| *** naohirot has joined #openstack-ironic | 00:04 | |
| *** romcheg has quit IRC | 00:05 | |
| *** baoli has joined #openstack-ironic | 00:11 | |
| *** davideagnello has quit IRC | 00:12 | |
| *** davideagnello has joined #openstack-ironic | 00:12 | |
| *** baoli has quit IRC | 00:16 | |
| *** shadower has quit IRC | 00:23 | |
| *** shadower has joined #openstack-ironic | 00:23 | |
| *** baoli has joined #openstack-ironic | 00:29 | |
| *** baoli has joined #openstack-ironic | 00:33 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Fix iRMC vmedia deploy failure due to already attached image https://review.openstack.org/225521 | 00:37 |
|---|---|---|
| *** chlong has joined #openstack-ironic | 00:42 | |
| *** dimsum__ has joined #openstack-ironic | 00:45 | |
| *** dimsum__ has quit IRC | 00:47 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor IRMCVirtualMediaAgentDeploy by applying new BootInterface https://review.openstack.org/221577 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add portgroups to support LAG interfaces - RPC https://review.openstack.org/206243 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add portgroups to support LAG interfaces - net https://review.openstack.org/206245 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add portgroups to support LAG interfaces - DB https://review.openstack.org/206232 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add portgroups to support LAG interfaces - objs https://review.openstack.org/206238 | 00:49 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Add network provider interface and implementations https://review.openstack.org/139687 | 00:50 |
| openstackgerrit | William Stevenson proposed openstack/ironic: Update the deploy drivers with network flipping logic https://review.openstack.org/213262 | 00:50 |
| *** achanda has quit IRC | 00:55 | |
| *** davideagnello has quit IRC | 00:58 | |
| *** derekh has quit IRC | 01:14 | |
| *** baoli has quit IRC | 01:17 | |
| *** Sukhdev has joined #openstack-ironic | 01:32 | |
| *** garthb has joined #openstack-ironic | 01:34 | |
| *** Sukhdev has quit IRC | 01:37 | |
| *** dimsum__ has joined #openstack-ironic | 01:45 | |
| *** baoli has joined #openstack-ironic | 01:49 | |
| *** dimsum__ has quit IRC | 02:00 | |
| *** chenglch has joined #openstack-ironic | 02:01 | |
| *** chenglch has quit IRC | 02:12 | |
| *** Haomeng has joined #openstack-ironic | 02:17 | |
| *** chenglch has joined #openstack-ironic | 02:22 | |
| *** baoli has quit IRC | 02:28 | |
| *** Sukhdev has joined #openstack-ironic | 02:38 | |
| *** davideagnello has joined #openstack-ironic | 02:59 | |
| *** dimsum__ has joined #openstack-ironic | 03:01 | |
| *** naohirot has quit IRC | 03:01 | |
| *** davideagnello has quit IRC | 03:04 | |
| *** dimsum__ has quit IRC | 03:08 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Make end-points discoverable via Ironic API https://review.openstack.org/205895 | 03:09 |
| *** praneshp_ has joined #openstack-ironic | 03:12 | |
| *** praneshp has quit IRC | 03:14 | |
| *** praneshp_ is now known as praneshp | 03:14 | |
| *** garthb has quit IRC | 03:44 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Config drive support for whole disk images for iscsi based deploy https://review.openstack.org/225115 | 03:46 |
| *** baoli has joined #openstack-ironic | 03:59 | |
| *** naohirot has joined #openstack-ironic | 03:59 | |
| *** baoli has quit IRC | 04:03 | |
| *** davideagnello has joined #openstack-ironic | 04:12 | |
| *** Nisha has joined #openstack-ironic | 04:25 | |
| *** VikasC has joined #openstack-ironic | 04:27 | |
| *** achanda has joined #openstack-ironic | 04:34 | |
| *** garthb has joined #openstack-ironic | 04:54 | |
| *** davideagnello has quit IRC | 05:05 | |
| *** dimsum__ has joined #openstack-ironic | 05:05 | |
| *** dimsum__ has quit IRC | 05:11 | |
| *** Nisha_away has joined #openstack-ironic | 05:26 | |
| *** Nisha has quit IRC | 05:30 | |
| *** chenglch has quit IRC | 05:30 | |
| *** garthb has quit IRC | 05:33 | |
| *** rameshg87 has joined #openstack-ironic | 05:46 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add support for in-band cleaning in ISCSIDeploy https://review.openstack.org/220898 | 05:46 |
| *** praneshp has quit IRC | 05:48 | |
| *** praneshp has joined #openstack-ironic | 05:48 | |
| *** e0ne has joined #openstack-ironic | 06:10 | |
| *** lsmola has joined #openstack-ironic | 06:13 | |
| *** e0ne has quit IRC | 06:14 | |
| *** Marga_ has quit IRC | 06:18 | |
| *** Marga_ has joined #openstack-ironic | 06:19 | |
| openstackgerrit | Michael Davies proposed openstack/ironic-lib: Switch default partition type to gpt https://review.openstack.org/225556 | 06:21 |
| *** Marga_ has quit IRC | 06:21 | |
| *** Sukhdev has quit IRC | 06:37 | |
| *** ukalifon has joined #openstack-ironic | 06:42 | |
| *** e0ne has joined #openstack-ironic | 07:06 | |
| *** davideagnello has joined #openstack-ironic | 07:06 | |
| *** dimsum__ has joined #openstack-ironic | 07:08 | |
| *** jamielennox is now known as jamielennox|away | 07:09 | |
| *** davideagnello has quit IRC | 07:11 | |
| *** dimsum__ has quit IRC | 07:13 | |
| *** chlong has quit IRC | 07:17 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add support for in-band cleaning in IloVirtualMediaIscsiDeploy https://review.openstack.org/224174 | 07:19 |
| *** dtantsur|afk is now known as dtantsur | 07:32 | |
| dtantsur | Morning Ironic | 07:32 |
| *** yog_ has joined #openstack-ironic | 07:33 | |
| mrda | hi dtantsur | 07:33 |
| *** ifarkas has joined #openstack-ironic | 07:33 | |
| mrda | dtantsur: just a question on ironic-lib - where are we at? We haven't started using yet, right? But what is the plan for when we will switch over? | 07:34 |
| dtantsur | mrda, after the feature freeze :) and we still need https://review.openstack.org/#/c/212495/ to land | 07:35 |
| dtantsur | mrda, for now, we've made the first release and added it to g-r | 07:35 |
| dtantsur | mrda, we've also frozen changing ironic parts that went to ironic-lib, so that they don't get desynced | 07:35 |
| mrda | ok, thanks | 07:36 |
| mrda | just emans we'd better switch over soon or else the new stuff added to ironic-lib won't be available until we do :) | 07:36 |
| dtantsur | yeah | 07:37 |
| * mrda just posted a patch for ironic-lib | 07:37 | |
| *** achanda has quit IRC | 07:38 | |
| *** romainh has joined #openstack-ironic | 07:39 | |
| *** achanda has joined #openstack-ironic | 07:39 | |
| *** ndipanov has joined #openstack-ironic | 07:40 | |
| *** baoli has joined #openstack-ironic | 07:41 | |
| *** Nisha_away has quit IRC | 07:43 | |
| *** baoli has quit IRC | 07:46 | |
| *** Marga_ has joined #openstack-ironic | 07:52 | |
| *** praneshp has quit IRC | 07:57 | |
| *** Marga_ has quit IRC | 07:57 | |
| *** Haomeng|2 has joined #openstack-ironic | 07:59 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic-specs: Using multicast for image transfer https://review.openstack.org/225570 | 08:02 |
| *** Haomeng has quit IRC | 08:02 | |
| *** lucasagomes has joined #openstack-ironic | 08:03 | |
| openstackgerrit | Bob Ball proposed openstack/ironic: Handle unquoted node names from virt types https://review.openstack.org/224770 | 08:07 |
| *** athomas has joined #openstack-ironic | 08:10 | |
| *** pradipta has joined #openstack-ironic | 08:11 | |
| openstackgerrit | Merged stackforge/proliantutils: Fix ribcl to return correct number of processors https://review.openstack.org/221678 | 08:11 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic-specs: Using multicast for image transfer https://review.openstack.org/225570 | 08:12 |
| lucasagomes | BobBall, ^ there's a pep8 error in that patch | 08:12 |
| lucasagomes | and morning all! | 08:13 |
| yuriyz | morning lucasagomes Ironic | 08:15 |
| betherly | Morning ironic! | 08:15 |
| dtantsur | morning lucasagomes, yuriyz, betherly! | 08:18 |
| *** pelix has joined #openstack-ironic | 08:19 | |
| *** Nisha_away has joined #openstack-ironic | 08:19 | |
| *** achanda has quit IRC | 08:20 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic: Fix conductor deregistration on non init conductor https://review.openstack.org/225573 | 08:20 |
| BobBall | thanks lucasagomes... Serves me right for thinking I didn't need to re-run the local tests just for a minor doc change :D | 08:23 |
| lucasagomes | BobBall, heh yeah! Happens :-( | 08:23 |
| openstackgerrit | Bob Ball proposed openstack/ironic: Handle unquoted node names from virt types https://review.openstack.org/224770 | 08:24 |
| BobBall | I did run pep8 this time :P | 08:24 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Validate the input of properties of nodes. https://review.openstack.org/215505 | 08:25 |
| *** rcarrillocruz has joined #openstack-ironic | 08:40 | |
| rcarrillocruz | morning folks | 08:40 |
| openstackgerrit | Merged openstack/ironic: Fix iRMC vmedia deploy failure due to already attached image https://review.openstack.org/225521 | 08:40 |
| *** baoli has joined #openstack-ironic | 08:42 | |
| *** MattMan has joined #openstack-ironic | 08:42 | |
| *** chlong has joined #openstack-ironic | 08:44 | |
| *** Nisha_away has quit IRC | 08:45 | |
| *** alexpilotti has joined #openstack-ironic | 08:46 | |
| *** e0ne has quit IRC | 08:46 | |
| *** baoli has quit IRC | 08:47 | |
| openstackgerrit | Merged openstack/ironic: Implement take_over for iscsi_ilo driver https://review.openstack.org/217068 | 08:49 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Refactor agent {prepare,tear_down}_cleaning into deploy_utils https://review.openstack.org/220895 | 08:50 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add support for in-band cleaning in ISCSIDeploy https://review.openstack.org/220898 | 08:51 |
| lucasagomes | rameshg87, hi there, yeah that's a good suggestion on the base.py | 08:51 |
| rameshg87 | lucasagomes: hi | 08:51 |
| lucasagomes | rameshg87, I wonder if I should remove it there or just leave it as a plumbing work for when we have abort for cleaning state | 08:52 |
| lucasagomes | probably that will be misleading tho | 08:52 |
| rameshg87 | lucasagomes: yeah, I wondered if it was needed since it is used only for inband clean steps for now | 08:52 |
| lucasagomes | rameshg87, yeah | 08:52 |
| lucasagomes | I will remove it | 08:52 |
| lucasagomes | probably add a TODO there talking about it | 08:52 |
| rameshg87 | lucasagomes: yeah, may be we can add it later | 08:52 |
| rameshg87 | okay | 08:53 |
| lucasagomes | rameshg87, thanks much! | 08:53 |
| rameshg87 | lucasagomes: wc :) | 08:53 |
| *** boris-42 has quit IRC | 08:53 | |
| *** Marga_ has joined #openstack-ironic | 08:53 | |
| *** Marga_ has quit IRC | 08:58 | |
| *** derekh has joined #openstack-ironic | 08:59 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add support for in-band cleaning in IloVirtualMediaIscsiDeploy https://review.openstack.org/224174 | 09:00 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Fix final comments in RAID commits https://review.openstack.org/225590 | 09:06 |
| *** davideagnello has joined #openstack-ironic | 09:07 | |
| *** romcheg has joined #openstack-ironic | 09:10 | |
| rameshg87 | lucasagomes: one quick question | 09:10 |
| rameshg87 | lucasagomes: in https://review.openstack.org/#/c/219603/4 | 09:10 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic: Fix conductor deregistration on non init conductor https://review.openstack.org/225573 | 09:10 |
| rameshg87 | lucasagomes: is it not possible to commit something to stable/kilo that's not from master branch | 09:11 |
| rameshg87 | lucasagomes: because there are some differences in documentation in master vs stable/kilo branch | 09:11 |
| *** davideagnello has quit IRC | 09:11 | |
| rameshg87 | lucasagomes: in master branch, ilo driver's don't pass auth_token | 09:11 |
| rameshg87 | lucasagomes: in stable/kilo branch, it passes | 09:11 |
| *** mgoddard has joined #openstack-ironic | 09:12 | |
| lucasagomes | rameshg87, hi there... hmmm I think you always have to cherry-pick it from master | 09:13 |
| lucasagomes | and resolve those conflicts manually | 09:14 |
| rameshg87 | lucasagomes: okay, so first cherry-pick, and then edit ? | 09:14 |
| lucasagomes | and then propose the result of it (keeping the conflicted files in the commit message) | 09:14 |
| rameshg87 | lucasagomes: but why so ? is it a workflow requirement ? :) | 09:14 |
| lucasagomes | rameshg87, yeah | 09:14 |
| lucasagomes | rameshg87, I'm not sure to be honest | 09:14 |
| lucasagomes | lemme see if there's a guide | 09:14 |
| rameshg87 | okay | 09:15 |
| lucasagomes | rameshg87, https://wiki.openstack.org/wiki/StableBranch#Appropriate_Fixes | 09:16 |
| lucasagomes | rameshg87, if it's only doc, it may fall on that rule that the change does not need to come from master because it doesn't touch production code | 09:17 |
| rameshg87 | a change must be a backport of a change already merged onto master, unless the change simply does not make sense on master. | 09:17 |
| rameshg87 | lucasagomes: also ^^^ | 09:17 |
| lucasagomes | yeah | 09:17 |
| rameshg87 | lucasagomes: okay, so I needn't do a cherry-pick and can have an independently raised review ? | 09:18 |
| lucasagomes | rameshg87, seems like that yes | 09:18 |
| lucasagomes | I'm updating the comment on that patch | 09:18 |
| rameshg87 | lucasagomes: okay, let me just move on with that then .. | 09:18 |
| rameshg87 | lucasagomes: thanks | 09:19 |
| rameshg87 | I will fix comments in it | 09:19 |
| lucasagomes | rameshg87, ok thanks! | 09:19 |
| *** boris-42 has joined #openstack-ironic | 09:24 | |
| xek | lucasagomes, Hi, I made a change to the patch https://review.openstack.org/#/c/224079/ which deals with the random timeouts I was having | 09:26 |
| lucasagomes | xek, nice one! I will take a look soonish | 09:27 |
| xek | lucasagomes, namely, I commented out remotable decorators on object methods (there is a note in the commit description) | 09:28 |
| xek | lucasagomes, ok, thanks! | 09:28 |
| *** romcheg has quit IRC | 09:28 | |
| *** romcheg has joined #openstack-ironic | 09:30 | |
| *** Marga_ has joined #openstack-ironic | 09:30 | |
| *** athomas has quit IRC | 09:32 | |
| *** dtantsur is now known as dtantsur|brb | 09:32 | |
| openstackgerrit | sonu proposed openstack/python-ironicclient: Added unit test cases for command-line shell https://review.openstack.org/213654 | 09:33 |
| *** Marga_ has quit IRC | 09:35 | |
| *** mgoddard has quit IRC | 09:35 | |
| *** mgoddard has joined #openstack-ironic | 09:36 | |
| *** athomas has joined #openstack-ironic | 09:42 | |
| *** baoli has joined #openstack-ironic | 09:43 | |
| *** dimsum__ has joined #openstack-ironic | 09:47 | |
| *** baoli has quit IRC | 09:48 | |
| *** Marga_ has joined #openstack-ironic | 09:52 | |
| *** Marga_ has quit IRC | 09:54 | |
| *** Marga__ has joined #openstack-ironic | 09:54 | |
| lucasagomes | rameshg87, actually looking at the clean_step decorator, it seems that it can be an async clean step there | 09:54 |
| lucasagomes | rameshg87, depending on the return value | 09:54 |
| rameshg87 | lucasagomes: but for async clean step, do we change the provision state back to CLEANWAIT ? | 09:55 |
| lucasagomes | https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L1000-L1002 | 09:56 |
| rameshg87 | I mean aysnc out-of-band clean step | 09:56 |
| lucasagomes | rameshg87, we do yes | 09:56 |
| * lucasagomes finds the place | 09:56 | |
| rameshg87 | lucasagomes: it just returns CLEANWAIT iirc | 09:56 |
| rameshg87 | lucasagomes: we don't change the provision state | 09:56 |
| lucasagomes | rameshg87, https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L995-L1002 | 09:56 |
| lucasagomes | we do | 09:57 |
| rameshg87 | oh | 09:57 |
| lucasagomes | we get the result, if it's cleanwait indicating it's an async task | 09:57 |
| rameshg87 | lucasagomes: okay, new thing for me | 09:57 |
| lucasagomes | we set it back to cleanwait state | 09:57 |
| rameshg87 | lucasagomes: so it's upto the async task (probably a periodic task) to do resume | 09:57 |
| lucasagomes | rameshg87, ok, so I will keep the stuff in the decorator right? | 09:57 |
| lucasagomes | seems so | 09:57 |
| rameshg87 | lucasagomes: before it triggers continue clean to the conductor | 09:58 |
| *** e0ne has joined #openstack-ironic | 09:58 | |
| rameshg87 | lucasagomes: +1, we need it in the decorator then :) | 09:58 |
| lucasagomes | rameshg87, yeah, I was just checking it before removing to make sure | 09:58 |
| * lucasagomes still needs to learn more about cleaning | 09:58 | |
| * rameshg87 too | 09:58 | |
| *** Marga__ has quit IRC | 09:59 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Move iLO documentation to wiki https://review.openstack.org/225602 | 10:00 |
| *** naohirot has quit IRC | 10:00 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Allow abort for CLEANWAIT states https://review.openstack.org/201552 | 10:03 |
| lucasagomes | rameshg87, ^ that should do it | 10:03 |
| rameshg87 | lucasagomes: thanks | 10:03 |
| *** romainh has quit IRC | 10:25 | |
| *** rcarrillocruz has quit IRC | 10:26 | |
| *** Nisha has joined #openstack-ironic | 10:28 | |
| *** dtantsur|brb is now known as dtantsur | 10:33 | |
| sambetts | Morning all! | 10:43 |
| sambetts | Thanks for the +2 on my driver dtantsur | 10:43 |
| sambetts | ! | 10:43 |
| *** baoli has joined #openstack-ironic | 10:44 | |
| dtantsur | morning sambetts, you're welcome :) | 10:45 |
| *** baoli has quit IRC | 10:48 | |
| lucasagomes | sambetts, morning | 10:48 |
| sambetts | Hey lucasagomes | 10:49 |
| *** mgoddard has quit IRC | 10:53 | |
| *** Marga_ has joined #openstack-ironic | 10:55 | |
| *** mgoddard has joined #openstack-ironic | 10:58 | |
| *** Marga_ has quit IRC | 10:59 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic-specs: Proxy configuration for agent images downloading https://review.openstack.org/225181 | 11:00 |
| BobBall | Hey all... Anyone got a hint for rebuilding a dracut initramfs? I've got one from RHEL for the deploy image, but I need to make some tweaks to the python in the initramfs... cpio -i only extracts the kernel microcode. I can't help but think I'm being blind to something obvious because it shouldn't be _hard_ to modify an initramfs! :) | 11:00 |
| *** rcarrillocruz has joined #openstack-ironic | 11:03 | |
| sambetts | lucasagomes: Thanks for the +2 :D | 11:04 |
| lucasagomes | sambetts, np :-D | 11:05 |
| *** bradjones has joined #openstack-ironic | 11:05 | |
| *** bradjones has quit IRC | 11:05 | |
| *** bradjones has joined #openstack-ironic | 11:05 | |
| lucasagomes | BobBall, hmm I think you may be better off recreating the image with DIB | 11:06 |
| lucasagomes | you can use the initramfs/dracut tools to inspect and modify it if you want to | 11:07 |
| lucasagomes | but microcode is a bit more complicated | 11:07 |
| BobBall | I don't want to mess with the microcode - just that cpio is the only thing extracted | 11:07 |
| BobBall | Very hesitant to recreate as I don't know how it was created in the first place :D | 11:07 |
| *** davideagnello has joined #openstack-ironic | 11:08 | |
| *** bradjones has quit IRC | 11:08 | |
| BobBall | lsinitrd shows all files, including the microcode, but the standard approach for extracting from an initramfs of using cpio -i < <img> only extracts the microcode - not the rest of the files. | 11:08 |
| *** bradjones has joined #openstack-ironic | 11:08 | |
| *** bradjones has quit IRC | 11:08 | |
| *** bradjones has joined #openstack-ironic | 11:08 | |
| BobBall | I haven't been able to find how to use dracut to extract / repack an existing initramfs so if you have pointers that would be amazing :) | 11:08 |
| *** Haomeng has joined #openstack-ironic | 11:09 | |
| lucasagomes | not sure either, lemme see, not sure if customizing the microcode is actually possible tho | 11:10 |
| BobBall | I don't want to change microcode - just the python scripts also included in the initramfs | 11:10 |
| lucasagomes | BobBall, right, can't you do a dracut --include (or something like that) to inject files and recreate the ramdisk? | 11:11 |
| lucasagomes | you already have the python scripts handy right? | 11:11 |
| lucasagomes | BobBall, one thing I wonder is that, I'm not sure there are python scripts on those ramdisks. If you have python script you have the interpreter which is pretty heavy | 11:12 |
| BobBall | heh. That'll do, thanks. How did I miss that. I'll try doing that | 11:12 |
| *** Haomeng|2 has quit IRC | 11:12 | |
| BobBall | Yeah, it is heavy :) lsinitrd shows the interpreter + python scripts | 11:12 |
| BobBall | This is for the deploy ramdisk initramfs | 11:12 |
| *** davideagnello has quit IRC | 11:13 | |
| lucasagomes | alright ok | 11:13 |
| lucasagomes | (just saying cause usually we use this https://github.com/openstack/diskimage-builder/blob/master/elements/deploy-ironic/init.d/80-deploy-ironic ) | 11:13 |
| *** bradjones has quit IRC | 11:14 | |
| lucasagomes | and a rhel ramdisk could be created with: "$ ramdisk-image-create -o <output file> rhel deploy-ironic dracut-ramdisk" | 11:14 |
| BobBall | Ah yes - sorry... I was mixing things up. There is indeed a python interpreter in the initramfs but I want to change the shell script providing find_disk ;) | 11:14 |
| *** bradjones has joined #openstack-ironic | 11:14 | |
| *** bradjones has quit IRC | 11:14 | |
| *** bradjones has joined #openstack-ironic | 11:14 | |
| lucasagomes | BobBall, cool, yeah that function really sucks | 11:15 |
| BobBall | the initramfs is 54MB ;) | 11:15 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 11:15 |
| BobBall | Sucks most because it doesn't know about 'xvda' for Xen disks! | 11:15 |
| lucasagomes | fwiw we have deprecated this bash ramdisk this cycle, it will be around for the next cycle but will probably be removed later | 11:15 |
| BobBall | Patch will be coming at some point... | 11:15 |
| BobBall | Ah | 11:16 |
| lucasagomes | BobBall, cool. That's good to fix such things | 11:16 |
| lucasagomes | if you have time you can fix it for IPA (if needed) as well | 11:16 |
| lucasagomes | that's our default ramdisk now | 11:16 |
| lucasagomes | https://github.com/openstack/ironic-python-agent | 11:16 |
| lucasagomes | and it's python, so no more bash! | 11:16 |
| lucasagomes | BobBall, that's the bits where we look for the disk in IPA https://github.com/openstack/ironic-python-agent/blob/88e22ad852fed2dc8ce0c63c067d868f11572ca9/ironic_python_agent/utils.py#L287 | 11:17 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: Expose serial, wwn and vendor on the BlockDevice object https://review.openstack.org/225102 | 11:17 |
| lucasagomes | BobBall, also, for the ramdisk to find xvda you can modify a config option in Ironic called "disk_devices" | 11:18 |
| lucasagomes | you can include xvda there | 11:18 |
| TheJulia | Good morning everyone | 11:19 |
| lucasagomes | and it will look for /dev/xvda | 11:19 |
| BobBall | Yes, except diskimage-builder/elements/ramdisk-base/extra-data.d/scripts/d/init-func then discards it | 11:19 |
| BobBall | https://github.com/openstack/diskimage-builder/blob/master/elements/ramdisk-base/extra-data.d/scripts/d/init-func#L132 | 11:19 |
| BobBall | It just returns the empty string, even though it was passed a valid - and existing - disk in ;) | 11:20 |
| lucasagomes | oh dear | 11:20 |
| lucasagomes | *sigh* | 11:20 |
| BobBall | Same problem in next function | 11:20 |
| lucasagomes | BobBall, yeah that's a not so smart assumption indeed | 11:20 |
| lucasagomes | heh yeah thanks for fixing it | 11:20 |
| lucasagomes | TheJulia, morning | 11:20 |
| BobBall | Well - I was just going to add xvd* to the list, since I don't really understand why it's got those distinctions... | 11:21 |
| * lucasagomes neither | 11:21 | |
| BobBall | Might make more sense to just strip out [0-9]* in the generic case | 11:21 |
| BobBall | but I assume it was written that way for a reason | 11:21 |
| lucasagomes | yeah, git blame or ask at #tripleo | 11:22 |
| BobBall | i.e. unless we're certain how the partitions are numbered then don't guess how to remove them or add them | 11:22 |
| *** lucasagomes is now known as lucas-hungry | 11:29 | |
| * lucas-hungry go grab some lunch | 11:29 | |
| *** pradipta has quit IRC | 11:30 | |
| *** romainh has joined #openstack-ironic | 11:34 | |
| dtantsur | sambetts, ifarkas, just FYI: I've created stable/liberty for python-ironic-inspector-client | 11:35 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor agent_ilo driver to use new boot interface https://review.openstack.org/217102 | 11:36 |
| sambetts | dtantsur: :) nice! | 11:41 |
| *** rameshg87 has left #openstack-ironic | 11:46 | |
| *** trown|outttypeww is now known as trown | 11:49 | |
| *** Marga_ has joined #openstack-ironic | 11:56 | |
| *** Haomeng|2 has joined #openstack-ironic | 11:58 | |
| *** rcarrillocruz has quit IRC | 12:00 | |
| *** Marga_ has quit IRC | 12:00 | |
| *** rcarrillocruz has joined #openstack-ironic | 12:00 | |
| *** Haomeng has quit IRC | 12:00 | |
| *** maurosr is now known as murilovf | 12:02 | |
| *** murilovf is now known as maurosr | 12:02 | |
| *** nicodemos has joined #openstack-ironic | 12:05 | |
| dtantsur | lucas-hungry, sambetts, could you please review an inspector-related IPA patch https://review.openstack.org/#/c/225092 ? | 12:06 |
| *** priteau has joined #openstack-ironic | 12:12 | |
| *** baoli has joined #openstack-ironic | 12:17 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Fix default log file location https://review.openstack.org/225647 | 12:23 |
| *** VikasC has quit IRC | 12:26 | |
| *** lucas-hungry is now known as lucasagomes | 12:30 | |
| lucasagomes | dtantsur, yup | 12:30 |
| jroll | g'morning BobBall, lucasagomes, dtantsur, sambetts, TheJulia, anyone else maybe lurking :) | 12:32 |
| lucasagomes | jroll, hi there! g'morning! | 12:32 |
| jroll | xek: hey, is 224079 just a partial bug fix with removing the remoteable decorators? | 12:33 |
| dtantsur | morning jroll! | 12:33 |
| jroll | xek: s/just/now/ that is :) | 12:33 |
| xek | jroll, no, adding indirection_api caused the calls that were decorated to be remoted via RPC | 12:34 |
| xek | I had to remove the decorators, because I didn't want to change the current behavior | 12:34 |
| jroll | xek: oh, so once that's flipped on the api was RPCing to the conductor for all objects methods? | 12:35 |
| xek | jroll, yes | 12:35 |
| jroll | heh | 12:35 |
| jroll | neat | 12:35 |
| * jroll wonders if we want to do that or not | 12:35 | |
| jroll | it's weird that was timing out | 12:35 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: WIP: Include token value for no-auth mode https://review.openstack.org/225187 | 12:36 |
| xek | jroll, I think in the future, we would like to replace the current explicit RPC calls with remotable object methods | 12:38 |
| *** linggao has joined #openstack-ironic | 12:39 | |
| *** athomas has quit IRC | 12:39 | |
| jroll | xek: ah, I see, e.g. objects.Node.do_node_deploy() rather than rpcapi.do_node_deploy() | 12:40 |
| xek | jroll, f.ex. the update_node RPC call -> node.update() | 12:40 |
| xek | jroll, yes | 12:40 |
| jroll | yep | 12:40 |
| *** rloo has joined #openstack-ironic | 12:41 | |
| jroll | awesome, thanks for helping with this | 12:41 |
| *** Nisha has quit IRC | 12:41 | |
| sambetts | Morning jroll o/ | 12:42 |
| jroll | ohai \o | 12:42 |
| *** e0ne has quit IRC | 12:44 | |
| *** athomas has joined #openstack-ironic | 12:45 | |
| *** dimsum__ has quit IRC | 12:47 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor IRMCVirtualMediaIscsiDeploy by applying new BootInterface https://review.openstack.org/221371 | 12:47 |
| *** dimsum__ has joined #openstack-ironic | 12:47 | |
| *** dimsum__ is now known as dims | 12:48 | |
| *** dims is now known as Guest31942 | 12:49 | |
| *** Guest31942 is now known as dims__ | 12:49 | |
| *** Marga_ has joined #openstack-ironic | 12:56 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Support IPA in raid_device plugin https://review.openstack.org/225658 | 12:57 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: WIP: Include token value for no-auth mode https://review.openstack.org/225187 | 13:00 |
| *** Marga_ has quit IRC | 13:01 | |
| *** davideagnello has joined #openstack-ironic | 13:09 | |
| BobBall | Any suggestions for how to debug 'iSCSI connection was not seen by the file system after attempting to verify 3 times' during deploy? Is there a way to stop the node from shutting down automatically and/or log me into a shell so I can pause the environment and try and debug what's going on? | 13:12 |
| *** davideagnello has quit IRC | 13:13 | |
| *** cdearborn has joined #openstack-ironic | 13:13 | |
| jroll | xek: I bet if we start ironic-conductor before ironic-api, we wouldn't get those timeouts... noticing the RPC call that's timing out is sent before the RPC service is up on the conductor | 13:14 |
| *** liliars has joined #openstack-ironic | 13:16 | |
| lucasagomes | BobBall, hmmmm that's a bit hard to debug, I would say increase the number of attempts in the ironic-conductor | 13:17 |
| lucasagomes | BobBall, now to log in the ramdisk, you have access to the console? If so, I would say that the esiest way to do it is to create a ramdisk image with a user | 13:18 |
| dtantsur | lucasagomes, when you have a second of time, could you rereview https://review.openstack.org/#/c/225102 please? | 13:18 |
| lucasagomes | or if you are tweaking the bash, you can call "bash" there and it will give you a shell | 13:18 |
| lucasagomes | dtantsur, ack, I opened the first in a tab haven't looked into it yet | 13:19 |
| BobBall | OK - but either way I have to rebuild the ramdisk. OK. | 13:19 |
| xek | jroll, so the test script is started before the services are up? | 13:19 |
| dtantsur | lucasagomes, thanks! I know, review pressure is bad before the release :) | 13:19 |
| lucasagomes | BobBall, unfortunately yes :-( | 13:19 |
| lucasagomes | BobBall, other ramdisks, such as the coreos one you can just pass stuff in the command line (nologin, or ssh_key=<key>) | 13:20 |
| lucasagomes | but the bash ramdisk is a bit limited | 13:20 |
| BobBall | I see, thanks | 13:21 |
| *** marios has quit IRC | 13:22 | |
| jroll | xek: well, the errors are in devstack set up. we turn up nova, then ironic services. so nova is polling for nodes before ironic services are up (which is fine) however I think it's hitting a race where ironic-api is up but ironic-conductor is not. does that make sense? | 13:22 |
| *** marios has joined #openstack-ironic | 13:24 | |
| jroll | xek: anyway, I think it's best to leave remotable disabled for now since that's a massive behavior change. left you a review, just a couple minor things if you want to fix them up quickly :) | 13:24 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Allow empty lookup attributes if node_not_found_hook is provided https://review.openstack.org/225668 | 13:26 |
| xek | jroll, ah yes, I saw that the script stopped on wait_for_nova_resources but didn't analyze it further | 13:27 |
| xek | jroll, getting to it | 13:27 |
| jroll | cool, thanks for doing this work! | 13:28 |
| *** superdan is now known as dansmith | 13:34 | |
| dtantsur | jroll, devananda, do we have any kind of abandon policy for stale patches? | 13:39 |
| jroll | dtantsur: we do not | 13:40 |
| jroll | (afaik) | 13:41 |
| dtantsur | I see | 13:41 |
| jroll | people get cranky when you do it, apparently | 13:41 |
| dtantsur | jroll, well, yeah, but what should we do about assigned bugs with patches -1'ed for 2 months? | 13:42 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Refactor IRMCVirtualMediaAgentDeploy by applying new BootInterface https://review.openstack.org/221577 | 13:42 |
| jroll | dtantsur: well, we should definitely unassign the bug... then try to get someone to pick up the patch and run with it? | 13:42 |
| jroll | dtantsur: alternatively, if another patch fixes the bug, it makes sense to abandon the old patch | 13:43 |
| dtantsur | makes sense | 13:43 |
| jroll | just my 2c though :) | 13:43 |
| *** r-daneel has joined #openstack-ironic | 13:46 | |
| *** e0ne has joined #openstack-ironic | 13:47 | |
| jroll | rloo: good morning :) | 13:48 |
| jroll | rloo: this is all that we need for in-band raid, right? mind fixing that up today or is that a ramesh thing? https://review.openstack.org/#/c/224938/ | 13:48 |
| lucasagomes | jroll, ++ yeah, I think nova is doing something about it no? At least I got some patches abandoned there | 13:49 |
| jroll | lucasagomes: yeah, I remember some back and forth as to whether that's the right thing to do, idk what folks decided | 13:50 |
| jroll | anyway, we don't have a policy around it, but if you think a patch should be abandoned I would say just do it | 13:50 |
| jroll | (assuming you're comfortable with that, knowing it might frustrate a contributor) | 13:50 |
| lucasagomes | +1 if it's there for ages with no activity at all I think it's fine to abandon | 13:51 |
| dtantsur | jroll, what's the status of https://bugs.launchpad.net/ironic/+bug/1382589 ? | 13:51 |
| openstack | Launchpad bug 1382589 in Ironic "Conductor stops heartbeating when exception is raised in heartbeat worker" [Medium,In progress] - Assigned to Jim Rollenhagen (jim-rollenhagen) | 13:51 |
| jroll | dtantsur: you think the inband iscsideploy cleaning feature thing is safe to land this week or no? | 13:51 |
| * jroll checks | 13:51 | |
| dtantsur | jroll, it's disabled by default, so I'd land it, if we can | 13:52 |
| jroll | ok cool, will review | 13:52 |
| jroll | dtantsur: on that bug, "We need to think about how to handle other exceptions that may come up" is why I only tagged it partial | 13:52 |
| jroll | maybe we should just close it, idk | 13:53 |
| jroll | I don't see many other errors that could happen there | 13:53 |
| dtantsur | jroll, could you mark it as done then? | 13:53 |
| jroll | sure | 13:53 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Add alembic migrations for the inspector database https://review.openstack.org/223665 | 13:53 |
| jroll | dtantsur: done, sorry about that | 13:54 |
| dtantsur | JoshNang, hi! what is left in https://bugs.launchpad.net/ironic/+bug/1493554 ? could you please update the bug? | 13:54 |
| openstack | Launchpad bug 1493554 in Ironic "Returning processutils output in IPA clean step causes API error" [Medium,In progress] - Assigned to Josh Gachnang (joshnang) | 13:54 |
| openstackgerrit | Merged openstack/ironic: Handle unquoted node names from virt types https://review.openstack.org/224770 | 13:54 |
| dtantsur | jroll, no prob, I'm just doing some housekeeping :) | 13:54 |
| jroll | dtantsur: also, you said help wanted with triage etc, you just need more brain power on it or what? | 13:55 |
| dtantsur | jroll, I will start dropping hard-to-triage bugs to the status report. yes, brain power will be needed to try figure out root cause | 13:55 |
| jroll | got it, thanks | 13:56 |
| *** sdake has joined #openstack-ironic | 13:57 | |
| *** Marga_ has joined #openstack-ironic | 13:57 | |
| dtantsur | lucasagomes, now your turn :) what's the status of https://bugs.launchpad.net/ironic/+bug/1177655 ? | 13:57 |
| openstack | Launchpad bug 1177655 in Ironic "kernel boot command line for baremetal assumes block device UUID is correct root" [Medium,In progress] - Assigned to Lucas Alvares Gomes (lucasagomes) | 13:57 |
| openstackgerrit | Grzegorz Grasza (xek) proposed openstack/ironic: Fix rolling upgrades by implementing indirection_api https://review.openstack.org/224079 | 13:58 |
| dtantsur | lucasagomes, and the same question for https://bugs.launchpad.net/ironic/+bug/1451885 | 13:58 |
| openstack | Launchpad bug 1451885 in Ironic "Clean up driver_internal_info after tear_down" [Low,In progress] - Assigned to Lucas Alvares Gomes (lucasagomes) | 13:58 |
| * lucasagomes looks | 13:58 | |
| lucasagomes | dtantsur, we can close that first bug, because it's being fixed as part of a spec | 13:59 |
| lucasagomes | https://review.openstack.org/#/c/178229/ | 13:59 |
| dtantsur | lucasagomes, please do :) | 13:59 |
| lucasagomes | Blake did rebase it and reproposed, it seems to be failing some tests now I will take a look | 14:00 |
| * lucasagomes closes it | 14:00 | |
| dtantsur | lucasagomes, thanks. what about the 2nd? | 14:00 |
| *** david-ly_ is now known as david-lyle | 14:01 | |
| * lucasagomes reading | 14:01 | |
| *** Marga_ has quit IRC | 14:02 | |
| rloo | hi everyone | 14:02 |
| lucasagomes | dtantsur, yeah urgh... it's a bit messy right now because we have some bits where the conductor (manager.py) add stuff to the driver_internal_info | 14:02 |
| lucasagomes | lemme untag myself from that bug | 14:02 |
| rloo | jroll: catchin' up on stuff. haven't gotten to 224938 yet. Let me see. | 14:02 |
| dtantsur | lucasagomes, thanks! | 14:03 |
| jroll | rloo: merge conflicts :( | 14:03 |
| rloo | jroll: that i can probably fix, even though I have no idea what state the patch is in :-) | 14:03 |
| dtantsur | NobodyCam, hey! I've reached letter N, so now a question for you: what's the status of https://bugs.launchpad.net/ironic/+bug/1282836 ? | 14:04 |
| openstack | Launchpad bug 1282836 in Ironic "PXE driver does not check if tftpd is running" [Low,In progress] - Assigned to Chris Krelle (nobodycam) | 14:04 |
| jroll | rloo: heh, ok, I should have known ramesh took it over but didn't check until now :) | 14:04 |
| *** afaranha has joined #openstack-ironic | 14:04 | |
| *** jlvillal has joined #openstack-ironic | 14:05 | |
| rloo | dtantsur: as far as abandoning patches/bugs. Nova has some timeframe they use, can't recall if it is 1 or 2 months or something, and i think they even have some automated script. but i'm not sure we want to go that route. seems ... kinda cold. dunno. | 14:05 |
| rloo | jroll: no worries, i coauthored so can't +2 on that anyway. | 14:05 |
| dtantsur | rloo, I'll be doing it manually and with warmth :D | 14:05 |
| dtantsur | and g'morning rloo! | 14:05 |
| rloo | dtantsur: ++ for warmth! :D | 14:06 |
| *** mgoddard1 has joined #openstack-ironic | 14:07 | |
| *** mgoddard has quit IRC | 14:08 | |
| rloo | who approved 'memory_consumed_by_agent' config? the help string isn't that clear to me. guess i need to look at the patch one day. | 14:08 |
| jroll | rloo: may have been me :/ | 14:09 |
| rloo | 'memory_consumed_by_agent' doesn't quite make sense to me either. how do you know how much was already consumed. I suspect it means something else. | 14:09 |
| *** e0ne has quit IRC | 14:10 | |
| rloo | jroll: no worries. i'll look later. that's the merge conflict/culprit :) | 14:10 |
| jroll | rloo: it's more of "expected memory consumed by agent" | 14:10 |
| jroll | but yeah, we can revisit | 14:10 |
| rloo | jroll: words are hard. as we all know! | 14:10 |
| jroll | :) | 14:11 |
| *** lazy_prince has joined #openstack-ironic | 14:11 | |
| sambetts | Morning rloo | 14:12 |
| jroll | rloo: would you like to review this before I land it then? :) https://review.openstack.org/#/c/220898/9/etc/ironic/ironic.conf.sample | 14:12 |
| *** krtaylor has quit IRC | 14:13 | |
| rloo | jroll: it doesn't make sense to me. what does 'IPA deploy ramdisk should be used... before this CONF option is set to True' mean? | 14:13 |
| rloo | jroll: if it is deprecated, why doesn't it show up in the original? | 14:14 |
| *** e0ne has joined #openstack-ironic | 14:14 | |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Convert eDeploy data so that rules can process it https://review.openstack.org/225168 | 14:14 |
| jroll | rloo: so this enables cleaning for the pxe_* drivers... but can only be used with IPA ramdisk, not bash ramdisk. and it's being added as deprecated because the bash ramdisk is deprecated and we'll consolidate that with agent.enable_cleaning or whatever it's called | 14:15 |
| rloo | jroll: sigh. so early in the morning and i have to think about words. | 14:16 |
| rloo | jroll: i think what it is saying is that if you set this to True, you need to be using IPA deploy ramdisk. | 14:16 |
| jroll | rloo: correct | 14:16 |
| rloo | jroll: and what does it mean if it is set to false? no cleaning at all? | 14:17 |
| jroll | rloo: correct | 14:17 |
| rloo | jroll: what's the diff between that new config and the existing 'clean_nodes'? | 14:18 |
| jroll | rloo: the new one only controls pxe_*, so the idea is you could have some agent_* nodes clean while pxe_* nodes do not | 14:19 |
| *** yog_ has quit IRC | 14:19 | |
| rloo | jroll: is 'clean_nodes' specific to agent then? | 14:19 |
| BobBall | omg... lucasagomes FYI the iscsiadm sessions were/are borked. Perhaps my continual earlier failures found a condition where iscsi sessions weren't cleaned up properly. Anyway, one of the iscsiadm commands has been failing routinely. A server reboot fixed the initial condition, now I just need to clear out all of the old sessions which still exist | 14:19 |
| jroll | rloo: clean_nodes is global afaik | 14:20 |
| lucasagomes | BobBall, ew, Ironic should try to delete the sessions no? | 14:20 |
| rloo | jroll: can you give me 30 min to think about this and get the other patch merged first. i can't multitask. | 14:20 |
| jroll | rloo: yeah no worries, there's 2+2 on there so I'll leave it to you | 14:20 |
| BobBall | Yes, of course it should. But for some reason it wasn't (as I say - unusual code path perhaps) | 14:20 |
| *** krtaylor has joined #openstack-ironic | 14:22 | |
| *** david-lyle has quit IRC | 14:22 | |
| lucasagomes | yeah :-/ | 14:24 |
| dtantsur | jroll, wdyt about approving Cisco IMC driver https://review.openstack.org/#/c/219253/ ? It got 2x +2, we're waiting for your ack | 14:25 |
| BobBall | So I couldn't even get iscsiadm to delete the nodes as it was trying to connect to the target (which didn't exist) to remove them(!) | 14:25 |
| BobBall | Brain surgery time. | 14:25 |
| *** david-lyle has joined #openstack-ironic | 14:26 | |
| jroll | dtantsur: it's probably fine, I just hope we don't end up with a bunch of backports if things are broken :) | 14:26 |
| dtantsur | sambetts, ^^ your promise needed :D | 14:27 |
| jroll | dang, that's quite large | 14:27 |
| jroll | sambetts: just curious, was there a spec for this? | 14:27 |
| * jroll touching blueprint stuff | 14:27 | |
| sambetts | jroll: Yup merged a while abck | 14:27 |
| *** Marga_ has joined #openstack-ironic | 14:27 | |
| lucasagomes | BobBall, :-( yeah the code that should delete that session is here: https://github.com/openstack/ironic/blob/68ec9be6a0abe4946edf55fa8337e2c9854bc397/ironic/drivers/modules/deploy_utils.py#L795 | 14:27 |
| *** Marga_ has quit IRC | 14:28 | |
| *** Marga_ has joined #openstack-ironic | 14:28 | |
| BobBall | Curious. that's what I was trying to run manually - perhaps it's an iscsiadm bug then; the delete was trying to contact the node, but of course there was no server to contact, so it just timed out | 14:28 |
| jroll | sambetts: got it, thanks | 14:29 |
| BobBall | Didn't check the exit code though - if I can reproduce I'll check that to see if it's 0 or 21. | 14:29 |
| *** tsekiyama has joined #openstack-ironic | 14:31 | |
| sambetts | dtantsur, jroll: I very much hope not to have to do any backports | 14:31 |
| lucasagomes | BobBall, thanks, feel free to open a bug about it if you find any | 14:31 |
| BobBall | Yeah... I'll need much more info before I can open a bug that wouldn't be closed as incomplete :D | 14:32 |
| jroll | sambetts: maaaan, lack of pypi for this library makes me sad :( | 14:32 |
| *** mgoddard has joined #openstack-ironic | 14:32 | |
| *** mgoddard1 has quit IRC | 14:32 | |
| sambetts | jroll: Unfortunatly its not my department thats in charge of the library :( | 14:33 |
| jroll | yeah, I figured | 14:33 |
| sambetts | I need to send some emails, and pester some people I guess | 14:33 |
| sambetts | dtantsur: i'm looking at this patch https://review.openstack.org/#/c/225494, and if scroll to the bottom of this, even those its +1'd on the patch, it doesn't look right in the logs http://logs.openstack.org/94/225494/1/check/gate-ironic-inspector-dsvm/7879bb7/logs/devstacklog.txt.gz | 14:38 |
| sambetts | s/even those/even though/ | 14:38 |
| *** absubram has joined #openstack-ironic | 14:38 | |
| dtantsur | sambetts, you mean that ERROR in the end? I'm seeing it for some time already | 14:39 |
| dtantsur | but clearly devstack returned success, and our tests passed | 14:39 |
| sambetts | dtantsur: I can't see it running the exercise though, aren't there any logs for that? i see it starting inspector but then nothing else | 14:41 |
| dtantsur | sambetts, but that's interesting, it's probably am issue in our client | 14:41 |
| dtantsur | sambetts, it's another file | 14:41 |
| *** lazy_prince has quit IRC | 14:41 | |
| dtantsur | sambetts, http://logs.openstack.org/94/225494/1/check/gate-ironic-inspector-dsvm/7879bb7/console.html | 14:42 |
| *** absubram has quit IRC | 14:42 | |
| dtantsur | anyway, it does not look related to the patch itself | 14:43 |
| sambetts | dtantsur: Nope, just wanted to check :) | 14:44 |
| dtantsur | I'm looking into it now... | 14:44 |
| dtantsur | sambetts, the same problem with ironic jobs without inspector: http://logs.openstack.org/38/224938/2/check/gate-tempest-dsvm-ironic-pxe_ssh/15115ef/logs/devstacklog.txt.gz | 14:45 |
| dtantsur | so it does not seem to be our fault :) I'll have a look at the openstackclient bug tracker | 14:45 |
| sambetts | dtantsur: ah, yeah, I'm surprised that the stack is finishing successfully with an error like that | 14:46 |
| dtantsur | sambetts, reported: https://bugs.launchpad.net/python-openstackclient/+bug/1498038 | 14:48 |
| openstack | Launchpad bug 1498038 in python-openstackclient "openstack complete: ERROR: openstack 'module' object has no attribute 'UpdateDataSource'" [Undecided,New] | 14:48 |
| sambetts | Cool :) | 14:48 |
| dtantsur | thanks for noticing! | 14:48 |
| *** achanda has joined #openstack-ironic | 14:51 | |
| *** wshao has joined #openstack-ironic | 14:51 | |
| *** mgoddard has quit IRC | 14:53 | |
| *** xek has quit IRC | 14:53 | |
| *** mgoddard has joined #openstack-ironic | 14:54 | |
| openstackgerrit | Monty Taylor proposed openstack/python-oneviewclient: Change ignore-errors to ignore_errors https://review.openstack.org/225822 | 14:54 |
| lucasagomes | dtantsur, ur patch to use the retry values from the python-ironicclient didn't made it into nova yet? | 14:54 |
| dtantsur | lucasagomes, only one of them, IIRC | 14:54 |
| dtantsur | lucasagomes, https://review.openstack.org/#/c/211097/ is still up | 14:55 |
| lucasagomes | dtantsur, ack | 14:55 |
| lucasagomes | :-( yeah that was the one | 14:55 |
| lucasagomes | urgh | 14:55 |
| lucasagomes | dtantsur, I wonder if for that scheduler race, if we could make it fail fast would help somehow | 14:55 |
| devananda | g'morning, all | 14:56 |
| dtantsur | morning devananda! | 14:56 |
| lucasagomes | dtantsur, when we try to set the instance_uuid, if that wasn't retriable I think it would mitigate a bit the problem | 14:56 |
| lucasagomes | devananda, good morning | 14:56 |
| rloo | morning devananda | 14:56 |
| rloo | jroll: the enable_cleaning config doesn't make sense to me: https://review.openstack.org/#/c/220898/9 | 14:57 |
| rloo | jroll: should I just comment in the patch? | 14:57 |
| rloo | jroll: although the spec was approved with that config mentioned. | 14:57 |
| jroll | rloo: yeah, let's start in the patch I guess | 14:57 |
| jroll | devananda: mornin | 14:57 |
| rloo | jroll: ok, commenting. | 14:58 |
| sambetts | g'morning devananda | 14:58 |
| *** achanda has quit IRC | 15:01 | |
| *** devlaps has joined #openstack-ironic | 15:02 | |
| dtantsur | devananda, what's the status of https://bugs.launchpad.net/ironic/+bug/1427923 ? Are you working on it? | 15:03 |
| openstack | Launchpad bug 1427923 in Ironic "boot device API blocks while waiting on the BMC" [Medium,In progress] - Assigned to Devananda van der Veen (devananda) | 15:03 |
| devananda | ahh, no. let's unassign | 15:05 |
| dtantsur | thanks | 15:06 |
| jroll | sambetts: +A on https://review.openstack.org/#/c/219253/ | 15:07 |
| jroll | sambetts: can you post a follow up patch please? | 15:07 |
| *** thiagop has joined #openstack-ironic | 15:07 | |
| sambetts | jroll: :O thanks!! | 15:09 |
| jroll | :) | 15:09 |
| sambetts | Is gerrit being slow for anyone else? | 15:09 |
| jroll | it's always slow for me | 15:10 |
| dtantsur | sambetts, looks as usual for me | 15:10 |
| thiagop | good morning Ironic | 15:10 |
| jroll | \o thiagop | 15:10 |
| sambetts | dtantsur: Must have been my end, its sped up again now :/ | 15:10 |
| openstackgerrit | Merged openstack/ironic: Refactor agent {prepare,tear_down}_cleaning into deploy_utils https://review.openstack.org/220895 | 15:10 |
| sambetts | thiagop: o/ | 15:11 |
| lucasagomes | sambetts, w00t | 15:13 |
| sambetts | :D | 15:13 |
| lucasagomes | yeah jenkins is a bit slow today | 15:13 |
| dtantsur | JoshNang or jroll, if you have a minute, could you have a look at https://review.openstack.org/#/c/225102 please? I feel sorry for constantly pinging you, but switch to IPA is a bit harder than I expected :) | 15:14 |
| lucasagomes | well actually it seems grand in zuul | 15:14 |
| JayF | I'm core on IPA as well, fwiw :) | 15:18 |
| * JayF puts it in a tab | 15:18 | |
| dtantsur | JayF, oh man, sorry for forgetting you :) | 15:18 |
| JayF | dtantsur: would you be a fan of a patch that puts the hw info sent over from the agent | 15:18 |
| JayF | dtantsur: into node.extra or node.properties? | 15:19 |
| JayF | dtantsur: our original downstream agent patch did that and we have reasons to want to readd it | 15:19 |
| *** devlaps has quit IRC | 15:19 | |
| dtantsur | JayF, I'm ok with it, if it's optional. It essentially duplicates what inspector does, but not everyone uses inspector. | 15:19 |
| thiagop | Guys, if you have a slot, take a look on the OneView driver review: https://review.openstack.org/#/c/191822/ | 15:20 |
| thiagop | we did several improvements on it since the last time I asked you to look | 15:20 |
| dtantsur | JayF, not sure about optional, I think it won't hurt if we just do it always | 15:20 |
| thiagop | seems a LOT mature for us now | 15:20 |
| JayF | dtantsur: any reason you didn't use lsblk -oVENDOR? | 15:21 |
| trown | JayF: dtantsur, what if we send too much to overflow those columns? | 15:21 |
| dtantsur | JayF, people told me it's not working on Ubuntu | 15:21 |
| trown | that was a problem when we stored stuff on node.extra before | 15:21 |
| dtantsur | JayF, or was it SERIAL? I don't remember.. | 15:21 |
| JayF | dtantsur: IDK, I just looked for Vendor | 15:21 |
| JayF | dtantsur: sounds like you thought of it at all, which is what I wanted to be sure of | 15:21 |
| dtantsur | JayF, the problem with changing lsblk is that, nothing is optional there. If it dfoes not support any of colummns, the whole call will fail | 15:22 |
| dtantsur | I'm open to do it all via lsblk, but it might break IPA+DIB+Ubuntu LTS | 15:23 |
| JayF | dtantsur: +2'd, just wanted to make sure you had thought of it | 15:23 |
| JayF | dtantsur: going to wait to land it to see if JoshNang or jroll wants a look | 15:23 |
| JoshNang | sure, i can take a look | 15:23 |
| dtantsur | thanks guys! | 15:24 |
| JoshNang | np! | 15:24 |
| *** devlaps has joined #openstack-ironic | 15:28 | |
| JoshNang | dtantsur: +A | 15:31 |
| dtantsur | \o/ | 15:31 |
| *** rameshg87 has joined #openstack-ironic | 15:35 | |
| rameshg87 | rloo: hi | 15:35 |
| rloo | hi | 15:35 |
| *** mtanino has joined #openstack-ironic | 15:35 | |
| rameshg87 | rloo: can we talk about https://review.openstack.org/#/c/220898/ now ? | 15:35 |
| rloo | sure | 15:35 |
| *** lazy_prince has joined #openstack-ironic | 15:35 | |
| rameshg87 | rloo: [conductor]clean_nodes=true is for enabling cleaning operations from all interfaces | 15:36 |
| rameshg87 | rloo: we have cleaning operations from management interface for ilo driver in upstream today | 15:36 |
| rloo | rameshg87: right | 15:36 |
| JayF | trown: in that case, Truncation++ would be a stupid simple first way to handle it | 15:36 |
| rameshg87 | rloo: someone using pxe_ilo driver would have enabled cleaning (or it would have got enabled automatically when they upgraded to kilo)\ | 15:37 |
| rloo | rameshg87: and those ilo cleaning operations don't care if dib ramdisk or ipa? | 15:37 |
| rameshg87 | rloo: no, they are out of band | 15:37 |
| rloo | rameshg87: is there any way to know, via code, whether a node is using dib ramdisk? | 15:37 |
| rameshg87 | rloo: unfortunately no | 15:38 |
| rameshg87 | rloo: if pass_deploy_info is invoked during deploy, then it's a dib ramdisk | 15:38 |
| trown | JayF: you guys do not use Swift? | 15:38 |
| rameshg87 | rloo: but there is no way to check if active instances used dib ramdisk or not | 15:38 |
| JayF | trown: We do have swift, yes, but I'm not sure how that applies to shipping node hardware information into ironic db? | 15:39 |
| trown | JayF: that is what we did in inspector | 15:39 |
| JayF | aha | 15:39 |
| trown | https://blueprints.launchpad.net/ironic-inspector/+spec/store-introspection-data | 15:39 |
| rloo | rameshg87: so IF we put in this new [iscsi_deploy] enable_cleaning=False and you want to delete it in M*, what will the code do? | 15:40 |
| rloo | rameshg87: what will the code do in M* when you delete that config? | 15:40 |
| rameshg87 | rloo: we would simple remove the if CONF.iscsi_deploy.enable_cleaning, then return statements | 15:40 |
| rameshg87 | rloo: the code will do what is supposed to do by default | 15:41 |
| rloo | rameshg87: so the default now is NOT to enable any deploy cleaning, even though conductor/clean_nodes=True | 15:41 |
| rameshg87 | rloo: for example, prepare_cleaning would start creating ports and boot up deploy ramdisk | 15:41 |
| rameshg87 | rloo: yes | 15:41 |
| rloo | rameshg87: but in M*, with conductor/clean_nodes=True, it will be enabled. | 15:41 |
| rameshg87 | rloo: only for drivers using iscsi based deploy | 15:41 |
| rloo | rameshg87: that isn't backwards compatible. | 15:42 |
| rameshg87 | rloo: why ? | 15:42 |
| rloo | rameshg87: what if we don't even merge the code now and wait for M*. then we won't need the config. | 15:42 |
| rameshg87 | rloo: we would end up not having feature for one cycle | 15:42 |
| *** Sukhdev has joined #openstack-ironic | 15:42 | |
| JayF | trown: in this case, I'm not sure inspector is a good fit b/c we want to track something across the fleet that changes | 15:43 |
| rameshg87 | rloo: which could have been enabled very easily | 15:43 |
| JayF | trown: imagine you have part model A in your fleet and you want to swap with part model B, we just wanted to store that info in Ironic so it's quickly queryable how many machines have the new | 15:43 |
| rloo | rameshg87: for users that have agent ramdisk, the default now would be no iscsi_deploy/cleaning, unless they change the config. | 15:43 |
| JayF | trown: and given each provisioned instance represents a custoemr, having it quickly cross-ref'd with instance uuid is helpful | 15:43 |
| rloo | rameshg87: for users that have agent ramdisk, in M*, they would get iscsi_deploy/cleaning ? | 15:43 |
| rameshg87 | rloo: in M*, please using iscsi based drivers [pxe_ilo, pxe_ipmitool, etc] would get cleaning enabled by default | 15:44 |
| rameshg87 | s/please/people | 15:45 |
| trown | JayF: that makes sense, I think eventually we would want a nice way to use data from inspector to do that type of query | 15:45 |
| dtantsur | sambetts, when you have a minute, could you please review https://review.openstack.org/#/c/225173 ? that's our last feature before we call all BP's done | 15:45 |
| trown | JayF: but we are not there yet | 15:45 |
| rameshg87 | rloo: for now, we would disable this by default as we want to make sure we don't break users | 15:45 |
| rloo | rameshg87: but they wouldn't get that by default in L*. SO isn't that backwards incompatible? | 15:45 |
| rameshg87 | rloo: they would get it if they are interested in using it | 15:46 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Add alembic migrations for the inspector database https://review.openstack.org/223665 | 15:46 |
| rloo | rameshg87: I feel like I'm missing something. | 15:46 |
| rloo | rameshg87: but what if they are not interested in using it now. | 15:46 |
| dtantsur | going afk until the meeting | 15:46 |
| rloo | rameshg87: or using it in M* | 15:46 |
| *** dtantsur is now known as dtantsur|afk | 15:46 | |
| rameshg87 | rloo: they can go and change the CONF variable if they are interested in it now | 15:46 |
| *** garthb has joined #openstack-ironic | 15:47 | |
| rloo | rameshg87: but what if they are NOT interested in using it now, in L* and NOT interested in using it in M* | 15:47 |
| rameshg87 | rloo: in M*, they would get it anyway like rest of agent_* drivers | 15:47 |
| jroll | rloo: rameshg87 should we take this to the meeting instead? | 15:47 |
| rameshg87 | jroll: sure, but meeting is time capped :) | 15:47 |
| rameshg87 | jroll: you can join in if you are interested | 15:48 |
| rloo | jroll: we could, but now or then, I still don't understand yet. | 15:48 |
| jroll | rameshg87: O' | 15:48 |
| jroll | rameshg87: I'm listening, at least :) | 15:48 |
| jroll | feel free to continue, just a thought | 15:48 |
| rameshg87 | rloo: but, honestly I don't get your concern alternately | 15:48 |
| rameshg87 | rloo: let me see if I get it right | 15:49 |
| rloo | rameshg87: so you're saying that if users are interested in isci-deploy-cleaning in L*, they have to set the config. If they do not want that cleaning, they don't have to set any config. | 15:49 |
| rameshg87 | rloo: yes | 15:49 |
| rloo | rameshg87: and in M*, if they are not interested, too bad, they will get cleaning. | 15:49 |
| rameshg87 | rloo: we did the same for agent_* drivers | 15:49 |
| rameshg87 | rloo: we enabled cleaning by default for everyone using agent_* drivers | 15:49 |
| rameshg87 | rloo: the CONF variable directly set clean_nodes to true | 15:50 |
| rameshg87 | I mean we did the same in kilo | 15:50 |
| rloo | rameshg87: Didn't we set conductor/clean_nodes=True, but all clean steps have a config with priority=0? | 15:51 |
| rameshg87 | rloo: no | 15:51 |
| rameshg87 | or did we ? :) | 15:51 |
| * rameshg87 checks | 15:51 | |
| JayF | We set clean_nodes=true | 15:51 |
| JayF | and only step was erase_devices for agent driver | 15:51 |
| JayF | iirc | 15:51 |
| JayF | Today in gate, agent runs w/erase_devices as only cleaning step <-- I know that's accurate | 15:52 |
| JayF | except we override that config on kernel command line b/c believe it or not, bad idea to shred in devstack :) | 15:52 |
| rloo | rameshg87: I see ilo has clean_priority_erase_devices=<None>. I | 15:52 |
| rameshg87 | https://github.com/openstack/ironic/blob/master/etc/ironic/ironic.conf.sample#L811-L821 | 15:52 |
| rloo | rameshg87: the help sez that default is 10 if unset. I wonder why it wasn't just set to 10 instead of None. | 15:52 |
| rameshg87 | JayF: but didn't the erase device happen automatically for people upgrading from juno to kilo ? | 15:53 |
| rameshg87 | any idea ? | 15:53 |
| JayF | rameshg87: yes, it did | 15:53 |
| JayF | rameshg87: I'm 99% sure of it | 15:53 |
| *** mgoddard has quit IRC | 15:53 | |
| JayF | rameshg87: was a pretty big moment for those of us working on OnMetal when it was enabled by default, tbh | 15:53 |
| JayF | rameshg87: given our first interaction w/the project was talking deva into how adding some form of "decommissioning" was a godo idea :) | 15:54 |
| rameshg87 | JayF: okay | 15:54 |
| * lucasagomes remembers that; and the summit talk too | 15:54 | |
| *** mgoddard has joined #openstack-ironic | 15:54 | |
| rameshg87 | rloo: so we did this in kilo iiuc | 15:54 |
| rloo | rameshg87, JayF: maybe enabling that by default should have been with some api version bump. | 15:54 |
| *** romcheg has quit IRC | 15:55 | |
| rloo | rameshg87: to get back to your patch now. You aren't enabling it by default. | 15:55 |
| rloo | rameshg87: and you want to enable it by default in M* | 15:55 |
| rloo | rameshg87: that's one of my issues. | 15:55 |
| rameshg87 | rloo: we could enable it straightaway now | 15:55 |
| rloo | rameshg87: how would you enable it straightaway? | 15:56 |
| *** rameshg871 has joined #openstack-ironic | 15:58 | |
| rameshg871 | rloo: I am sorry, got disconnected | 15:58 |
| *** rameshg87 has quit IRC | 15:58 | |
| *** rameshg871 is now known as rameshg87 | 15:58 | |
| rameshg87 | rloo: but the only issue is we would break the workflow of users using the dib ramdisk, which is bad in my opinion | 15:58 |
| rameshg87 | rloo: I would put it like | 15:58 |
| rameshg87 | rloo: we are providing the feature, but not enabling by default today because "we might break some users if we enabled it by default" | 15:59 |
| rameshg87 | rloo: we would enable it by default in M*, because "there will no longer be users whom we worry we would break" | 15:59 |
| jroll | this may be the worst suggestion I've ever made, but: what if we didn't add a config, booted the ramdisk every time, and only clean if it pings back to /lookup instead of wherever the bash ramdisk pings to, so we know it's an agent | 16:00 |
| rloo | rameshg87: let's say there was no issue with DIB/agent and we didn't need that config. how would a user turn off deploy/cleaning? | 16:00 |
| *** mitchjameson has joined #openstack-ironic | 16:00 | |
| rameshg87 | rloo: why would they need to disable only deploy/cleaning ? we don't give that luxury today for drivers that enable in-band cleaning today ? | 16:01 |
| *** wshao has quit IRC | 16:01 | |
| rameshg87 | rloo: if we want to do, probably it should be another work, right ? | 16:02 |
| rameshg87 | jroll: do you mean put a periodic task ? | 16:02 |
| rloo | jroll, rameshg87: is this a feature that we want in L*, given that feature freeze was last week? | 16:02 |
| rameshg87 | jroll: but we don't know if at all users switched to deploy_kernel/deploy_ramdisk | 16:02 |
| jroll | rameshg87: no, I mean always enable cleaning, then if the bash ramdisk vendor passthru gets hit after booting the ramdisk, then "finish" cleaning | 16:03 |
| rameshg87 | rloo: idk, we can take a decision | 16:03 |
| rameshg87 | jroll: but bash ramdisk might fail because we don't pass those other parameters like iscsi_target_iqn | 16:03 |
| jroll | rloo: yeah, that's my other question... though other than the config thing it doesn't seem risky | 16:03 |
| jroll | rameshg87: ugh | 16:03 |
| rameshg87 | jroll: or it might hit and report an error | 16:04 |
| rameshg87 | jroll: for booting up the bash ramdisk successfully we need all the parameters iirc | 16:04 |
| jroll | right, I didn't think about that | 16:05 |
| rameshg87 | rloo: how about taking a vote during the meeting ? | 16:05 |
| jroll | well, rloo has a good point there, I didn't think about the behavior change from L -> M | 16:06 |
| rloo | rameshg87: seems like two votes. 1. let a feature in after feature freeze; 2. the config issue wrt this particular feature | 16:06 |
| jroll | (somehow) | 16:06 |
| jroll | it seems like at this point, we should bump it | 16:06 |
| rameshg87 | rloo: and cancel second vote if 1st one is no ;-) | 16:06 |
| rameshg87 | finishes quickly | 16:06 |
| jroll | lucasagomes: for removing the bash ramdisk endpoints, you were going to do that in M or beginning of N? | 16:06 |
| rloo | rameshg87: yeah, 2nd vote isn't needed if 1st vote is no :) | 16:07 |
| rameshg87 | rloo: agreed | 16:07 |
| lucasagomes | jroll, I think we should do that in N now | 16:07 |
| lucasagomes | cause it took too long to get stuff merged | 16:07 |
| lucasagomes | and tripleo-ci is not yet using IPA | 16:07 |
| jroll | lucasagomes: yeah, so there's a TC proposal for deprecation period being "at least one release cycle and at least 3 months", so we could do it later in M, in theory | 16:08 |
| jroll | however I'm fine with N, this is a pretty big drop | 16:08 |
| lucasagomes | jroll, yeah that seems fair enough | 16:08 |
| lucasagomes | keeping it for 1 cycle more, deprecated, is not a big deal | 16:09 |
| jroll | yeah | 16:09 |
| jroll | rameshg87: rloo ^ more food for thought :P | 16:09 |
| * lucasagomes didn't read the scrollback | 16:09 | |
| rameshg87 | jroll: but N is still a longer period | 16:09 |
| rameshg87 | jroll: not enabling in-band cleaning for iscsi based drivers seems bad to me (unless we find a better way to do this without a CONF option) | 16:10 |
| JayF | jroll: -1 I really dislike that idea (reading back from 9am) | 16:10 |
| JayF | jroll: Basically if cleaning is on, you never want a case where it can "complete" without doing anything | 16:11 |
| JayF | jroll: without that being the default or explicitly the behavior the deployer requested | 16:11 |
| JayF | i.e. I would not want someone with BASH ramdisk thinking that cleaning is doing anything on their machine, which I think that would imply | 16:11 |
| JayF | jroll: re: auto-disable cleaning for bash ramdisk | 16:11 |
| rameshg87 | JayF: so are you saying you don't want to auto detect it's a bash ramdisk and silently come out ? | 16:12 |
| lucasagomes | which patch is that ? | 16:12 |
| rameshg87 | lucasagomes: there isn't a patch for that, it was only a proposal | 16:12 |
| JayF | rameshg87: exactly. How would an operator know, looking at Ironic only, if a machine was actually clean or if it was skipped due to the ramdisk | 16:13 |
| jroll | JayF: I mean, I tend to agree, it needs thought | 16:14 |
| jroll | just a start to a better proposal, idk | 16:14 |
| lucasagomes | right yeah we currently have no way to identify whether it's the IPA or bash ramdisk :-( | 16:16 |
| lucasagomes | the way it was architect was to be a drop in replacement (so no extra config to differentiate) | 16:16 |
| lucasagomes | looks like we want to have a way to tell which ramdisk the node is using | 16:16 |
| rloo | lucasagomes: the only way to know, is via the vendor passthrus? | 16:16 |
| lucasagomes | rloo, yeah, dependeing on which endpoint the ramdisk calls | 16:17 |
| lucasagomes | but cleaning also happens prior to the deployment so... | 16:17 |
| rameshg87 | rloo: just added to the agent | 16:18 |
| rameshg87 | I mean agenda :) | 16:18 |
| * rameshg87 has agent in the tip of the tongue | 16:18 | |
| rameshg87 | rloo: will be back for the meeting | 16:18 |
| *** rameshg87 is now known as rameshg87-afk | 16:18 | |
| jroll | lucasagomes: do you have an example of one of those hacks to deal with wsme 0.8 weirdness? | 16:19 |
| lucasagomes | jroll, lol hmm well the problem that was fix in wsme was that | 16:21 |
| jroll | lucasagomes: specifically, trying to deal with the test failures here https://review.openstack.org/#/c/206232/ | 16:21 |
| jroll | where a field is in the object but not the API | 16:21 |
| lucasagomes | if you try to do for e.g a POST with some fields that doesn't exist in a resource it will fail now | 16:21 |
| * lucasagomes looks | 16:21 | |
| *** ukalifon has quit IRC | 16:21 | |
| lucasagomes | jroll, right and who populates that field? | 16:22 |
| lucasagomes | jroll, is the API? The way we do for things like that (e.g node_uuid to node_id) is to have the API layer to convert from one to another | 16:22 |
| jroll | lucasagomes: there are later patches adding them to the objects and api layer | 16:23 |
| jroll | just split for ease of review | 16:23 |
| openstackgerrit | Merged openstack/ironic-python-agent: Expose serial, wwn and vendor on the BlockDevice object https://review.openstack.org/225102 | 16:24 |
| lucasagomes | jroll, right hmm so we also should split that test to not POST those fields if they do not exist | 16:24 |
| lucasagomes | we can append the fields to the dict in the patch that is actually adding it to the API | 16:24 |
| jroll | lucasagomes: nothing changed the API or API tests, that's the thing | 16:25 |
| jroll | nothing is posting those fields | 16:25 |
| *** ifarkas has quit IRC | 16:25 | |
| jroll | lucasagomes: even existing, unchanged tests are failing | 16:25 |
| lucasagomes | jroll, https://github.com/openstack/ironic/blob/master/ironic/tests/api/utils.py#L93 | 16:26 |
| lucasagomes | I think you may be looking for it ^ | 16:26 |
| jroll | lucasagomes: OH | 16:26 |
| jroll | yes, perfect, ty | 16:26 |
| lucasagomes | because in the api.utils.py we use the same dict as the db.utils | 16:26 |
| jroll | yep | 16:26 |
| lucasagomes | I know right? Don't be sad :-( | 16:27 |
| *** siva has joined #openstack-ironic | 16:27 | |
| *** siva is now known as Guest25749 | 16:28 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Refactor deploy_utils methods https://review.openstack.org/222264 | 16:28 |
| jroll | lucasagomes: lol, it isn't that terrible | 16:29 |
| lucasagomes | jroll, well it's a bit confusing tho | 16:29 |
| jroll | sure | 16:30 |
| lucasagomes | s/bit/lot | 16:30 |
| lucasagomes | I'm too optimistic these days, vacation is coming | 16:30 |
| jroll | lol | 16:31 |
| thiagop | lucasagomes: o/ | 16:34 |
| lucasagomes | thiagop, hi there! | 16:34 |
| thiagop | lucasagomes: coming to Brazil? | 16:35 |
| * thiagop wonders if coming to Brazil and going to Japan right after is a good idea... | 16:35 | |
| lucasagomes | thiagop, not yet, I'm going to iceland this time. But I will be in Brazil after the summit, in december | 16:35 |
| * thiagop is jetlagged just of thinking about it | 16:35 | |
| lucasagomes | thiagop, yeah, it's totally the other way around | 16:36 |
| lucasagomes | thiagop, going to bauru! man, I was looking at the weather there and it was suppose to be winter right now, but it's 35C | 16:36 |
| lucasagomes | dahell | 16:36 |
| lucasagomes | https://www.google.ie/search?q=bauru+weather&ie=utf-8&oe=utf-8&gws_rd=cr&ei=JTIAVoaKNYGngwTOhqbICQ | 16:37 |
| thiagop | lucasagomes: I was in São Paulo last week. 10-13° | 16:37 |
| thiagop | the weather for these lands is crazy latelly | 16:37 |
| lucasagomes | yeah são paulo is usually colder, probably because the sun doesn't reach the houses there due the polution and/or buildings around (jk) | 16:37 |
| lucasagomes | thiagop, indeed | 16:38 |
| thiagop | lol | 16:38 |
| *** romainh has left #openstack-ironic | 16:38 | |
| thiagop | lucasagomes: I had a worse impression of SP than I have now after spending a week there | 16:38 |
| lucasagomes | thiagop, oh yeah, SP is complicated. The thing I don't like there most is the commuting | 16:40 |
| * jroll wants to go to brazil and lay on the beach | 16:40 | |
| lucasagomes | jroll, ++ | 16:40 |
| thiagop | jroll: consider going to Pipa beach, brazillian northeast | 16:41 |
| jroll | thiagop: thanks for the recommendation, but I'll forget by the time I actually try to go | 16:41 |
| *** e0ne has quit IRC | 16:41 | |
| thiagop | jroll: try not to get hypnotized by the pics on google | 16:41 |
| jroll | lol | 16:42 |
| jroll | oh my, yes please | 16:42 |
| lucasagomes | lol | 16:42 |
| * lucasagomes always wanted to go to the lençois maranhenses thing | 16:42 | |
| lucasagomes | it's pretty awesome: https://www.google.ie/search?q=len%C3%A7ois+maranhenses&source=lnms&tbm=isch&sa=X&ved=0CAcQ_AUoAWoVChMIjY_Pk8mIyAIVBdGACh2pEwTp | 16:43 |
| thiagop | lucasagomes: commuting? | 16:43 |
| *** stendulker has joined #openstack-ironic | 16:43 | |
| lucasagomes | thiagop, yeah in SP it takes too long to go from one place to another | 16:43 |
| lucasagomes | even by metro/train | 16:43 |
| * lucasagomes is used to the country side | 16:43 | |
| thiagop | thiagop: aw, yeah. Things are very far away from one another | 16:44 |
| thiagop | lucasagomes: ^ | 16:44 |
| thiagop | lençois are awesome. I have a friend who walked that whole thing by foot | 16:45 |
| lucasagomes | thiagop, really? damn I thought it was very long | 16:46 |
| thiagop | It is indeed. Sort of 5-6 days trip. And one on the group had a severe heatstroke. But, you know... we backpackers are crazy. | 16:49 |
| thiagop | (and also well prepared to these kinds of risks) | 16:50 |
| *** chlong has quit IRC | 16:50 | |
| lucasagomes | lol | 16:50 |
| lucasagomes | thiagop, nice one! | 16:50 |
| lucasagomes | yeah I hope next year I'm going to plan some trips in brazil | 16:50 |
| lucasagomes | for the end of the year I will mostly just go see family/friends and I was have a weeding to attend | 16:51 |
| thiagop | lucasagomes: I myself like woods and mountains better than desert, but | 16:51 |
| lucasagomes | thiagop, same here | 16:51 |
| thiagop | meeting in 9min, right? | 16:51 |
| jroll | 8 | 16:52 |
| jroll | :P | 16:52 |
| trown | 7 | 16:53 |
| thiagop | 6 | 16:54 |
| *** dtantsur|afk is now known as dtantsur | 16:54 | |
| *** derekh has quit IRC | 16:55 | |
| *** ijw has joined #openstack-ironic | 16:56 | |
| *** xek has joined #openstack-ironic | 16:58 | |
| lucasagomes | oh now! | 17:00 |
| lucasagomes | JoshNang, :-( | 17:00 |
| *** rameshg87-afk is now known as rameshg87 | 17:00 | |
| lucasagomes | devananda, NobodyCam ? | 17:01 |
| JoshNang | :( | 17:01 |
| devananda | meeting time :) | 17:01 |
| lucasagomes | JoshNang, sad to hear you go but I wish you all the best on ur new job! | 17:01 |
| * lucasagomes replied the ML | 17:01 | |
| JoshNang | lucasagomes: thanks! | 17:01 |
| lucasagomes | JoshNang, wish I could -2 the decision tho, jk :-) | 17:02 |
| rameshg87 | JoshNang: all the best for your new job | 17:02 |
| *** mitchjameson has quit IRC | 17:02 | |
| *** chlong has joined #openstack-ironic | 17:02 | |
| *** bizarrochristy has joined #openstack-ironic | 17:02 | |
| *** bizarrochristy has quit IRC | 17:03 | |
| JoshNang | rameshg87: thanks :) | 17:03 |
| JoshNang | lucasagomes: heh | 17:03 |
| jlvillal | lekha: Any interest in being a part of the testing/quality team for Ironic? :) | 17:06 |
| lekha | jlvillal: sure! | 17:06 |
| dtantsur | don't surrender so quickly :D | 17:07 |
| lekha | :) | 17:08 |
| jlvillal | lekha: Woo hoo :) I'll add your name on the white board. | 17:08 |
| jlvillal | lekha: https://etherpad.openstack.org/p/IronicWhiteBoard | 17:08 |
| *** rameshg871 has joined #openstack-ironic | 17:09 | |
| * JoshNang runs to an eye appointment, thanks for all the kind words y'all :) | 17:11 | |
| *** devlaps has quit IRC | 17:12 | |
| krtaylor | jlvillal, me too :) | 17:12 |
| *** rameshg87 has quit IRC | 17:12 | |
| krtaylor | jlvillal, I should have more time in the next few weeks | 17:12 |
| *** Sukhdev has quit IRC | 17:12 | |
| *** rameshg871 is now known as rameshg87 | 17:13 | |
| jlvillal | krtaylor: Great :) | 17:14 |
| *** lazy_prince has quit IRC | 17:15 | |
| *** Sukhdev has joined #openstack-ironic | 17:16 | |
| *** Sukhdev has quit IRC | 17:16 | |
| dtantsur | dear launchpad, timeout error in 5 seconds does not sound sane >_< | 17:20 |
| *** achanda has joined #openstack-ironic | 17:20 | |
| jroll | trown: I'm going to have to make a launchpad and such for IPA, too, and a pypi thing | 17:22 |
| jroll | or not pypi apparently | 17:22 |
| trown | jroll: do you think we need a launchpad? where are current IPA bugs filed? | 17:23 |
| *** davideagnello has joined #openstack-ironic | 17:23 | |
| dtantsur | in ironic itself | 17:23 |
| jroll | trown: to use openstack/releases we need a launchpad | 17:23 |
| lekha | jlvillal: Thanks for adding that :) | 17:23 |
| trown | ah, for tracking what went in to the release | 17:24 |
| jlvillal | lekha: Thanks for the patch! | 17:24 |
| *** e0ne has joined #openstack-ironic | 17:24 | |
| jroll | yeah | 17:24 |
| lucasagomes | jroll, can we still use the bug tracker in ironic? | 17:24 |
| lucasagomes | or we should start using the one for ipa? (I'm more happy just using ironic to be honest) | 17:24 |
| jlvillal | lekha: If you didn't know, we have an ironic meeting over on #openstack-meeting-3 at this time. | 17:24 |
| jroll | lucasagomes: in theory, yes, but then release emails become a pain | 17:24 |
| lucasagomes | right | 17:25 |
| lekha | jlvillal: ha thanks. Joining | 17:25 |
| lucasagomes | yeah we can see how it works | 17:25 |
| dansmith | jroll: xek: One more plea on that review to implement the new class_action method, and a link to how we did it preemptively in nova | 17:25 |
| jroll | dansmith: nice, thanks | 17:25 |
| dansmith | jroll: xek: I promise it will make M easier, but if you decide to move on with what's here, that's fine | 17:25 |
| jroll | dansmith: I mean, I'm clearly the expert and you're the peanut gallery here, but I'll take your feedback into consideration :) | 17:26 |
| * jroll changes vote | 17:27 | |
| jroll | thanks again dansmith | 17:27 |
| dansmith | jroll: heh | 17:27 |
| dansmith | basically, the old way of doing things is overly hard and complex and if you do it that way, I'm worried you'll think less of me | 17:28 |
| dansmith | ...once you start to feel that pain | 17:28 |
| jroll | nod | 17:29 |
| openstackgerrit | Sam Betts proposed openstack/ironic: Add Cisco IMC PXE Driver https://review.openstack.org/219253 | 17:30 |
| openstackgerrit | Sam Betts proposed openstack/ironic: Clean up CIMC driver docs and comments https://review.openstack.org/225958 | 17:30 |
| dtantsur | sambetts, not sure why you rebased it ^^ | 17:31 |
| jroll | -.- | 17:32 |
| sambetts | dtantsur: -_- it wasn't meant to rebase that one... | 17:32 |
| jroll | I'm going to re-approve that | 17:32 |
| xek | dansmith, jroll, I'll get right to it in a couple of hours :) I'm changing offices so my dev workstation is on the move right now... | 17:32 |
| dansmith | xek: cool, I'll be glad to throw my +1 on it quickly after you do if you ping me.. I think otherwise it's cool | 17:32 |
| dtantsur | sambetts, don't touch it after reapproval and before it merges :D | 17:33 |
| sambetts | dtantsur: yup, no more touching, I thought it had made it through the gate already so I rebased against master :'( | 17:34 |
| *** rameshg871 has joined #openstack-ironic | 17:38 | |
| dhellmann | devananda: double checking that ironic-lib 0.1.0 is what you want for stable/liberty | 17:41 |
| devananda | dhellmann: we were just discussing that in meeting | 17:41 |
| *** rameshg87 has quit IRC | 17:41 | |
| devananda | it looks like we couldn't switch to actually use it for L, because dep freeze | 17:41 |
| devananda | dtantsur: ^^ can you confirm the version? | 17:42 |
| dhellmann | devananda: so, you don't care about a stable branch? I can drop it from my list if that's the case. | 17:42 |
| dtantsur | dhellmann, devananda, correct, we're not using ironic-lib in liberty at all | 17:42 |
| devananda | yea, no need for stable branch for that. we'll switch projects to use what ever version is currently released when M opens | 17:43 |
| dhellmann | cool, it's easy enough to drop it for now | 17:43 |
| *** praneshp has joined #openstack-ironic | 17:45 | |
| *** harlowja has joined #openstack-ironic | 17:46 | |
| *** vdrok_ has joined #openstack-ironic | 17:49 | |
| *** mgoddard has quit IRC | 17:51 | |
| *** baoli has quit IRC | 17:59 | |
| *** pelix has quit IRC | 18:00 | |
| *** baoli has joined #openstack-ironic | 18:01 | |
| jroll | devananda: infra resources are precious, yes, but we have to get full tempest running somehow :/ | 18:01 |
| devananda | jroll: curious, how does sdague feel about that job? | 18:01 |
| vdrok_ | jroll: devananda and there will be a problem on how to know how many baremetal nodes are available, as some tests require 3, and they may run un parallel | 18:02 |
| vdrok_ | so i'll try to skip tests for now | 18:02 |
| lucasagomes | rloo, mind if I update the config for #224938 ? | 18:02 |
| dtantsur | see you tomorrow | 18:03 |
| devananda | jroll: some of the integration tests (ironic + nova) that I've wanted were, well, not useful in his opinion. I have not grok'd his reasoning but am curious if you have | 18:03 |
| rameshg871 | lucasagomes: did it come correctly for you ? | 18:03 |
| *** dtantsur is now known as dtantsur|afk | 18:03 | |
| lucasagomes | rameshg871, will check it in a min | 18:03 |
| * rameshg871 wonders if ordering is decided based on TZ | 18:03 | |
| devananda | lol | 18:04 |
| rloo | lucasagomes: I'm fine as long as you can still +2 it. Otherwise no :) | 18:04 |
| lucasagomes | rloo, I will stay a bit longer to review some of the bugs | 18:04 |
| jroll | devananda: sdague feels we need to get full tempest running | 18:04 |
| lucasagomes | rloo, and that exception | 18:04 |
| jroll | devananda: and afaik has been asking for it for a long time | 18:05 |
| rloo | thx lucasagomes, although the whole idea was not to burn ourselves out at release time :-) | 18:05 |
| *** ukalifon1 has joined #openstack-ironic | 18:05 | |
| lucasagomes | rloo, indeed, but it's all good | 18:06 |
| lucasagomes | we should get these things done | 18:06 |
| rloo | lucasagomes: :) | 18:06 |
| * rameshg871 leaves for the day | 18:06 | |
| rameshg871 | see you tomorrow folks | 18:06 |
| jlvillal | rameshg871: Ciao | 18:06 |
| *** rameshg871 has quit IRC | 18:06 | |
| devananda | jroll: "full tempest" is a strange phrase since we know that there are many nova features that our driver doesn't support | 18:07 |
| devananda | like snapshots and cinder volumes | 18:07 |
| jroll | devananda: sure, you know what I really mean | 18:07 |
| devananda | jroll: but I dont know what sdague means | 18:07 |
| jroll | then talk to him? | 18:08 |
| devananda | eh, sure | 18:08 |
| devananda | I mean, I have, several times over the last year. it just sounded like you had more recent infor | 18:08 |
| jroll | well | 18:08 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Agent Inband RAID configuration available in cleaning https://review.openstack.org/224938 | 18:08 |
| lucasagomes | rloo, ^ | 18:08 |
| jroll | devananda: at the nova midcycle, people asked when we're going to do it because they've been asking for this for a year | 18:09 |
| jroll | devananda: so I said I'll do my best to do the thing, and mtreinish volunteered to help where he could | 18:09 |
| lucasagomes | that's what I get generating with a fresh tox env | 18:09 |
| rloo | thx lucasagomes, i should look anyway, cuz there were changes since i first proposed that patch :) | 18:09 |
| jroll | anyway, bbiaf | 18:10 |
| devananda | jroll: cool. adam_g was working on that until, well, he wasn't. iirc, it got mired in "how do we disable X, Y, and Z feature tests" along with "how long is this really going to take? OMG 4 HRS!" | 18:11 |
| lucasagomes | rloo, the only thing I worry is that by having that set to 0 now we should keep it like that on future releases | 18:12 |
| lucasagomes | it's not something we want to change all the time | 18:12 |
| lucasagomes | do we plan to have that raid stuff disabled by default in the future too? | 18:12 |
| rloo | lucasagomes: i have no idea. this is ramesh's baby. i am just trying to help. | 18:12 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor iscsi_ilo driver to use new boot interface https://review.openstack.org/216538 | 18:13 |
| rloo | lucasagomes: it depends on whether you want to do RAID during cleaning, or just during 'zapping'. | 18:13 |
| rloo | lucasagomes: and originally, it was just 'zapping' -- so 0 made sense. | 18:13 |
| rloo | lucasagomes: if that is what/how we think operators will use raid, then it is fine. but again, i don't know. | 18:13 |
| lucasagomes | right, I bet it depends on the raid too, if it's soft raid it totally needs to be enabled by default | 18:14 |
| lucasagomes | hardware raid, that's questinoble, for cleaning I think it would makes more sense to disable it | 18:14 |
| * lucasagomes don't know the definitive answer really | 18:14 | |
| *** harshs has joined #openstack-ironic | 18:14 | |
| *** trown is now known as trown|lunch | 18:15 | |
| *** tsekiyam_ has joined #openstack-ironic | 18:15 | |
| rloo | lucasagomes: alternatively, there are other configs for clean steps related to IPA, that use <None> as the value, which means take the value from the hw manager. | 18:15 |
| *** linggao_ has joined #openstack-ironic | 18:15 | |
| lucasagomes | rloo, yeah, the erease device right? | 18:15 |
| lucasagomes | if not specified it defaults to 10 | 18:15 |
| rloo | lucasagomes: yup, that one. | 18:15 |
| lucasagomes | (or something like that_ | 18:15 |
| lucasagomes | )* | 18:15 |
| lucasagomes | hmmmmmmmmmm | 18:15 |
| rloo | lucasagomes: maybe worth commenting/asking. | 18:16 |
| lucasagomes | rloo, you know, I wonder if someone will have hardware and software raid in their setup | 18:16 |
| lucasagomes | and this option to enable is global | 18:16 |
| lucasagomes | rloo, yeah i will | 18:17 |
| lucasagomes | because one must rebuild the soft raid on cleaning, but the hardware one, that's not desirable for some people | 18:17 |
| * lucasagomes comments | 18:17 | |
| *** mgoddard has joined #openstack-ironic | 18:18 | |
| *** tsekiyama has quit IRC | 18:18 | |
| *** linggao has quit IRC | 18:18 | |
| devananda | lucasagomes: wy not? | 18:19 |
| *** vdrok_ has quit IRC | 18:19 | |
| lucasagomes | devananda, because some people may think that just ereasing the device is enough | 18:19 |
| devananda | ooh | 18:19 |
| lucasagomes | I don't think everyone agrees that you should remount the hardware raid all the time | 18:19 |
| lucasagomes | you can do it to clean individuals disks in the raid | 18:19 |
| devananda | because software raid contains some metadata which is only destroyed by rebuilding? | 18:19 |
| *** tsekiyam_ has quit IRC | 18:19 | |
| lucasagomes | but that's a bit over concern... | 18:20 |
| lucasagomes | devananda, the software raid will be deleted once we clean the local disks right? So it should be rebuilt | 18:20 |
| devananda | lucasagomes: depends on how IPA is cleaning: zeroing the individual disks? mounting the logical volume and zeroing it? | 18:21 |
| lucasagomes | devananda, right, I don't think it does that right now | 18:21 |
| lucasagomes | JayF, ^ | 18:21 |
| lucasagomes | devananda, I think it simple zero all disk devices it found | 18:22 |
| JayF | lucasagomes: yes | 18:22 |
| JayF | The idea for in-band raid is that it had to be implemented as a cleaning step so you could inject it in cleaning post-secure-erase | 18:22 |
| JayF | to rebuild the raid each time | 18:22 |
| devananda | gotcha. so yes, you'll need to rebuild a software raid after that | 18:22 |
| devananda | but not a hardware raid | 18:22 |
| lucasagomes | yeah | 18:22 |
| lucasagomes | devananda, and adding a global config for that seems a bit odd no? | 18:23 |
| *** harshs has quit IRC | 18:23 | |
| lucasagomes | because that will affects drivers using hardware raid and drivers using software raid | 18:23 |
| devananda | lucasagomes: agreed | 18:23 |
| *** cdearborn-mobile has joined #openstack-ironic | 18:25 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Refactor agent_ilo driver to use new boot interface https://review.openstack.org/217102 | 18:26 |
| *** cdearborn-mobile has quit IRC | 18:27 | |
| *** Sukhdev has joined #openstack-ironic | 18:27 | |
| *** harshs has joined #openstack-ironic | 18:30 | |
| *** stendulker has quit IRC | 18:32 | |
| lucasagomes | there are two bug fixes for this release that is going to conflict: https://review.openstack.org/#/c/205895 with https://review.openstack.org/#/c/201552/ | 18:35 |
| *** rbudden has joined #openstack-ironic | 18:36 | |
| rloo | lucasagomes: you mean the microversion bump? | 18:39 |
| lucasagomes | rloo, yeah | 18:39 |
| lucasagomes | rloo, I'm reviewing #205895, if it's good enough i will rebase the other one on top of it | 18:39 |
| rloo | lucasagomes: whichever gets approved first, wins. | 18:39 |
| lucasagomes | to avoid the conflict | 18:39 |
| lucasagomes | rloo, right, yeah but let's try make it less painful :-) | 18:39 |
| rloo | lucasagomes: well, one of them has to be changed. if you change before, that could limit things. | 18:40 |
| rloo | lucasagomes: what's wrong with changing after? | 18:40 |
| rloo | lucasagomes: but i guess it is your call :) | 18:40 |
| lucasagomes | rloo, that's a good a point, yeah let's see | 18:41 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/224614 | 18:55 |
| *** Marga_ has quit IRC | 19:00 | |
| rloo | lucasagomes: why is this a partial bug: https://review.openstack.org/#/c/201552/ | 19:01 |
| rloo | lucasagomes: is there more work to be done to close that bug? | 19:01 |
| lucasagomes | rloo, ahh because I submitted a new patch-set | 19:01 |
| lucasagomes | and previously I fixed it by editing directly in gerrit :-( | 19:01 |
| *** baoli has quit IRC | 19:02 | |
| *** gabriel-bezerra has quit IRC | 19:02 | |
| lucasagomes | rloo, my bad | 19:02 |
| rloo | lucasagomes: that didn't make any sense to me. i must be tired of reviewing. | 19:03 |
| lucasagomes | rloo, oh I mean... jroll noticed that too and I've edited the commit message directly in gerrit changing it from "Partial" to "Closes" | 19:03 |
| rloo | lucasagomes: oh. ok, so it should be closes. thx :) | 19:04 |
| *** bradjones has quit IRC | 19:04 | |
| lucasagomes | rloo, but more reviews came and I edited the code in my machine and resubmitted the patch, but in my machine the commit message was still "Partial" | 19:04 |
| lucasagomes | rloo, yeah | 19:04 |
| lucasagomes | should be Closes | 19:04 |
| rloo | lucasagomes: ha ha. got it. | 19:05 |
| lucasagomes | rloo, should I update it? Or we can mark the bug as closes if that merges as partial? | 19:06 |
| lucasagomes | closed* | 19:07 |
| rloo | easy enough to update it. You'd only lose tan's +1 on it now. | 19:07 |
| lucasagomes | rloo, ack | 19:07 |
| rloo | lucasagomes: so no one else asks the same question :) | 19:07 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Allow abort for CLEANWAIT states https://review.openstack.org/201552 | 19:08 |
| lucasagomes | rloo, done | 19:08 |
| rloo | thx lucasagomes | 19:08 |
| lucasagomes | np | 19:09 |
| *** gabriel-bezerra has joined #openstack-ironic | 19:10 | |
| *** trown|lunch is now known as trown | 19:10 | |
| lucasagomes | ok I will leave the indirection api and the config driver one to review tomorrow morning | 19:10 |
| lucasagomes | the rest are reviewed | 19:10 |
| lucasagomes | I will call it a day folks, have a good night everyone | 19:10 |
| *** lucasagomes is now known as lucas-dinner | 19:12 | |
| sinval | lucasagomes: good night | 19:12 |
| *** ijw has quit IRC | 19:16 | |
| thiagop | night lucas-dinner | 19:20 |
| *** Sukhdev has quit IRC | 19:29 | |
| sambetts | night all o/ | 19:33 |
| betherly | I'm off night all! | 19:34 |
| *** xek has quit IRC | 19:34 | |
| thiagop | night sambetts betherly | 19:34 |
| *** Marga_ has joined #openstack-ironic | 19:35 | |
| openstackgerrit | Merged openstack/python-oneviewclient: Change ignore-errors to ignore_errors https://review.openstack.org/225822 | 19:39 |
| *** Marga_ has quit IRC | 19:46 | |
| *** Marga_ has joined #openstack-ironic | 19:46 | |
| *** ijw has joined #openstack-ironic | 19:51 | |
| *** baoli has joined #openstack-ironic | 19:52 | |
| *** baoli has quit IRC | 19:53 | |
| *** ukalifon1 has quit IRC | 19:53 | |
| *** baoli has joined #openstack-ironic | 19:53 | |
| *** xek has joined #openstack-ironic | 19:55 | |
| *** ijw has quit IRC | 19:56 | |
| *** e0ne has quit IRC | 19:59 | |
| *** e0ne has joined #openstack-ironic | 20:00 | |
| *** xek has quit IRC | 20:05 | |
| *** mgoddard has quit IRC | 20:19 | |
| *** baoli has quit IRC | 20:20 | |
| *** nicodemos has quit IRC | 20:22 | |
| rloo | jroll: wrt etherpad and 'RELEASE STATUS', are you OK if I delete most of it, and just put a note about 4.2.0 & link to https://launchpad.net/ironic/+milestone/4.2.0? | 20:22 |
| jroll | rloo: +1 | 20:23 |
| rloo | jroll: done | 20:26 |
| jroll | thanks | 20:27 |
| *** e0ne has quit IRC | 20:29 | |
| jroll | so uh | 20:30 |
| jroll | how do I run ironic-dbsync against sqlite | 20:30 |
| jroll | owait, I have a devstack already | 20:32 |
| rloo | jroll: i could be wrong, but i thought someone said that ironic-dbsync (upgrade) didn't work with sqlite. | 20:33 |
| jroll | rloo: it doesn't | 20:36 |
| jroll | I was hoping there was a "batch" sort of thing to get it to the newest without each step | 20:36 |
| jroll | because it's ALTER statements that break it | 20:36 |
| rloo | jroll: oh. submit a bug/wishlist for it :) | 20:37 |
| jroll | rloo: there's plenty of bugs filed against alembic already, I believe it's WONTFIX | 20:38 |
| rloo | jroll: haha. or sobsob. | 20:38 |
| *** ijw has joined #openstack-ironic | 20:38 | |
| jroll | rloo: "lolsob" | 20:38 |
| rloo | jroll: i can't even get that right! | 20:39 |
| jroll | hahaha | 20:39 |
| *** linggao_ has quit IRC | 20:41 | |
| * EmilienM | 20:49 | |
| *** harshs has quit IRC | 20:56 | |
| openstackgerrit | Merged openstack/ironic-specs: [ironic-specs] Update links to Cloud Admin Guide https://review.openstack.org/224940 | 20:57 |
| *** Sukhdev has joined #openstack-ironic | 21:02 | |
| *** harshs has joined #openstack-ironic | 21:03 | |
| *** ijw has quit IRC | 21:03 | |
| *** harshs has quit IRC | 21:03 | |
| *** trown is now known as trown|outttypeww | 21:05 | |
| *** harshs has joined #openstack-ironic | 21:05 | |
| jlvillal | jroll: https://review.openstack.org/#/c/166386/16/tempest/services/baremetal/v1/json/baremetal_client.py Am I right that line 142 is wrong? :( | 21:05 |
| jlvillal | I think it needs to be **kwargs | 21:06 |
| jroll | jlvillal: depends if that method takes a bunch=of keyword=args, or just a dict | 21:07 |
| * jlvillal doesn't know enough about the code | 21:07 | |
| jlvillal | Yeah... | 21:07 |
| jlvillal | I'll look more. | 21:07 |
| jlvillal | Thanks | 21:08 |
| jlvillal | jroll: Never mind, it takes a dict. | 21:10 |
| jroll | woo | 21:10 |
| *** r-daneel has quit IRC | 21:14 | |
| *** ijw has joined #openstack-ironic | 21:19 | |
| jroll | JayF: would you mind reviewing https://review.openstack.org/#/c/215097/ at some point? idk this stuff well | 21:21 |
| *** achanda has quit IRC | 21:24 | |
| *** bnemec has quit IRC | 21:27 | |
| mrda | Morning Ironic | 21:27 |
| jroll | hey mrda \o | 21:28 |
| rloo | morning mrda | 21:28 |
| rloo | jroll: I just took a look at https://review.openstack.org/#/c/215254/ | 21:28 |
| thiagop | morning mrda | 21:28 |
| rloo | jroll: is it possible that the agent never uses any memory? | 21:28 |
| jroll | rloo: no, why? | 21:28 |
| rloo | why would we set the default for the config to zero then | 21:29 |
| rloo | jroll: https://review.openstack.org/#/c/215254/14/ironic/drivers/modules/agent.py | 21:29 |
| jroll | rloo: 0 == skip the check | 21:29 |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Allow vendor drivers to acquire shared tasks https://review.openstack.org/226058 | 21:29 |
| jroll | er, wait | 21:29 |
| jroll | rloo: idk, I assume that's useful if you don't know how much it would take and don't want to risk over-estimating? | 21:30 |
| rloo | jroll: it seems odd to me. that 'memory_consumed_by_agent' is the size of the IPA ramdisk? | 21:30 |
| jroll | rloo: it's how much memory to reserve to run the agent ramdisk | 21:31 |
| jroll | as the config help says | 21:31 |
| jroll | 'The memory size in MiB consumed by agent when it is booted on a bare metal node.' | 21:31 |
| *** priteau has quit IRC | 21:31 | |
| rloo | jroll: and how is anyone going to know what that memory size is. | 21:32 |
| *** alexpilotti has quit IRC | 21:32 | |
| rloo | jroll: the help sez 'This may be set according to the memory consumed by the agent ramdisk image'. | 21:33 |
| jroll | rloo: boot one and run 'free -m' or something? | 21:33 |
| jroll | rloo: there's a number of ways to figure it out, and there's a number of things that number depends on that may be specific to a given deployment | 21:33 |
| rloo | jroll: hmm. well, zero seems like an odd default cuz i would suspect it will never be 0 unless the agent does nothing. | 21:34 |
| jroll | rloo: see the discussion here https://review.openstack.org/#/c/215254/9/ironic/drivers/modules/agent.py | 21:35 |
| rloo | jroll: sigh. so it doesn't disable the check as lucas had suggested. | 21:37 |
| jroll | rloo: it disables the agent's contribution to that check, and only strictly checks the image size | 21:37 |
| jroll | rloo: if that check fails with the config set to 0, the deploy will definitely fail | 21:37 |
| rloo | jroll: right. | 21:37 |
| rloo | jroll: so help string can be improved. | 21:38 |
| rloo | jroll: you ok if i change/suggest the config name to be something like 'agent_memory'? | 21:38 |
| jroll | rloo: well, the help string doesn't mention disabling it... | 21:39 |
| jroll | rloo: and changing config names makes me sad, but I might be fine with it if you have a compelling 'why' for changing the name | 21:39 |
| *** Sukhdev has quit IRC | 21:39 | |
| rloo | jroll: it doesn't disable the check -- the check is done, but doesn't factor in the memory used by the agent. | 21:39 |
| jroll | rloo: well, it does, it factors in the 0MB you told it to factor in | 21:40 |
| rloo | jroll: right, it factors in the 0MB, which (highly likely) doesn't reflect the actual memory used by the agent | 21:40 |
| jroll | rloo: sure, I guess, feel free to change the help message, I don't see much value in changing the config name though | 21:41 |
| *** bnemec has joined #openstack-ironic | 21:41 | |
| jroll | I need to walk my dog, I'll be back in a few | 21:42 |
| rloo | jroll: wrt the config name. 1. I think 'consumed' is too verbose, 'used' might be better. 2. I don't like the past tense used in the name | 21:42 |
| rloo | jroll: I'm not going to change anything right now. will think about it. just my first impression seeing the config name, I thought 'what?!!' | 21:42 |
| *** achanda has joined #openstack-ironic | 21:44 | |
| *** achanda has quit IRC | 21:44 | |
| *** achanda has joined #openstack-ironic | 21:44 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Added code coverage output directory to eslintignore https://review.openstack.org/215209 | 21:47 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Removed libs from eslintignore https://review.openstack.org/215214 | 21:47 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Added coverage threshold failure https://review.openstack.org/213816 | 21:47 |
| thiagop | I'm calling it a day | 21:47 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Removed X-Client header https://review.openstack.org/214406 | 21:47 |
| thiagop | good evening everyone | 21:47 |
| *** thiagop has quit IRC | 21:47 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Updated eslint rules. https://review.openstack.org/215795 | 21:48 |
| *** harshs has quit IRC | 21:48 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Removed gitignore files for previous build system https://review.openstack.org/215180 | 21:49 |
| *** Sukhdev has joined #openstack-ironic | 21:54 | |
| *** harshs has joined #openstack-ironic | 21:55 | |
| jroll | rloo: ok, yeah, I'm open to suggestions | 21:56 |
| *** Sukhdev has quit IRC | 21:56 | |
| *** praneshp has quit IRC | 21:57 | |
| *** Sukhdev has joined #openstack-ironic | 21:57 | |
| *** Sukhdev has quit IRC | 22:01 | |
| *** Sukhdev has joined #openstack-ironic | 22:02 | |
| *** cdearborn has quit IRC | 22:07 | |
| *** thrash is now known as thrash|g0ne | 22:14 | |
| *** rbudden has quit IRC | 22:24 | |
| *** romcheg has joined #openstack-ironic | 22:25 | |
| *** Guest25749 has quit IRC | 22:29 | |
| *** alexpilotti has joined #openstack-ironic | 22:34 | |
| *** dims__ has quit IRC | 22:37 | |
| *** praneshp has joined #openstack-ironic | 22:46 | |
| *** praneshp_ has joined #openstack-ironic | 22:49 | |
| *** praneshp has quit IRC | 22:51 | |
| *** praneshp_ is now known as praneshp | 22:51 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic-python-agent: Add more info to checksum exception https://review.openstack.org/217369 | 22:54 |
| *** derekh has joined #openstack-ironic | 23:00 | |
| devananda | jroll: boot driver interface split is still on the LP list. I don't think we talked about that one this morning | 23:01 |
| devananda | jroll: actually, is LP updated? | 23:02 |
| *** mtanino has quit IRC | 23:05 | |
| *** alexpilotti has quit IRC | 23:16 | |
| *** ijw has quit IRC | 23:16 | |
| *** dims has joined #openstack-ironic | 23:17 | |
| *** dims has quit IRC | 23:23 | |
| *** jamielennox|away is now known as jamielennox | 23:23 | |
| *** dims has joined #openstack-ironic | 23:23 | |
| *** ijw has joined #openstack-ironic | 23:23 | |
| openstackgerrit | Merged openstack/ironic: Fix final comments in RAID commits https://review.openstack.org/225590 | 23:29 |
| *** derekh has quit IRC | 23:32 | |
| *** Haomeng has joined #openstack-ironic | 23:36 | |
| *** dims has quit IRC | 23:39 | |
| *** Haomeng|2 has quit IRC | 23:39 | |
| *** dims has joined #openstack-ironic | 23:40 | |
| *** dims_ has joined #openstack-ironic | 23:43 | |
| *** dims has quit IRC | 23:45 | |
| *** lucas-dinner has quit IRC | 23:47 | |
| *** Sukhdev has quit IRC | 23:53 | |
| *** romcheg has quit IRC | 23:58 | |
| *** achanda_ has joined #openstack-ironic | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!