | 23:00 |
devananda | they need to create a client connection to the RPC bus -- which they do not do today | 23:01 |
jroll | devananda: maybe we choose a sane default and ship that hardcoded in the backport. I don't love that, though... | 23:01 |
meghal | devananda: in that case it would be better that ironic-api initiated it and asked all to refresh | 23:01 |
jroll | devananda: they do these days! cleaning does it. | 23:01 |
devananda | jroll: nice! | 23:01 |
devananda | jroll: oh ha, that. | 23:01 |
thiagop | folks, calling it a day | 23:01 |
*** Sukhdev has joined #openstack-ironic | 23:01 | |
thiagop | see you tomorrow | 23:01 |
devananda | g'night, thiagop | 23:01 |
jroll | later thiagop | 23:01 |
*** thiagop has quit IRC | 23:02 | |
devananda | jroll: you know how I do not love the rpc callback thing that cleaning does ... | 23:02 |
devananda | jroll: where is that created? | 23:02 |
devananda | i dont see any rpcapi in conductorManager | 23:03 |
devananda | oh. new layer. ironic/common/rpc | 23:03 |
*** r-daneel has quit IRC | 23:03 | |
devananda | nope ... conductor only calls "rpc.get_notifier" | 23:04 |
jroll | devananda: I don't love it either | 23:04 |
jroll | ummm | 23:04 |
jroll | it totally works. somehow. | 23:04 |
*** jobewan has quit IRC | 23:04 | |
devananda | jroll: got it. agent_base_vendor.py pulls in rpcapi | 23:05 |
jroll | ah, yes | 23:06 |
meghal | devananda: jroll need to go…will try to send out draft implementation based on what jroll put in the bug and we can take it from there | 23:06 |
jroll | meghal: cool, sounds good. thanks for the help here :) | 23:06 |
devananda | meghal: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/agent_base_vendor.py#L218-L222 | 23:07 |
devananda | something like that might work | 23:07 |
jroll | an rpc call doesn't solve the 'conductor crashed' problem, though | 23:07 |
devananda | jroll: true | 23:07 |
devananda | periodic task has latency and can be turned off | 23:07 |
devananda | jroll: yet another case where a real DLM might help | 23:08 |
jroll | devananda: that wouldn't help us remember to update the hash ring :P | 23:08 |
jroll | nor help us not have local state in the conductor | 23:09 |
*** david-lyle has quit IRC | 23:09 | |
devananda | jroll: to get a notification that a service has stopped | 23:09 |
*** david-lyle has joined #openstack-ironic | 23:09 | |
jroll | oh, right | 23:09 |
devananda | so each conductor initiates the appropriate action fromthat event | 23:10 |
meghal | jroll: np | 23:10 |
devananda | meghal: thanks & good night :) | 23:10 |
meghal | devananda: jroll I guess this issues are similar to service group drivers in nova | 23:10 |
jroll | meghal: yeah | 23:10 |
meghal | need to start thinking how events related to nodes leaving/joining clusters can be pushed to interested entities | 23:11 |
openstackgerrit | John L. Villalovos proposed openstack/ironic: WIP: Initial func work to keep safe https://review.openstack.org/235612 | 23:15 |
meghal | see you later | 23:15 |
meghal | good night | 23:15 |
*** meghal has left #openstack-ironic | 23:15 | |
*** vilobhmm11 has joined #openstack-ironic | 23:16 | |
*** chlong has quit IRC | 23:16 | |
*** wshao has joined #openstack-ironic | 23:17 | |
*** hblixt has joined #openstack-ironic | 23:20 | |
*** BadCub_ has joined #openstack-ironic | 23:27 | |
*** wendar_ has joined #openstack-ironic | 23:29 | |
*** wendar has quit IRC | 23:30 | |
*** persia has quit IRC | 23:30 | |
*** persia has joined #openstack-ironic | 23:30 | |
*** achanda_ has joined #openstack-ironic | 23:30 | |
*** chenke_ has joined #openstack-ironic | 23:30 | |
*** linggao_ has joined #openstack-ironic | 23:31 | |
*** SpamapS has quit IRC | 23:31 | |
*** Marga_ has joined #openstack-ironic | 23:31 | |
*** mkovacik_ has joined #openstack-ironic | 23:31 | |
*** dansmith has quit IRC | 23:31 | |
*** dansmith has joined #openstack-ironic | 23:32 | |
*** achanda has quit IRC | 23:32 | |
*** BadCub has quit IRC | 23:32 | |
*** BadCub_ is now known as BadCub | 23:32 | |
*** mkovacik has quit IRC | 23:34 | |
*** chenke has quit IRC | 23:34 | |
*** linggao has quit IRC | 23:34 | |
*** sbadia has quit IRC | 23:34 | |
*** bapalm has quit IRC | 23:34 | |
*** mtreinish has quit IRC | 23:34 | |
*** SpamapS has joined #openstack-ironic | 23:35 | |
*** mtreinish has joined #openstack-ironic | 23:39 | |
*** bapalm has joined #openstack-ironic | 23:39 | |
*** sbadia has joined #openstack-ironic | 23:42 | |
*** alexpilotti has quit IRC | 23:54 | |
*** afaranha has quit IRC | 23:54 | |
*** coolsvap has quit IRC | 23:54 | |
*** phschwartz has quit IRC | 23:54 | |
*** wshao has quit IRC | 23:54 | |
*** sambetts has quit IRC | 23:54 | |
*** bradjones has quit IRC | 23:54 | |
*** JoshNang has quit IRC | 23:54 | |
*** lintan has quit IRC | 23:54 | |
*** betherly has quit IRC | 23:54 | |
*** arif-ali has quit IRC | 23:54 | |
*** johnthetubaguy has quit IRC | 23:54 | |
*** gmmaha has quit IRC | 23:54 | |
*** SpamapS has quit IRC | 23:54 | |
*** wendar_ has quit IRC | 23:54 | |
*** vilobhmm11 has quit IRC | 23:54 | |
*** dims_ has quit IRC | 23:54 | |
*** davidlenwell has quit IRC | 23:54 | |
*** jamielennox has quit IRC | 23:54 | |
*** greghaynes has quit IRC | 23:54 | |
*** mitz_ has quit IRC | 23:54 | |
*** xek has quit IRC | 23:54 | |
*** vishwanathj has quit IRC | 23:54 | |
*** romcheg has quit IRC | 23:54 | |
*** cppforlife_ has quit IRC | 23:54 | |
*** igordcard_ has quit IRC | 23:54 | |
*** zigo has quit IRC | 23:54 | |
*** tonyb has quit IRC | 23:54 | |
*** krotscheck has quit IRC | 23:54 | |
*** eglute has quit IRC | 23:54 | |
*** jroll has quit IRC | 23:54 | |
*** natorious has quit IRC | 23:54 | |
*** comstud has quit IRC | 23:54 | |
*** mgagne has quit IRC | 23:54 | |
*** TheJulia has quit IRC | 23:54 | |
*** dtantsur|afk has quit IRC | 23:54 | |
*** gabriel has quit IRC | 23:54 | |
*** mjturek1 has quit IRC | 23:54 | |
*** vdrok has quit IRC | 23:54 | |
*** mikal has quit IRC | 23:54 | |
*** crinkle has quit IRC | 23:54 | |
*** trown|outttypeww has quit IRC | 23:54 | |
*** jrist has quit IRC | 23:54 | |
*** sinval has quit IRC | 23:54 | |
*** mariojv has quit IRC | 23:54 | |
*** marios has quit IRC | 23:54 | |
*** lekha||away has quit IRC | 23:54 | |
*** shakamunyi has quit IRC | 23:54 | |
*** sirushti has quit IRC | 23:54 | |
*** zsmithnyc has quit IRC | 23:54 | |
*** Marga_ has quit IRC | 23:54 | |
*** persia has quit IRC | 23:54 | |
*** adam_g has quit IRC | 23:54 | |
*** v92 has quit IRC | 23:54 | |
*** zer0c00l has quit IRC | 23:54 | |
*** kbyrne has quit IRC | 23:54 | |
*** toabctl has quit IRC | 23:54 | |
*** maurosr has quit IRC | 23:54 | |
*** bapalm has quit IRC | 23:54 | |
*** sbadia has quit IRC | 23:54 | |
*** MattMan has quit IRC | 23:54 | |
*** devananda has quit IRC | 23:54 | |
*** _degorenko has quit IRC | 23:54 | |
*** tristanC has quit IRC | 23:55 | |
*** jlvillal has quit IRC | 23:55 | |
*** anteaya has quit IRC | 23:55 | |
*** david-lyle has quit IRC | 23:55 | |
*** harshs has quit IRC | 23:55 | |
*** mgoddard__ has quit IRC | 23:55 | |
*** ndipanov has quit IRC | 23:55 | |
*** kozhukalov has quit IRC | 23:55 | |
*** jaypipes has quit IRC | 23:55 | |
*** Madasi has quit IRC | 23:55 | |
*** verdurin has quit IRC | 23:55 | |
*** keekz has quit IRC | 23:55 | |
*** coreycb has quit IRC | 23:55 | |
*** balajin_ has quit IRC | 23:55 | |
*** davideag_ has quit IRC | 23:55 | |
*** logan2 has quit IRC | 23:55 | |
*** morgan has quit IRC | 23:55 | |
*** yuanying has quit IRC | 23:55 | |
*** liliars has quit IRC | 23:55 | |
*** rcarrillocruz has quit IRC | 23:55 | |
*** agordeev has quit IRC | 23:55 | |
*** huats has quit IRC | 23:55 | |
*** achanda_ has quit IRC | 23:55 | |
*** hblixt has quit IRC | 23:55 | |
*** Sukhdev has quit IRC | 23:55 | |
*** openstackgerrit has quit IRC | 23:55 | |
*** amit213 has quit IRC | 23:55 | |
*** aweeks has quit IRC | 23:55 | |
*** dhellmann has quit IRC | 23:55 | |
*** JayF has quit IRC | 23:55 | |
*** NobodyCam has quit IRC | 23:55 | |
*** EmilienM has quit IRC | 23:55 | |
*** mtreinish has quit IRC | 23:55 | |
*** chenke_ has quit IRC | 23:55 | |
*** shadower has quit IRC | 23:55 | |
*** daemontool_ has quit IRC | 23:55 | |
*** yuriyz has quit IRC | 23:55 | |
*** HenryG has quit IRC | 23:55 | |
*** d0ugal has quit IRC | 23:55 | |
*** pleia2 has quit IRC | 23:55 | |
*** slagle has quit IRC | 23:55 | |
*** linggao_ has quit IRC | 23:55 | |
*** praneshp has quit IRC | 23:55 | |
*** getvasanth__ has quit IRC | 23:55 | |
*** UForgotten_ has quit IRC | 23:55 | |
*** bigjools has quit IRC | 23:55 | |
*** alex_xu has quit IRC | 23:55 | |
*** dasp has quit IRC | 23:55 | |
*** BadCub has quit IRC | 23:55 | |
*** Lydia_K has quit IRC | 23:55 | |
*** bodgix has quit IRC | 23:55 | |
*** kragniz has quit IRC | 23:55 | |
*** bnemec has quit IRC | 23:55 | |
*** Ng has quit IRC | 23:55 | |
*** sc68cal has quit IRC | 23:55 | |
*** _fortis has quit IRC | 23:55 | |
*** mrda has quit IRC | 23:55 | |
*** aignatov has quit IRC | 23:55 | |
*** clif_h has quit IRC | 23:55 | |
*** dansmith has quit IRC | 23:55 | |
*** cemason has quit IRC | 23:55 | |
*** notq has quit IRC | 23:55 | |
*** jasondotstar has quit IRC | 23:55 | |
*** gridinv has quit IRC | 23:55 | |
*** lynxman has quit IRC | 23:55 | |
*** morgabra has quit IRC | 23:55 | |
*** sinh has quit IRC | 23:55 | |
*** cinerama has quit IRC | 23:55 | |
*** goodygum has quit IRC | 23:55 | |
*** mkovacik_ has quit IRC | 23:55 | |
*** mdbooth has quit IRC | 23:55 | |
*** harlowja has quit IRC | 23:55 | |
*** mordred has quit IRC | 23:55 | |
*** krtaylor has quit IRC | 23:55 | |
*** purp has quit IRC | 23:55 | |
*** ekarlso has quit IRC | 23:55 | |
*** thrash|g0ne has quit IRC | 23:55 | |
*** kevinbenton has quit IRC | 23:55 | |
*** intr1nsic has quit IRC | 23:55 | |
*** EntropyWorks has quit IRC | 23:55 | |
*** raginbajin has quit IRC | 23:55 | |
*** early has quit IRC | 23:55 | |
*** smoriya has quit IRC | 23:55 | |
*** mmnelemane has quit IRC | 23:55 | |
*** Sukhdev has joined #openstack-ironic | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!