| mrda | Shrews: you're not here are you? | 00:00 |
|---|---|---|
| JoshNang | looks like it changed recently with the sync from oslo incubator: https://github.com/openstack/ironic/commit/9a8145cc22f9f19e04ac19c9eca522972dbc09e6 | 00:00 |
| NobodyCam | jroll: good question.. one sec | 00:00 |
| jroll | NobodyCam: never mind, ignore that :) | 00:00 |
| jroll | NobodyCam: I re-approved | 00:00 |
| NobodyCam | ack | 00:00 |
| NobodyCam | test_manager.py and manager.py | 00:00 |
| jroll | yeah | 00:01 |
| *** alexpilotti has quit IRC | 00:01 | |
| *** jrist has joined #openstack-ironic | 00:02 | |
| jroll | NobodyCam: or, I'm trying to. it's not going through | 00:02 |
| mrda | Has anyone here does alter_tables for alembic migrations? | 00:02 |
| NobodyCam | nope... ironic/drivers/modules/ilo/common.py and ironic/tests/conductor/test_manager.py | 00:02 |
| *** alexpilotti has joined #openstack-ironic | 00:02 | |
| jroll | mrda: I haven't. I'm thinking; just because we allow dots doesn't mean we necessarily have to allow that many characters, right? | 00:03 |
| jroll | I guess it's the right thing to do | 00:03 |
| rloo | thx JoshNang. I'll add a comment about that. | 00:03 |
| mrda | wll, if we want fqdns, we need longer than 63 | 00:03 |
| *** naohirot has joined #openstack-ironic | 00:03 | |
| mrda | and if more than 63 we might as well go out to the RFC length of 255 | 00:03 |
| *** alexpilotti has quit IRC | 00:03 | |
| mrda | s/wll/well/ | 00:04 |
| rloo | jroll, mrda: yeah, if we are going to increase the size of name, we should do it now/in kilo | 00:04 |
| jroll | yeah, I guess so | 00:04 |
| jroll | right | 00:04 |
| mrda | I have code, just not a migration script | 00:04 |
| jroll | ok, +a went through now | 00:04 |
| jroll | mrda: I thought alembic could generate migrations? | 00:04 |
| *** alexpilotti has joined #openstack-ironic | 00:04 | |
| jroll | http://docs.openstack.org/developer/ironic/cmds/ironic-dbsync.html#revision | 00:05 |
| mrda | I've got a running devstack, modified models.py for the length increase, and run the ironic-dysync script, but the migration file that gets generated is a noop | 00:05 |
| jroll | oh huh | 00:05 |
| mrda | so it doesn't auto-detect a length change | 00:05 |
| mrda | or it's pebkak from me | 00:05 |
| jroll | maybe you have to do the migration first and alembic picks up the difference from the db? | 00:05 |
| jroll | yeah, dunno | 00:06 |
| *** alexpilotti has quit IRC | 00:06 | |
| jroll | mrda: I think you want alter_column('nodes', 'name', type_=String(255)) or something | 00:07 |
| *** mtanino has quit IRC | 00:08 | |
| mrda | bingo | 00:08 |
| mrda | http://stackoverflow.com/questions/17174636/can-alembic-autogenerate-column-alterations | 00:08 |
| mrda | I won't make this change to our alembic env.py (yet :) | 00:08 |
| mrda | but doing it locally gives me the SQL for a update in column width | 00:09 |
| jroll | oh | 00:09 |
| jroll | no | 00:09 |
| jroll | do it! | 00:09 |
| mrda | op.alter_column('nodes', 'name', | 00:09 |
| mrda | existing_type=mysql.VARCHAR(length=63), | 00:09 |
| mrda | type_=sa.String(length=255), | 00:09 |
| mrda | existing_nullable=True) | 00:09 |
| mrda | jroll: nah, it adds in some other stuff for Boolean types that I'm unsure of | 00:09 |
| *** jamielennox is now known as jamielennox|away | 00:09 | |
| jroll | hm | 00:10 |
| mrda | I think I annoyed enough people already this week than to break alembic for ironic | 00:10 |
| *** hj-hp has quit IRC | 00:10 | |
| mrda | I might raise a bug though, so it can be done properly | 00:11 |
| jroll | +1 | 00:11 |
| * jroll wonders who mrda annoyed :P | 00:11 | |
| * BadCub needs to ponder dinner-y things | 00:14 | |
| *** hj-hp has joined #openstack-ironic | 00:15 | |
| *** achanda has quit IRC | 00:15 | |
| *** BadCub is now known as BadCub_Hiding | 00:16 | |
| jroll | NobodyCam: you still here? | 00:18 |
| *** romcheg has quit IRC | 00:18 | |
| NobodyCam | do I have to do something if I am? | 00:18 |
| NobodyCam | lol | 00:18 |
| *** absubram has joined #openstack-ironic | 00:18 | |
| jroll | lol | 00:18 |
| jroll | tests failed after that rebase | 00:18 |
| jroll | http://logs.openstack.org/15/157715/19/check/gate-ironic-python27/0f67281/console.html | 00:18 |
| jroll | missing inspect_ports key | 00:19 |
| *** jmccrory has quit IRC | 00:19 | |
| NobodyCam | I didnn't edit that file :( | 00:20 |
| NobodyCam | wasn't in conflict | 00:20 |
| jroll | I mean, I can fix too | 00:20 |
| jroll | doesn't matter as long as we get it done | 00:20 |
| *** kkoski has quit IRC | 00:21 | |
| *** kkoski has joined #openstack-ironic | 00:21 | |
| NobodyCam | one sec I */should/* still have hte branch | 00:21 |
| *** jamielennox|away is now known as jamielennox | 00:23 | |
| *** hj-hp has quit IRC | 00:25 | |
| *** ramineni has joined #openstack-ironic | 00:26 | |
| NobodyCam | was that lost somewhere. i'm looking at ilo/common now | 00:30 |
| NobodyCam | or is the test wrong. | 00:31 |
| jroll | NobodyCam: probably test is wrong | 00:33 |
| NobodyCam | thats what I'm thinking too.. one sec | 00:33 |
| ramineni | NobodyCam: ilo inspection also has some code changes in the same area , making the merge failed and test case failure :( | 00:33 |
| ramineni | NobodyCam: fixing it | 00:34 |
| openstackgerrit | Adam Gandelman proposed openstack/python-ironicclient: Add better framework for functionally testing verioned APIs https://review.openstack.org/165665 | 00:34 |
| NobodyCam | ramineni: awesome thank you... | 00:35 |
| NobodyCam | jroll: ^^^^^ | 00:35 |
| jroll | woot | 00:35 |
| adam_g | jroll, hoping that can get us toward testing the different API paths via client tests | 00:35 |
| adam_g | ^ | 00:35 |
| * NobodyCam runs away for dinner-y things | 00:35 | |
| jroll | adam_g: awesome, remind me to look again after k3 :P | 00:35 |
| adam_g | jroll, sure | 00:36 |
| *** absubram has quit IRC | 00:36 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: Fix is_hostname_safe for RFC compliance https://review.openstack.org/165666 | 00:40 |
| mrda | jroll: ^^^ | 00:40 |
| jroll | nice | 00:40 |
| *** Haomeng has joined #openstack-ironic | 00:40 | |
| *** Haomeng|2 has quit IRC | 00:43 | |
| *** Nisha has quit IRC | 00:43 | |
| mrda | devananda, dtantsur, BadCub: I'd like https://bugs.launchpad.net/ironic/+bug/1433832 to be targeted for K because I don't think it's wise to allow the Node name change (165443) that we introduced overnight to stay as is. I think it will create a mess that we'll just have to clean up later. | 00:45 |
| openstack | Launchpad bug 1433832 in Ironic "Fix is_hostname_safe for RFC compliance" [Undecided,In progress] - Assigned to Michael Davies (mrda) | 00:45 |
| mrda | ...fortunately there's a review up for it already :) | 00:46 |
| openstackgerrit | Michael Davies proposed openstack/ironic: Fix is_hostname_safe for RFC compliance https://review.openstack.org/165666 | 00:49 |
| *** jmccrory has joined #openstack-ironic | 00:50 | |
| *** jmccrory1 has joined #openstack-ironic | 00:52 | |
| *** jmccrory has quit IRC | 00:54 | |
| jroll | I see the bug in https://review.openstack.org/#/c/165528/1/ironic/conductor/manager.py | 00:57 |
| jroll | this is going to be fun. | 00:57 |
| *** jmccrory1 has quit IRC | 00:59 | |
| openstackgerrit | Merged openstack/ironic: Disable cleaning by default https://review.openstack.org/165241 | 00:59 |
| JoshNang | jroll: what's the bug? | 00:59 |
| jroll | this | 01:00 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 01:00 |
| jroll | I *think* | 01:00 |
| JoshNang | hmm? | 01:00 |
| jroll | actually I'm not so sure anymore | 01:01 |
| jroll | but see the spawn_after calls | 01:01 |
| jroll | or just the diff between patchsets | 01:01 |
| jroll | also surprised this doesn't need unit test updates | 01:01 |
| *** absubram has joined #openstack-ironic | 01:01 | |
| JoshNang | fwiw, i don't think we need to spawn after from _do_node_clean to _do_next_clean_step | 01:02 |
| jroll | I honestly have no idea, but it's a hunch | 01:02 |
| jroll | like, it got to the spawn_after call | 01:02 |
| jroll | and then just nothign | 01:02 |
| JoshNang | oh weird | 01:02 |
| jroll | could also be no free workers, but you'd think that would log somehow | 01:02 |
| jroll | see http://logs.openstack.org/28/165528/1/check/check-tempest-dsvm-ironic-pxe_ssh/532cf44/logs/screen-ir-cond.txt.gz#_2015-03-18_20_18_30_880 | 01:02 |
| JoshNang | hmmm | 01:03 |
| *** kkoski has quit IRC | 01:04 | |
| JoshNang | seems like a deadlock tbh | 01:04 |
| jroll | should probably set the spawn after hook thing too | 01:04 |
| jroll | actually that's almost certainly it | 01:05 |
| * jroll does stuff | 01:05 | |
| JoshNang | ohh good call | 01:05 |
| jroll | do we have a cleaning error handler? | 01:06 |
| JoshNang | yeah | 01:06 |
| JoshNang | i think its cleaning_error_handler | 01:06 |
| JoshNang | yup it is | 01:07 |
| *** kkoski has joined #openstack-ironic | 01:07 | |
| *** chenglch has joined #openstack-ironic | 01:07 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Add Cleaning Operations for iLO drivers https://review.openstack.org/157715 | 01:08 |
| ramineni | jroll, NobodyCam: ^^ | 01:09 |
| jroll | thanks ramineni | 01:09 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 01:09 |
| jroll | JoshNang: let's see how this goes | 01:09 |
| ramineni | jroll: I felt some wrong with https://review.openstack.org/#/c/157715/20/ironic/tests/drivers/ilo/test_power.py , so changed the code there | 01:09 |
| jroll | I feel like it's a real bug | 01:09 |
| jroll | but can't see it | 01:09 |
| ramineni | jroll: it should test for power properties , but testing for whole driver properties | 01:10 |
| jroll | ramineni: indeed! | 01:10 |
| jroll | good catch | 01:10 |
| jroll | ramineni: I wonder how that passed tests before :/ | 01:11 |
| jroll | anyway, +2 | 01:11 |
| ramineni | jroll: it is calling get_properties in base.py i suppose | 01:11 |
| ramineni | jroll: which is getting all properties , including inspect also | 01:12 |
| *** stendulker has joined #openstack-ironic | 01:12 | |
| jroll | hm | 01:13 |
| JoshNang | jroll: looks reasonable. | 01:13 |
| jroll | why wouldn't power.properties only return power properties? | 01:13 |
| ramineni | jroll: prev code was task.driver.get_properties | 01:14 |
| *** kkoski has quit IRC | 01:14 | |
| *** kkoski has joined #openstack-ironic | 01:14 | |
| ramineni | jroll: which should be actually task.driver.power.get_properties | 01:15 |
| JoshNang | jroll: ohhh i wonder if spawn after doesn't get called because we're not in a with... as task context manager? | 01:15 |
| jroll | ramineni: ah, I see, awesome :) | 01:15 |
| jroll | JoshNang: OH | 01:15 |
| jroll | urgh | 01:15 |
| *** achanda has joined #openstack-ironic | 01:15 | |
| jroll | JoshNang: the first one is | 01:15 |
| JoshNang | right | 01:16 |
| jroll | JoshNang: no, we are | 01:16 |
| jroll | line 1009 | 01:16 |
| rloo | ramineni: it seems like I commented on that test in some patch recently. thx for fixing it. | 01:16 |
| jroll | JoshNang: though that's a different thread. fml. | 01:16 |
| jroll | so we aren't. | 01:17 |
| jroll | JoshNang: maybe that should just use _do_node_clean as the callback? | 01:17 |
| ramineni | rloo: :) | 01:17 |
| jroll | I'm going to whip that up and see what folks think | 01:17 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 01:19 |
| JoshNang | jroll: i think 903-910 can just be self._do_next_clean_step(...). we've already returned for all the possible rpc calls at that point | 01:19 |
| JoshNang | that might work too. | 01:20 |
| jroll | I'm not sure there's much difference either way | 01:20 |
| jroll | oh, that would be simpler though | 01:20 |
| *** achanda has quit IRC | 01:21 | |
| JoshNang | well, the calls from nova wouldn't hit that api endpoint. do_provisioning_action is hit when you're going like manage->available | 01:21 |
| * ramineni sees its office time .. will connect again in some time | 01:21 | |
| jroll | oh, huh | 01:22 |
| jroll | oh. | 01:22 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 01:23 |
| jroll | ok, that might be nicer | 01:23 |
| * jroll reviews himself for sanity | 01:23 | |
| * jroll cleans up a bit | 01:25 | |
| JoshNang | nit: the comment is no longer valid | 01:25 |
| JoshNang | (boldly nitting the guy cleaning up my patch..) | 01:25 |
| *** ramineni has quit IRC | 01:25 | |
| jroll | :P | 01:26 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 01:26 |
| jroll | JoshNang: any reason you switched from node to task.node there? | 01:26 |
| jroll | I just removed it | 01:26 |
| jroll | jesus it's 6:30 already? | 01:27 |
| JoshNang | jroll: probably not | 01:27 |
| jroll | k | 01:27 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 01:28 |
| jroll | rebase button ^ | 01:28 |
| JoshNang | thanks for that jroll | 01:30 |
| *** stendulker has quit IRC | 01:32 | |
| mrda | rloo: ? | 01:33 |
| jroll | np dude | 01:34 |
| *** rameshg87 has joined #openstack-ironic | 01:35 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: Fix is_hostname_safe for RFC compliance https://review.openstack.org/165666 | 01:37 |
| *** Marga_ has quit IRC | 01:43 | |
| *** ChuckC has joined #openstack-ironic | 01:49 | |
| rloo | mrda: ? | 02:06 |
| mrda | rloo: Aren't you supposed to be enjoying your evening by now? :) | 02:07 |
| mrda | I just wanted to point out that I addressed your comment. Thanks for that! | 02:07 |
| rloo | mrda: yeah. going to log off etc. thx -- hopefully it'll be approved before tomorrow morning, but if not, i'll look at it. g'night! | 02:08 |
| mrda | thanks Ruby, have a nice night! | 02:08 |
| rameshg87 | rloo: just wanted to let you know that i have addressed your last comments on https://review.openstack.org/#/c/164386/ in https://review.openstack.org/#/c/165297/ | 02:09 |
| rameshg87 | rloo: i have just W-1 for now to not cause merge conflicts with other folks trying to submit features | 02:09 |
| rloo | rameshg87: ok thx. those are nits so thx for making it w-1 for now! | 02:09 |
| rameshg87 | rloo: will rebase and enable it back once others are done | 02:09 |
| rloo | rameshg87: +1 | 02:09 |
| *** rloo has quit IRC | 02:10 | |
| rameshg87 | folks what do you think about https://bugs.launchpad.net/ironic/+bug/1432093 | 02:11 |
| openstack | Launchpad bug 1432093 in Ironic "Instance tear down fail with KeyError: 'is_whole_disk_image'" [Undecided,New] - Assigned to Ramakrishnan G (rameshg87) | 02:11 |
| rameshg87 | we introduced some changes which can fail with KeyError if we tear down/take over nodes which were create before these changes | 02:12 |
| *** kkoski has quit IRC | 02:12 | |
| jroll | rameshg87: I think .get() would fix that, and we should do so | 02:12 |
| rameshg87 | devananda: please prioritize this bug ^^^ | 02:12 |
| rameshg87 | jroll: yeah | 02:12 |
| jroll | rameshg87: totally a valid bug | 02:12 |
| jroll | are you fixing already? | 02:12 |
| rameshg87 | same with 'root_uuid_or_disk_id' | 02:12 |
| * jroll marks as high priority | 02:13 | |
| rameshg87 | jroll: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/pxe.py#L440-L441 | 02:13 |
| jroll | yeah, we need to keep this stuff in mind :/ | 02:13 |
| rameshg87 | jroll: we could have a CI too | 02:13 |
| jroll | rameshg87: that one is fine | 02:13 |
| jroll | rameshg87: because except KeyError | 02:13 |
| rameshg87 | jroll: one that provisions the instance before the change, and teardowns the instances after the change | 02:13 |
| rameshg87 | jroll: a valid use case for upgrade | 02:14 |
| *** harlowja_ is now known as harlowja_away | 02:14 | |
| jroll | rameshg87: yeah, CI to upgrade ironic between commits would be useful, also difficult :) | 02:14 |
| rameshg87 | jroll: yeah .. that's a typical use-case of people who are already using ironic and upgrade to later versions | 02:15 |
| jroll | rameshg87: I'm one of them, I know this all too well | 02:15 |
| rameshg87 | jroll: :) | 02:15 |
| jroll | :P | 02:16 |
| rameshg87 | jroll: i don't know how much time left | 02:16 |
| rameshg87 | jroll: i am rolling out a fix in 15 mins | 02:16 |
| jroll | rameshg87: we have plenty of time for bugs after k3 | 02:16 |
| rameshg87 | jroll: can we set target as k-3 | 02:16 |
| rameshg87 | ? | 02:16 |
| jroll | oh | 02:16 |
| jroll | idk, I guess so? | 02:16 |
| jroll | I assume devananda won't tag until it's morning for him | 02:16 |
| rameshg87 | for people who are upgrading to k3 | 02:16 |
| jroll | yeah | 02:16 |
| * jroll tags | 02:16 | |
| rameshg87 | should not face this issue (if someone is really concerned) | 02:16 |
| jroll | great catch btw | 02:17 |
| jroll | I'll be in and out tonight so poke me when you push the patch | 02:17 |
| rameshg87 | sure | 02:17 |
| rameshg87 | it was just a co-incidence, had an instance created long back | 02:17 |
| rameshg87 | :) | 02:17 |
| rameshg87 | it was lying around in my dev machine | 02:18 |
| jroll | heh | 02:19 |
| jroll | yeah :) | 02:19 |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 02:24 |
| openstackgerrit | Merged openstack/ironic-python-agent: Add fallback secure erase using shred https://review.openstack.org/165247 | 02:25 |
| *** achanda has joined #openstack-ironic | 02:27 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: is_whole_disk_image might not exist for previous instances https://review.openstack.org/165676 | 02:31 |
| rameshg87 | jroll: ^^^ | 02:31 |
| jroll | +2! | 02:32 |
| * jroll bbl | 02:32 | |
| rameshg87 | jroll: 1 min | 02:33 |
| rameshg87 | jroll: fixed a comment | 02:33 |
| rameshg87 | wrong comment in test | 02:33 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: is_whole_disk_image might not exist for previous instances https://review.openstack.org/165676 | 02:33 |
| *** absubram has quit IRC | 02:33 | |
| rameshg87 | jroll: ^^^ patch set 2 :) | 02:33 |
| *** ramineni has joined #openstack-ironic | 02:43 | |
| *** aswadr has joined #openstack-ironic | 02:43 | |
| *** lintan has joined #openstack-ironic | 02:47 | |
| jroll | rameshg87: lol, good catch | 02:48 |
| jroll | rameshg87: if you want it to be perfect, should be "doesn't exist", not "doesn't exists" | 02:48 |
| rameshg87 | jroll: oh grammar, i will live with that :) | 02:51 |
| rameshg87 | jroll: i was totally wrong in the previous comment, that's why i corrected it | 02:51 |
| jroll | :) | 02:51 |
| jroll | yeah | 02:51 |
| jroll | good catch | 02:51 |
| rameshg87 | JoshNang: you around ? | 02:57 |
| jroll | rameshg87: he's been working too hard, he better not be around :P | 02:58 |
| jroll | I can try to help? | 02:58 |
| rameshg87 | jroll: oh okay | 02:58 |
| rameshg87 | jroll: https://review.openstack.org/#/c/161453/12/ironic/dhcp/neutron.py | 02:58 |
| rameshg87 | jroll: L337 | 02:58 |
| rameshg87 | jroll: i didn't get his concern of not saving the cleaning neutron vif ids in ironic ports | 02:59 |
| jroll | rameshg87: neutron knows what's already there, ironic may be wrong | 02:59 |
| rameshg87 | jroll: but we still do the check in L349 na | 02:59 |
| rameshg87 | jroll: to check if the neutron port is there in the ironic port | 02:59 |
| jroll | rameshg87: we do similar things for provisioning/tenant networks downstream, we had problems using anything but neutron as the source of truth | 02:59 |
| rameshg87 | jroll: we still loop through each neutron port and check if mac address is there in ironic port | 03:00 |
| rameshg87 | jroll: otherway round is the same too - if we had the vifs saved in ironic port | 03:00 |
| rameshg87 | jroll: if operator deleted ironic port, we are not going to delete the neutron port either way | 03:01 |
| jroll | rameshg87: but if operator deletes the vif from ironic port, this will delete that port, the other way won't | 03:01 |
| rameshg87 | jroll: oh | 03:01 |
| rameshg87 | jroll: will some operator do that ? :) | 03:01 |
| jroll | rameshg87: same thing applies if the operator adds a port without ironic knowing | 03:02 |
| jroll | rameshg87: probably not, but you never know | 03:02 |
| jroll | rameshg87: this generally improves robustness for weird failure scenarios | 03:02 |
| rameshg87 | jroll: just concerned because we loop over (number of neutron ports in whole environment * number of ironic ports for that node) | 03:02 |
| jroll | rameshg87: e.g. if a port-create succeeds but the "save to ironic db" part fails, this will handle it | 03:02 |
| rameshg87 | jroll: hmm okay | 03:03 |
| jroll | rameshg87: you're concerned about that being slow, in a process that does things like run shred? :) | 03:03 |
| rameshg87 | jroll: hehe no | 03:03 |
| rameshg87 | jroll: i know we are just doing some logical operations here (no external) | 03:03 |
| *** coolsvap|afk is now known as coolsvap | 03:04 | |
| rameshg87 | jroll: but just felt we dont need to loop over if we just saved it | 03:04 |
| jroll | rameshg87: oh, hmm, I didn't see the list_ports parameters. I wonder if that could be better. | 03:04 |
| jroll | rameshg87: yeah, I see what you mean | 03:04 |
| jroll | rameshg87: I thought we were looping over all ports for that node :/ | 03:04 |
| jroll | rameshg87: but we don't tie the node id to the port at all. this makes me sad. | 03:05 |
| rameshg87 | jroll: i meant looping over all neutron ports in that environment | 03:05 |
| jroll | rameshg87: right, I assumed it was something like list_ports({node: node.uuid}), not all ports for the network. I re-read it now. | 03:06 |
| rameshg87 | jroll: yeah | 03:06 |
| jroll | which we might be able to do... | 03:07 |
| jroll | I'm going to leave a comment for now because it's too late for testing code | 03:07 |
| rameshg87 | jroll: yeah may be something to look for later | 03:07 |
| rameshg87 | jroll: i agree | 03:07 |
| rameshg87 | jroll: and one more | 03:08 |
| rameshg87 | jroll: https://review.openstack.org/#/c/161453/12/ironic/dhcp/neutron.py L304 | 03:08 |
| rameshg87 | jroll: we bind all the ironic ports to the same network | 03:08 |
| rameshg87 | jroll: i see JoshNang reply "I disagree, you can't know that PXE is enabled for all the ports." | 03:09 |
| rameshg87 | jroll: but i think nova doesn't know it either | 03:09 |
| rameshg87 | jroll: and nova always chooses ironic ports randomly and assign neutron networks to them | 03:09 |
| rameshg87 | jroll: i think we have got to have pxe enabled for all the ironic ports for deploy to be successful - this is my knowledge | 03:10 |
| rameshg87 | but i might be wrong | 03:10 |
| jroll | rameshg87: left a note for the first one, I think that could be a follow-up optimization | 03:10 |
| rameshg87 | yeah | 03:10 |
| * jroll thinks about the second | 03:10 | |
| jroll | so I tend to think nova's behavior is a bug, but not a major one (yet) as things could we physically or virtually wired in ways that break that | 03:11 |
| rameshg87 | jroll: if operator had 4 ports enrolled in ironic - all of them will get IP in the same subnet - like all 4 macs will get 192.168.1.2, 192.168.1.3, 192.168.1.4, 192.168.1.5 | 03:11 |
| jroll | rameshg87: assuming they're on the same physical network, that's true :) | 03:11 |
| *** absubram has joined #openstack-ironic | 03:12 | |
| rameshg87 | jroll: shouldn't be a problem, but in a neutron environment i think it's expected that neutron should be able to connect any ports to any network (somehow) :) | 03:12 |
| jroll | morgabra: ^ if you're around I would love another set of eyes on the implications/correctness of the comments in that link | 03:12 |
| jroll | rameshg87: yeah, except neutron barely works at all for ironic's purposes today :) | 03:13 |
| rameshg87 | :) | 03:13 |
| jroll | rameshg87: I'd have to think more tomorrow on this... I know josh's way won't break; I don't think your way will break. | 03:14 |
| jroll | does that make sense? | 03:14 |
| jroll | rameshg87: we also do really different DHCP and network things in our deployment, so it's hard to reason between that and upstream, so we very well might be wrong | 03:15 |
| jroll | (we being rackspace people) | 03:15 |
| rameshg87 | jroll: sure .. | 03:15 |
| rameshg87 | jroll: may be we can talk your tomorrow | 03:16 |
| jroll | rameshg87: indeed, my brain is fried for today | 03:16 |
| rameshg87 | :) | 03:16 |
| rameshg87 | jroll: have a good sleep .. | 03:16 |
| jroll | rameshg87: will do, have a good day :) | 03:16 |
| jroll | see you "tonight" | 03:16 |
| rameshg87 | jroll: good night | 03:16 |
| *** Marga_ has joined #openstack-ironic | 03:21 | |
| *** Marga_ has quit IRC | 03:21 | |
| *** Marga_ has joined #openstack-ironic | 03:22 | |
| * rameshg87 will be back in few hours | 03:27 | |
| *** rameshg87 has quit IRC | 03:27 | |
| *** coolsvap is now known as coolsvap|afk | 03:32 | |
| *** coolsvap|afk is now known as coolsvap | 03:39 | |
| *** achanda has quit IRC | 03:42 | |
| *** greghaynes has quit IRC | 03:43 | |
| *** achanda has joined #openstack-ironic | 03:44 | |
| *** stendulker has joined #openstack-ironic | 03:49 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver https://review.openstack.org/154808 | 03:51 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for iscsi_ilo driver https://review.openstack.org/154814 | 03:53 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for agent_ilo driver https://review.openstack.org/154816 | 03:54 |
| *** achanda has quit IRC | 04:05 | |
| *** jamielennox is now known as jamielennox|away | 04:06 | |
| *** lazy_prince is now known as killer_prince | 04:15 | |
| *** Marga_ has quit IRC | 04:18 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver https://review.openstack.org/154808 | 04:24 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for iscsi_ilo driver https://review.openstack.org/154814 | 04:26 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for agent_ilo driver https://review.openstack.org/154816 | 04:26 |
| *** Nisha has joined #openstack-ironic | 04:27 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Ilo drivers sets capabilities:boot_mode in node https://review.openstack.org/155731 | 04:27 |
| *** achanda has joined #openstack-ironic | 04:37 | |
| *** achanda has quit IRC | 04:41 | |
| *** Marga_ has joined #openstack-ironic | 04:47 | |
| *** pradipta has joined #openstack-ironic | 04:52 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Support agent_ilo driver to perform cleaning https://review.openstack.org/164646 | 04:53 |
| *** ijw has joined #openstack-ironic | 04:56 | |
| *** yuanying has quit IRC | 05:01 | |
| *** rameshg87 has joined #openstack-ironic | 05:02 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: follow up patch for ilo capabilities https://review.openstack.org/165310 | 05:02 |
| rameshg87 | Haomeng: are you planning to work on - https://review.openstack.org/143919 ? | 05:02 |
| rameshg87 | Haomeng: if not any time soon, i can attempt to continue it | 05:03 |
| *** yuanying has joined #openstack-ironic | 05:06 | |
| *** achanda has joined #openstack-ironic | 05:08 | |
| *** Haomeng|2 has joined #openstack-ironic | 05:11 | |
| rameshg87 | Haomeng|2: ^^^ | 05:12 |
| rameshg87 | Haomeng|2: don't know if you are logged in as Haomeng or Haomeng|2 | 05:12 |
| *** Haomeng has quit IRC | 05:14 | |
| *** Haomeng has joined #openstack-ironic | 05:21 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 05:22 |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 05:22 |
| *** Haomeng|2 has quit IRC | 05:24 | |
| openstackgerrit | Merged openstack/python-ironicclient: Add support for logical names https://review.openstack.org/158520 | 05:24 |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Support agent_ilo driver to perform cleaning https://review.openstack.org/164646 | 05:29 |
| Haomeng | rameshg87: hi, ramesh | 05:30 |
| *** yuanying has quit IRC | 05:30 | |
| Haomeng | rameshg87: yes, I am working on other things these days, so go ahead please, the unittest code is not working now, so if you can help, thank you | 05:30 |
| rameshg87 | Haomeng: sure, i will take over..thanks | 05:31 |
| Haomeng | rameshg87: thank you very much:) | 05:31 |
| Haomeng | rameshg87: the original unittest is working, but it not working with latest code recently:) | 05:31 |
| *** yuanying has joined #openstack-ironic | 05:31 | |
| *** greghaynes has joined #openstack-ironic | 05:39 | |
| *** coolsvap is now known as coolsvap|afk | 05:41 | |
| *** ChuckC has quit IRC | 05:45 | |
| *** yuanying has quit IRC | 05:46 | |
| *** yuanying has joined #openstack-ironic | 05:47 | |
| *** coolsvap|afk is now known as coolsvap | 05:48 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Automate uefi boot iso creation for iscsi_ilo driver https://review.openstack.org/155900 | 05:53 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Generate keystone_authtoken options in sample config file https://review.openstack.org/162919 | 05:57 |
| *** yuanying has quit IRC | 06:05 | |
| *** yuanying has joined #openstack-ironic | 06:08 | |
| *** subscope_ has joined #openstack-ironic | 06:09 | |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add empty unit test files for ris https://review.openstack.org/165704 | 06:09 |
| rameshg87 | ramineni: Nisha: ^^^ | 06:09 |
| rameshg87 | ramineni: Nisha: can you please help me get this in quickly | 06:10 |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add empty unit test files for ris https://review.openstack.org/165704 | 06:11 |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add empty unit test files for ris https://review.openstack.org/165704 | 06:15 |
| rameshg87 | ramineni: Nisha: ^^^ | 06:15 |
| *** killer_prince is now known as lazy_prince | 06:20 | |
| *** achanda has quit IRC | 06:28 | |
| *** yog_ has joined #openstack-ironic | 06:28 | |
| openstackgerrit | Yuiko Takada proposed stackforge/ironic-discoverd: Add supporting generate config files https://review.openstack.org/163281 | 06:33 |
| *** ramineni has quit IRC | 06:33 | |
| *** ramineni has joined #openstack-ironic | 06:33 | |
| *** pcaruana has quit IRC | 06:33 | |
| *** ramineni1 has joined #openstack-ironic | 06:34 | |
| *** ramineni has quit IRC | 06:34 | |
| openstackgerrit | Yuiko Takada proposed stackforge/ironic-discoverd: Add supporting generate config files https://review.openstack.org/163281 | 06:34 |
| openstackgerrit | Merged stackforge/proliantutils: ILO: Add empty unit test files for ris https://review.openstack.org/165704 | 06:35 |
| openstackgerrit | Yuiko Takada proposed stackforge/ironic-discoverd: Add supporting generate config files https://review.openstack.org/163281 | 06:35 |
| openstackgerrit | Yuiko Takada proposed stackforge/ironic-discoverd: Add supporting generate config files https://review.openstack.org/163281 | 06:37 |
| *** achanda has joined #openstack-ironic | 06:52 | |
| *** jcoufal has joined #openstack-ironic | 06:53 | |
| *** saripurigopi has joined #openstack-ironic | 06:54 | |
| *** achanda has quit IRC | 07:00 | |
| openstackgerrit | Michael Davies proposed openstack/ironic: New field 'name' not supported in port REST API https://review.openstack.org/163730 | 07:00 |
| *** achanda has joined #openstack-ironic | 07:04 | |
| *** ifarkas has joined #openstack-ironic | 07:05 | |
| *** subscope_ has quit IRC | 07:07 | |
| *** Marga_ has quit IRC | 07:09 | |
| *** ijw has quit IRC | 07:13 | |
| mrda | devananda: ^^^ I know it's getting late in the release cycle, but I think I've met all the requirements for logical_name in the ports section of the API. It's backwardly compatible, and we've restored get_rpc_node as per the suggestions. But if it's too late I get that too :) Just FYI... | 07:21 |
| *** achanda has quit IRC | 07:37 | |
| *** 21WABNX9V has joined #openstack-ironic | 07:38 | |
| openstackgerrit | Aparna proposed stackforge/proliantutils: ILO: Support for configuring httpboot through RIS https://review.openstack.org/163325 | 07:44 |
| *** athomas has joined #openstack-ironic | 07:46 | |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add unit tests for ris.RISOperations._rest_op https://review.openstack.org/165718 | 07:47 |
| *** chlong has quit IRC | 07:47 | |
| rameshg87 | ramineni1: Nisha: stendulker: ^^^ | 07:48 |
| stendulker | rameshg87: o/ | 07:48 |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add unit tests for ris.RISOperations._rest_op https://review.openstack.org/165718 | 07:49 |
| rameshg87 | ramineni1: Nisha: stendulker: raised unit tests review for main method in RIS ^^ please have a look | 07:49 |
| stendulker | rameshg87: ok | 07:50 |
| *** stendulker has quit IRC | 07:54 | |
| *** stendulker has joined #openstack-ironic | 07:54 | |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add unit tests for ris.RISOperations._rest_op https://review.openstack.org/165718 | 07:55 |
| *** mrda is now known as mrda-away | 07:58 | |
| *** romcheg has joined #openstack-ironic | 08:02 | |
| *** dlpartain has joined #openstack-ironic | 08:05 | |
| *** yuanying_ has joined #openstack-ironic | 08:05 | |
| *** romcheg has quit IRC | 08:07 | |
| *** romcheg has joined #openstack-ironic | 08:07 | |
| *** yuanying has quit IRC | 08:08 | |
| *** rsalevsky has joined #openstack-ironic | 08:09 | |
| *** dlpartain has left #openstack-ironic | 08:14 | |
| GheRivero | morning | 08:16 |
| Nisha | morning GheRivero | 08:19 |
| *** romcheg has quit IRC | 08:21 | |
| *** naohirot has quit IRC | 08:27 | |
| *** jistr has joined #openstack-ironic | 08:37 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver https://review.openstack.org/154808 | 08:38 |
| *** Haomeng|2 has joined #openstack-ironic | 08:51 | |
| *** Haomeng has quit IRC | 08:55 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for iscsi_ilo driver https://review.openstack.org/154814 | 08:55 |
| openstackgerrit | Ramakrishnan G proposed stackforge/proliantutils: ILO: Add unit tests for ris.RISOperations._rest_op https://review.openstack.org/165718 | 08:58 |
| *** andreykurilin_ has joined #openstack-ironic | 08:58 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for agent_ilo driver https://review.openstack.org/154816 | 08:58 |
| *** mgoddard has joined #openstack-ironic | 09:15 | |
| *** lucasagomes has joined #openstack-ironic | 09:16 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: follow up patch for ilo capabilities https://review.openstack.org/165310 | 09:17 |
| *** romcheg has joined #openstack-ironic | 09:17 | |
| *** logan2 has quit IRC | 09:22 | |
| *** logan2 has joined #openstack-ironic | 09:22 | |
| *** mgoddard has quit IRC | 09:23 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Support agent_ilo driver to perform cleaning https://review.openstack.org/164646 | 09:27 |
| *** andreykurilin_ has quit IRC | 09:27 | |
| *** andreykurilin_ has joined #openstack-ironic | 09:28 | |
| *** pas-ha has joined #openstack-ironic | 09:28 | |
| *** yog_ has quit IRC | 09:31 | |
| *** derekh has joined #openstack-ironic | 09:37 | |
| *** yog_ has joined #openstack-ironic | 09:47 | |
| *** andreykurilin_ has quit IRC | 09:47 | |
| *** ndipanov has joined #openstack-ironic | 09:48 | |
| *** Nisha has quit IRC | 09:49 | |
| *** dtantsur|afk is now known as dtantsur | 09:54 | |
| dtantsur | Late good morning :) | 09:54 |
| *** absubram has quit IRC | 10:00 | |
| *** yog_ has quit IRC | 10:02 | |
| *** pelix has joined #openstack-ironic | 10:02 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Ilo drivers sets capabilities:boot_mode in node https://review.openstack.org/155731 | 10:05 |
| *** saripurigopi has quit IRC | 10:08 | |
| *** alex_xu has quit IRC | 10:12 | |
| *** kalpase has joined #openstack-ironic | 10:13 | |
| *** alex_xu has joined #openstack-ironic | 10:14 | |
| *** MattMan has joined #openstack-ironic | 10:14 | |
| *** yog_ has joined #openstack-ironic | 10:15 | |
| *** lazy_prince has quit IRC | 10:17 | |
| openstackgerrit | Ghe Rivero proposed openstack/python-ironicclient: Use oslo.i18n lib https://review.openstack.org/162797 | 10:19 |
| *** yog_ has quit IRC | 10:19 | |
| openstackgerrit | Ghe Rivero proposed openstack/python-ironicclient: Clean openstack-common module list https://review.openstack.org/162799 | 10:23 |
| openstackgerrit | Ghe Rivero proposed openstack/python-ironicclient: Use oslo.i18n lib https://review.openstack.org/162797 | 10:24 |
| *** killer_prince has joined #openstack-ironic | 10:24 | |
| *** killer_prince has quit IRC | 10:24 | |
| *** lazy_prince has joined #openstack-ironic | 10:24 | |
| ramineni1 | dtantsur: please update your vote again on review https://review.openstack.org/#/c/157715/ | 10:28 |
| ramineni1 | dtantsur: it resulted in merge conflict while merging :( ..so rebased again | 10:29 |
| *** tiagogomes_ has joined #openstack-ironic | 10:31 | |
| dtantsur | sure, once tests pass again | 10:33 |
| *** saripurigopi has joined #openstack-ironic | 10:34 | |
| ramineni1 | dtantsur: from morning , jenkins is not running tests on the patch :( it looks like stuck , gave recheck twice | 10:35 |
| ramineni1 | dtantsur: any idea, what else could solve the issue | 10:35 |
| ramineni1 | dtantsur: running only check-tripleo-ironic-undercloud-precise-nonha test everytime | 10:37 |
| *** chenglch has quit IRC | 10:37 | |
| rameshg87 | lucasagomes: dtantsur: need one more +2 (hopefully) - https://review.openstack.org/#/c/165676/ | 10:39 |
| rameshg87 | lucasagomes: dtantsur: a bug for k3 | 10:39 |
| *** coolsvap is now known as coolsvap|afk | 10:41 | |
| lucasagomes | rameshg87, "Tests when 'is_whole_disk_image' exists in driver_internal_info" | 10:42 |
| lucasagomes | rameshg87, I don't see any assert there checking that | 10:42 |
| *** vdrok_afk is now known as vdrok | 10:43 | |
| lucasagomes | perhaps self.assertIn('is_whole_disk_image', self.node.internal_driver_info) ? | 10:43 |
| lucasagomes | driver_internal_info* | 10:43 |
| * rameshg87 checks | 10:43 | |
| rameshg87 | lucasagomes: it's there by default now | 10:44 |
| lucasagomes | I see the setUp adding it | 10:44 |
| lucasagomes | right, I would make it more explicity on the test itself | 10:44 |
| lucasagomes | but fine | 10:44 |
| rameshg87 | lucasagomes: https://github.com/openstack/ironic/blob/master/ironic/tests/db/utils.py#L74 | 10:44 |
| lucasagomes | rameshg87, right | 10:45 |
| lucasagomes | +a :) | 10:45 |
| rameshg87 | thanks :) | 10:45 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Fail deploy if root uuid or disk id isn't available https://review.openstack.org/143919 | 10:48 |
| stendulker | lucasgomes: hi | 10:49 |
| stendulker | lucasgomes: Please look at https://review.openstack.org/#/c/155731/ Ilo drivers sets capabilities:boot_mode in node. One you had reviewed few days back had to be rebased to handle merge conflicts. | 10:50 |
| dtantsur | ramineni1, please note that blindly rechecking is nearly never a good answer | 10:52 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Fail deploy if root uuid or disk id isn't available https://review.openstack.org/143919 | 10:52 |
| ramineni1 | dtantsur: oh | 10:53 |
| dtantsur | ramineni1, according to http://status.openstack.org/zuul/ most of tests has passed, but one is taking 9 hrs already | 10:53 |
| dtantsur | ramineni1, you may want to sync with infra on it. but chances are high that problems are caused by Kilo 3 | 10:53 |
| dtantsur | all the projects are testing and approving last minutes changes :) | 10:53 |
| ramineni1 | dtantsur: ok | 10:55 |
| *** yuanying_ has quit IRC | 10:57 | |
| dtantsur | ramineni1, +A and hope the check will be ok soon | 10:57 |
| ramineni1 | dtantsur: great..thanks | 10:58 |
| ramineni1 | dtantsur: ya, hopefully :) | 10:58 |
| * ramineni1 leaves for the day, connect back later again | 11:00 | |
| *** ramineni1 has quit IRC | 11:01 | |
| *** pcaruana has joined #openstack-ironic | 11:05 | |
| dtantsur | ifarkas, o/ mind reviewing https://review.openstack.org/#/c/165062/ ? | 11:05 |
| ifarkas | dtantsur, sure! | 11:06 |
| *** sambetts has quit IRC | 11:09 | |
| *** stendulker has quit IRC | 11:09 | |
| *** sambetts has joined #openstack-ironic | 11:11 | |
| *** saripurigopi has quit IRC | 11:13 | |
| *** coolsvap|afk is now known as coolsvap | 11:36 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Fail deploy if root uuid or disk id isn't available https://review.openstack.org/143919 | 11:37 |
| * rameshg87 goes out for a while, will be back in few hours | 11:40 | |
| *** rameshg87 has quit IRC | 11:40 | |
| *** pradipta has quit IRC | 11:40 | |
| openstackgerrit | Merged openstack/ironic: is_whole_disk_image might not exist for previous instances https://review.openstack.org/165676 | 11:41 |
| openstackgerrit | Ghe Rivero proposed openstack/ironic: Use oslo_log lib https://review.openstack.org/157602 | 11:43 |
| openstackgerrit | Harshada Mangesh Kakad proposed openstack/ironic: Implement get_boot_device for seamicro driver https://review.openstack.org/165788 | 11:48 |
| *** dprince has joined #openstack-ironic | 12:13 | |
| *** chlong has joined #openstack-ironic | 12:14 | |
| *** BadCub_Hiding is now known as BadCub | 12:19 | |
| BadCub | Morning Folks | 12:19 |
| Shrews | morning BadCub and everyone else | 12:25 |
| *** Marga_ has joined #openstack-ironic | 12:26 | |
| *** chlong has quit IRC | 12:26 | |
| lucasagomes | Shrews, BadCub good morning | 12:27 |
| BadCub | hey guys | 12:32 |
| *** ramineni has joined #openstack-ironic | 12:34 | |
| TheJulia | good morning | 12:40 |
| openstackgerrit | Merged openstack/ironic: Hide inspection_*_at fields if version < 1.6 https://review.openstack.org/165596 | 12:41 |
| NobodyCam | good morning Ironicers | 12:41 |
| *** achanda has joined #openstack-ironic | 12:42 | |
| *** chlong has joined #openstack-ironic | 12:43 | |
| dtantsur | Shrews, BadCub, TheJulia, NobodyCam, morning | 12:47 |
| dtantsur | aren't you early today, folks? :) | 12:47 |
| *** achanda has quit IRC | 12:47 | |
| BadCub | morning TheJulia, dtantsur NobodyCam. Shrews and lucasagomes | 12:47 |
| TheJulia | dtantsur: Well, it is approaching 9 AM here... so no. :) | 12:48 |
| *** rloo has joined #openstack-ironic | 12:49 | |
| BadCub | I couldn't sleep. LOL | 12:49 |
| TheJulia | Thats no good | 12:49 |
| BadCub | yeah | 12:50 |
| BadCub | at least we are close to getting reviews done :-) | 12:52 |
| NobodyCam | morning dtantsur BadCub TheJulia lucasagomes | 12:53 |
| * Shrews feels snubbed | 12:53 | |
| TheJulia | Good morning Shrews | 12:54 |
| Shrews | hey TheJulia! morning | 12:54 |
| BadCub | I said morning Shrews LOL | 12:54 |
| Shrews | BadCub: yeah, i was poking at NobodyCam :-P | 12:54 |
| BadCub | hehehehe.. | 12:54 |
| TheJulia | :) | 12:54 |
| *** lazy_prince has quit IRC | 12:55 | |
| NobodyCam | morning Shrews :) | 12:55 |
| Shrews | :-P | 12:55 |
| NobodyCam | i didn't see you | 12:55 |
| NobodyCam | :-p | 12:55 |
| Shrews | morning | 12:55 |
| *** killer_prince has joined #openstack-ironic | 12:55 | |
| NobodyCam | :) | 12:55 |
| *** killer_prince is now known as lazy_prince | 12:56 | |
| devananda | g'morning, all | 12:59 |
| *** athomas has quit IRC | 12:59 | |
| NobodyCam | good morning devananda :) | 13:00 |
| BadCub | morning devananda | 13:00 |
| dtantsur | devananda, morning, you're early too Oo | 13:01 |
| dtantsur | am I the only one to wake up at 10am today? :D | 13:01 |
| NobodyCam | dtantsur: must be nice to sleep in like that :-p | 13:02 |
| BadCub | 10am? I wish I could sleep till 10!! LOL | 13:02 |
| *** kalpase has left #openstack-ironic | 13:02 | |
| dtantsur | I usually can't, but seems like today is special :) | 13:02 |
| devananda | dtantsur: 10am is when I meet with ttx again (if not sooner) with the status of our release | 13:02 |
| rloo | morning everyone. So, no reviews need to be done today? I can take the day off? :-) | 13:03 |
| devananda | rloo: exactly :) | 13:03 |
| NobodyCam | morning rloo :) | 13:03 |
| dtantsur | rloo, yeah relax and enjoy :) and g'morning | 13:03 |
| BadCub | devananda: I think we are looking pretty good. We have 4 BPs left to get done, and a bunch of stuff merged last night | 13:03 |
| devananda | BadCub: you have an interesting definition of good .... | 13:03 |
| * BadCub puts on his optimistic face | 13:04 | |
| * rloo thinks 4 is a lot | 13:04 | |
| devananda | BadCub: we have 4 hours for things to move through the (very crowded) gate | 13:04 |
| *** athomas has joined #openstack-ironic | 13:04 | |
| NobodyCam | whats first up on the review line! | 13:05 |
| devananda | BadCub: seriously. we can't punt any longer. of those 4 BP, which are going to be completed and merged in trunk in the next 4 hours? | 13:05 |
| BadCub | teh gate is the big issue | 13:05 |
| devananda | there's a notethat https://review.openstack.org/#/c/165528/ is failing tests, but it's not. last run passed | 13:05 |
| devananda | dtantsur: you -1'd it -- can your nits be done in a follow on patch? | 13:06 |
| *** mjturek1 has joined #openstack-ironic | 13:06 | |
| BadCub | deva | 13:06 |
| BadCub | oops | 13:06 |
| dtantsur | devananda, definitely | 13:06 |
| dtantsur | lemme change my vote with this condition | 13:06 |
| BadCub | ILO cleaning is going to be a problem if we do not land the dependent patch | 13:06 |
| devananda | dtantsur: great! let's land that, and... | 13:07 |
| * dtantsur changed to +2 | 13:07 | |
| BadCub | can someone poke at https://review.openstack.org/#/c/161453/ add cleaning agent driver? | 13:07 |
| *** dlpartain has joined #openstack-ironic | 13:08 | |
| NobodyCam | looking now | 13:08 |
| *** jistr has quit IRC | 13:08 | |
| dtantsur | will look at patches after lunch | 13:08 |
| *** dlpartain has left #openstack-ironic | 13:08 | |
| dtantsur | (if you just don't land everything which I advice you to do :D ) | 13:08 |
| openstackgerrit | Harshada Mangesh Kakad proposed openstack/ironic: Implement get_boot_device for seamicro driver https://review.openstack.org/165788 | 13:08 |
| *** jistr has joined #openstack-ironic | 13:09 | |
| devananda | BadCub: https://review.openstack.org/#/c/157715/ is passing the latest tests (first iLO clenaing patch) | 13:12 |
| devananda | BadCub: it was already approved ,so that should merge soon | 13:12 |
| BadCub | I see that too | 13:12 |
| *** kkoski has joined #openstack-ironic | 13:14 | |
| BadCub | https://review.openstack.org/#/c/164646/ support agent_ilo driver to perform cleaning needs review | 13:15 |
| rloo | devananda: so I don't know that we should be modifying strings after k-3 (wrt https://review.openstack.org/#/c/165528/7/ironic/conductor/manager.py) but it should be ok to fix in L? | 13:16 |
| devananda | rloo: if we land that follow up within the next day or so, it's fine | 13:17 |
| rloo | devananda: oh, good to know. | 13:17 |
| devananda | rloo: the translation team needs time to do their work, but something small like that is a reasonable exception | 13:17 |
| rloo | devananda: (if Ironic even gets much attention from the translation team...) | 13:18 |
| devananda | rloo: (right ...) | 13:18 |
| *** igordcard has joined #openstack-ironic | 13:18 | |
| *** chlong has quit IRC | 13:21 | |
| lucasagomes | devananda, rloo g'morning | 13:23 |
| rloo | hi lucasagomes! | 13:23 |
| devananda | lucasagomes: what's difference betwen pxe_ipa and pxe_ipa-src and agent_ssh jobs ? | 13:24 |
| lucasagomes | devananda, pxe_ipa and pxe_ipa-src uses the PXE driver iscsi methodology with the IPA ramdisk | 13:24 |
| lucasagomes | the -src is because it's going to build IPA from source | 13:24 |
| lucasagomes | which should run on the openstack/ironic-python-agent gate | 13:25 |
| lucasagomes | without -src it just download a pre-built ramdisk | 13:25 |
| devananda | ahh. gotcha | 13:25 |
| lucasagomes | agent_ssh uses IPA with the agent deploy methodology (downloading the image from swift, and having IPA partitioning the disk, etc) | 13:25 |
| lucasagomes | devananda, I'm really bad with names :( | 13:26 |
| lucasagomes | I wanted to keep the pxe_ namespace, but maybe that's a bit misleading | 13:26 |
| devananda | lucasagomes: heh, names are hard. since they all use pxe for booting a ramdisk, and ssh for power control in the gate ... perhaps "iscsi" is better in the name | 13:28 |
| devananda | lucasagomes: that should probably get cleaned up later (post k3 rush) along with an inline comment explaining the job purpose | 13:28 |
| devananda | lucasagomes: so reviewers (and us, 3 monts from now!) can look at the jobs and where they're being run and remember why we did that :) | 13:28 |
| lucasagomes | devananda, yeah indeed, iscsi would better describe the job. But I was thinking about the consistency with the driver names in ironic pxe_ drivers etc... | 13:29 |
| lucasagomes | devananda, oh that would be awesome | 13:29 |
| lucasagomes | to have a description | 13:29 |
| * lucasagomes writes it down | 13:29 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Add Cleaning Operations for iLO drivers https://review.openstack.org/157715 | 13:33 |
| * Shrews steps away for an appointment. bbs | 13:33 | |
| * devananda uses the recheck hammer on ^ because jenkins04 was dead | 13:34 | |
| *** Marga_ has quit IRC | 13:35 | |
| lucasagomes | it's only me that dislike having cleaning priority level as config options? | 13:37 |
| devananda | lucasagomes: why do you dislike it? | 13:37 |
| lucasagomes | the more I think about zapping and cleaning, the more I think the steps should be plugins | 13:37 |
| dtantsur | lucasagomes, me too, but I didn't find a really good reason not too | 13:38 |
| devananda | lucasagomes: when using IPA, they are plugged through the hardware manager | 13:38 |
| lucasagomes | devananda, cause they all have defaults, and the more we add new steps more those defaults r going to change | 13:38 |
| lucasagomes | we shouldn't be changing defaults on config options very often | 13:38 |
| devananda | lucasagomes: though I belueve IPA should have configurable "stages" of cleaning, or "groups" or something | 13:38 |
| lucasagomes | I cna picture steps being added like "oh this needs to run before that" etc | 13:39 |
| lucasagomes | and that will cause defaults to be changed all along | 13:39 |
| devananda | lucasagomes: so priority levels are per-driver, and can accept duplicates | 13:39 |
| lucasagomes | devananda, yeah... it's something we can work later | 13:39 |
| devananda | lucasagomes: much like init.d priority ... there's plenty of precedent for using this model in a plugin system | 13:40 |
| devananda | (unless I completely misunderstood the implementation here) | 13:40 |
| lucasagomes | devananda, I think ur correct | 13:40 |
| lucasagomes | and it's something we can improve later | 13:41 |
| devananda | k k | 13:41 |
| BadCub | is anyone free to give a second +2/+A to: https://review.openstack.org/#/c/154816/ secure boot for agent_ilo driver ? | 13:42 |
| devananda | https://review.openstack.org/#/c/155900/ is lower priority, but fairly isolated. lucas, looks like you've reviewed before? | 13:43 |
| BadCub | brb | 13:44 |
| devananda | BadCub: the agent_ilo patch you linked depends on iscsi_ilo and pxe_ilo -- and pxe_ilo is a big one -- https://review.openstack.org/#/c/154808/48 | 13:45 |
| *** stendulker has joined #openstack-ironic | 13:45 | |
| *** kkoski has quit IRC | 13:46 | |
| *** mtanino has joined #openstack-ironic | 13:47 | |
| devananda | if we can get ^ in, we can probalby complete the feature, because the two subsequent patches are both relatively small and constrained to the driver | 13:47 |
| *** kkoski has joined #openstack-ironic | 13:49 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Stop waiting for power off to happen after introspection https://review.openstack.org/165062 | 13:50 |
| NobodyCam | JoshNang: BadCub: 161453 -1'd for conflict in execption class. | 13:54 |
| BadCub | thnx NobodyCam | 13:55 |
| openstackgerrit | Harshada Mangesh Kakad proposed openstack/ironic: Implement get_boot_device for seamicro driver https://review.openstack.org/165788 | 13:55 |
| BadCub | devananda: yes, NobodyCam was reviewing the dependent | 13:56 |
| *** rameshg87 has joined #openstack-ironic | 13:57 | |
| *** wuhg has joined #openstack-ironic | 13:59 | |
| jroll | mornin' | 13:59 |
| BadCub | we also need reviews on: iscsi_ilo https://review.openstack.org/#/c/154814/ & pxe_ilo https://review.openstack.org/#/c/154808/ | 13:59 |
| rameshg87 | morning all :) | 13:59 |
| BadCub | mornin jroll and rameshg87 | 13:59 |
| rameshg87 | BadCub: o/ | 14:00 |
| NobodyCam | morning rameshg87 :) | 14:01 |
| rameshg87 | NobodyCam: o/ | 14:01 |
| NobodyCam | :) | 14:01 |
| rameshg87 | devananda: Shrews: NobodyCam: have a question about https://review.openstack.org/#/c/155460/9/ironic/drivers/modules/pxe.py (merged patch) | 14:01 |
| devananda | ? | 14:02 |
| rameshg87 | devananda: Shrews: NobodyCam: shouldn't we move the instance to error state if we cannot get driver_internal_info['root_uuid'] | 14:02 |
| rameshg87 | devananda: unless we generate pxe config, the node is never going to boot if someone triggered a reboot | 14:02 |
| jroll | rameshg87: +1 | 14:04 |
| devananda | rameshg87: not in all cases. 1) node is booting from local disk (in which case we should skip that check!) 2) user just doesn't reboot it ... | 14:04 |
| rameshg87 | devananda: yeah for local boot we should skip that check | 14:04 |
| jroll | (1) is valid, I think (2) is not | 14:04 |
| devananda | rameshg87: (2) is applicable for instances deployed prior to this code being landed, because there was no root_uuid stored. we shouldn't just fail those instances when deploying this code | 14:04 |
| devananda | jroll: continuous deployment ... | 14:05 |
| jroll | with (2) you're setting your users up for failure | 14:05 |
| jroll | oh | 14:05 |
| devananda | you'd fail all the instances | 14:05 |
| jroll | yeah I didn't realize that part, sorry | 14:05 |
| devananda | if you roll over the conductors | 14:05 |
| rameshg87 | devananda: it's only for takeover, right ? | 14:05 |
| devananda | yah | 14:05 |
| *** kkoski has quit IRC | 14:05 | |
| devananda | rameshg87: correct. only affects calling prepare() on conductor other than the one which still has a pxe config file for that instance | 14:05 |
| devananda | which will happen during take over of nodes deployed before that patch | 14:06 |
| rameshg87 | devananda: yeah | 14:06 |
| * jroll isn't quite awake yet and isn't feeling awesome today | 14:06 | |
| devananda | so, before thta patch, take over was actually broken for non-local-booting nodes. | 14:06 |
| rameshg87 | devananda: yeah okay .. | 14:07 |
| rameshg87 | devananda: so it's still broken for old nodes (if they reboot) | 14:07 |
| devananda | correct | 14:07 |
| NobodyCam | JoshNang: happen to around yet? | 14:07 |
| rameshg87 | devananda: but works for newly deployed nodes | 14:07 |
| devananda | hence the warning :) | 14:07 |
| rameshg87 | okay | 14:07 |
| devananda | rameshg87: exactly | 14:07 |
| jroll | NobodyCam: I can take over things for josh if needed | 14:07 |
| rameshg87 | got it thanks :) | 14:07 |
| NobodyCam | jroll want to take a quick look at the comments on 161453? | 14:08 |
| NobodyCam | main one is the execption | 14:08 |
| jroll | yeah | 14:09 |
| jroll | dtantsur: https://review.openstack.org/#/c/161453/16/ironic/dhcp/neutron.py check if None and raise InvalidParameter or whatever? | 14:09 |
| dtantsur | jroll, yeah | 14:09 |
| NobodyCam | jroll: None should be valid there | 14:09 |
| jroll | what | 14:09 |
| jroll | None is not valid there | 14:10 |
| devananda | rameshg87: fwiw, i think we could clean that up in liberty | 14:10 |
| NobodyCam | jroll: wont none be "If None, will get the list of ports from the Ironic port objects." | 14:10 |
| rameshg87 | devananda: okay | 14:11 |
| NobodyCam | or did I miss read that | 14:11 |
| jroll | NobodyCam: this is cleaning_network_uuid == None, when creating ports. it fails hard | 14:11 |
| NobodyCam | ahh I should have had more coffee.. | 14:12 |
| jroll | ugh, if I rebase this and submit, it's going to also rebase https://review.openstack.org/#/c/165528/7 | 14:13 |
| jroll | should I wait for that to merge before I submit this? | 14:13 |
| jroll | devananda: ^ | 14:13 |
| BadCub | ugh | 14:13 |
| jroll | there's a chance I can get it to not submit it but I doubt it | 14:13 |
| devananda | yes, wait | 14:14 |
| devananda | that one is almost done in the gte | 14:14 |
| jroll | just saw on the news you might get $10k if you were hit in the target data breach, jfyi | 14:14 |
| devananda | do not touch 162919 or 165528 right now | 14:14 |
| devananda | jroll: ^ | 14:14 |
| jroll | ok | 14:14 |
| jroll | dtantsur: NobodyCam: the rebase was clean and only left one definition of NodeCleaningError | 14:15 |
| NobodyCam | ya should be through the gate in just a couple of minutes if I read zuul right | 14:15 |
| jroll | so I think if we just land it, it will do the right thing | 14:15 |
| jroll | and if it doesn't we can fix it quickly | 14:15 |
| jroll | thoughts? | 14:16 |
| dtantsur | jroll, that's strange, because they seemingly add exceptions to different places in the file... | 14:16 |
| *** ijw has joined #openstack-ironic | 14:16 | |
| *** ijw has quit IRC | 14:16 | |
| dtantsur | I don't have a strong opinion on this :) | 14:16 |
| jroll | dtantsur: indeed... I just rebased on master and: | 14:16 |
| jroll | $ grep NodeClean ironic/common/exception.py | wc -l | 14:17 |
| *** ijw has joined #openstack-ironic | 14:17 | |
| jroll | 1 | 14:17 |
| dtantsur | lucasagomes, please see my comment on https://review.openstack.org/#/c/157715/21/ironic/drivers/modules/ilo/management.py | 14:18 |
| dtantsur | tl;dr we should not fill logs with warnings without a serious reason | 14:18 |
| lucasagomes | dtantsur, right, hmmm I still find it odd | 14:19 |
| lucasagomes | dtantsur, because the step is also optional, not the parameter | 14:19 |
| *** dhellmann has quit IRC | 14:19 | |
| lucasagomes | one can set level to 0 and don't run that step | 14:19 |
| *** dhellmann has joined #openstack-ironic | 14:19 | |
| lucasagomes | now the step is enabled but it doesn't have a parameter so it just skip with an info | 14:19 |
| lucasagomes | it's like you've two ways to disable the step | 14:20 |
| dtantsur | lucasagomes, I guess the use case is to _sometimes_ change password | 14:20 |
| lucasagomes | either set level to 0 or don't pass the parameter | 14:20 |
| dtantsur | not all-or-nothing thing | 14:20 |
| lucasagomes | right | 14:20 |
| jroll | I mean, the concept of resetting to the same password every time seems odd to me anyway | 14:20 |
| lucasagomes | yeah | 14:20 |
| jroll | because we aren't clearing it | 14:20 |
| lucasagomes | jroll, I wonder if we could rotate the password | 14:21 |
| lucasagomes | like create a random one | 14:21 |
| jroll | it's likely to make sure the tenant didn't change the password | 14:21 |
| openstackgerrit | Merged openstack/ironic: Generate keystone_authtoken options in sample config file https://review.openstack.org/162919 | 14:21 |
| jroll | lucasagomes: yeah, that's something I want to do in L | 14:21 |
| dtantsur | jroll, ++ | 14:21 |
| lucasagomes | and put it there, if it's not specified on that parameter | 14:21 |
| openstackgerrit | Merged openstack/ironic: Use task.spawn_after to maintain lock during cleaning https://review.openstack.org/165528 | 14:21 |
| devananda | aaaand there's the merges | 14:21 |
| devananda | woot | 14:21 |
| jroll | if the tenant changes the password you're totally screwed | 14:21 |
| jroll | and can't get to cleaning | 14:21 |
| devananda | jroll: ok, rebase or what ever. nothing of ours is in the gaet now | 14:21 |
| jroll | yep, on it | 14:22 |
| lucasagomes | dtantsur, jroll right, yeah I just pointed that warning/info thing | 14:22 |
| lucasagomes | it's ok being info | 14:22 |
| lucasagomes | I understand it better now | 14:22 |
| lucasagomes | (still find it a bit odd, perhaps the message should be "skipping the step X because ..." instead of "missing parameter" | 14:23 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add whole disk image support for iscsi_ilo using agent ramdisk https://review.openstack.org/165847 | 14:25 |
| NobodyCam | awesome ty jroll :) | 14:25 |
| rameshg87 | devananda: BadCub: sorry, i missed a small vector for ilo driver ^^^ | 14:25 |
| rameshg87 | devananda: BadCub: please accept if possible (it's a 2 line change in ilo driver | 14:26 |
| rameshg87 | devananda: BadCub: same change i made for dib ramdisk deployment (but at that time my agent patch wasn't merged) | 14:26 |
| rameshg87 | devananda: BadCub: and missed to do this in both the patches :( | 14:26 |
| *** hj-hp has joined #openstack-ironic | 14:27 | |
| BadCub | will take a look rameshg87 :-) | 14:27 |
| devananda | rameshg87: reasonable to fix this after feature freeze? | 14:29 |
| devananda | looks that way to me | 14:29 |
| *** jcoufal_ has joined #openstack-ironic | 14:29 | |
| devananda | rameshg87: pls file a bug, tag it on the commit, and target to kilo rc1 | 14:29 |
| rameshg87 | devananda: fine with me .. | 14:29 |
| rameshg87 | devananda: thanks :) | 14:29 |
| devananda | we'll get to it soon :) | 14:29 |
| rameshg87 | i will W-1 for now ? | 14:29 |
| rameshg87 | i mean shall i W-1 for now ? | 14:29 |
| devananda | rigt now the gate is actually our limit. fewer patches that we approve == faster merging for all of openstack | 14:29 |
| devananda | rameshg87: nah. | 14:30 |
| *** hj-hp has quit IRC | 14:30 | |
| devananda | just track it on LP | 14:30 |
| devananda | we should be ignoring patches not targeted to k3 at the moment | 14:30 |
| lucasagomes | I want to push the +a button for https://review.openstack.org/#/c/143919/ in case someone wants to take a look before I do it | 14:30 |
| lucasagomes | that bug is annoying and hard to debug | 14:30 |
| * lucasagomes hit that once | 14:30 | |
| *** jcoufal has quit IRC | 14:32 | |
| BadCub | is anyone looking at: https://review.openstack.org/#/c/155900/ Automate UEFI iso boot creation for iscsi_ilo driver? | 14:32 |
| *** absubram has joined #openstack-ironic | 14:32 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Documentation update https://review.openstack.org/165852 | 14:33 |
| NobodyCam | BadCub: I can look at that one | 14:34 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 14:34 |
| BadCub | thnx NobodyCam :-) | 14:34 |
| jroll | NobodyCam: dtantsur: ^^ | 14:34 |
| dtantsur | jroll, ack | 14:34 |
| devananda | jroll: i ws about to ask how soon those would be up | 14:34 |
| jroll | NobodyCam: dtantsur: it's still adding the exception... I can't find where that was added | 14:34 |
| NobodyCam | will look back in a sec ... TY jroll | 14:34 |
| dtantsur | jroll, https://review.openstack.org/#/c/157715/21/ironic/common/exception.py | 14:34 |
| jroll | the agent_ilo cleaning patch was rebased on top of this one, I wonder if that's where it was | 14:34 |
| devananda | lucasagomes: it's a bug. pls wait until after FF | 14:35 |
| *** hj-hp has joined #openstack-ironic | 14:35 | |
| jroll | dtantsur: oh, I need to depend on that patch eh | 14:35 |
| dtantsur | jroll, yes please | 14:35 |
| lucasagomes | devananda, fair enuff | 14:35 |
| *** achanda has joined #openstack-ironic | 14:35 | |
| devananda | literally folks, stop approving non-targeted ptaches right now ... | 14:35 |
| dtantsur | ack | 14:35 |
| lucasagomes | devananda, https://review.openstack.org/#/c/154808/48/ironic/drivers/modules/pxe.py | 14:35 |
| tiagogomes_ | Hi, I am trying to use Ironic, but when I try to add a node with ` ironic node-update $NODE_UUID add properties/cpus=2` I get a "Value: '2'. Wrong type" error. This happens with the ironic python client 0.4.1. Any knows what should I feed in properties/cpu? | 14:35 |
| lucasagomes | it's changing the default value of config options :/ | 14:35 |
| lucasagomes | I wonder if it's a concern here, cause it may break things | 14:36 |
| devananda | argh | 14:36 |
| lucasagomes | it doesn't seem to have a clear path to update | 14:36 |
| lucasagomes | from Juno to Kilo | 14:36 |
| jroll | tiagogomes_: that's a bad bug we need to fix. use properties/cpus="\"2\"" | 14:36 |
| lucasagomes | it's changing the bootloader, changing the default template | 14:36 |
| jroll | tiagogomes_: (if I remember correctly) | 14:37 |
| devananda | lucasagomes: adding a new one, or changing it??? | 14:37 |
| lucasagomes | and the pxe binary | 14:37 |
| devananda | stendulker: ^^ | 14:37 |
| lucasagomes | devananda, changing it | 14:37 |
| dtantsur | devananda, wdyt about https://review.openstack.org/#/c/165788/ ? should we: 1. land now, 2. land after FF, 3. land in L? | 14:37 |
| devananda | lucasagomes: for uefi only. not for legacy mode | 14:37 |
| lucasagomes | devananda, yes for uefi | 14:37 |
| lucasagomes | the patch is about security boot | 14:38 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 14:38 |
| jroll | dtantsur: fixed | 14:38 |
| dtantsur | thnx | 14:38 |
| tiagogomes_ | jroll it didn't worked, `Value: '1024'. Wrong type` | 14:38 |
| lucasagomes | but it was supported in Juno I think, but differently, I wonder if someone using the Juno version can update to Kilo with those changes | 14:38 |
| BadCub | thnx jroll :-) | 14:38 |
| stendulker | devananda: hi | 14:38 |
| NobodyCam | BadCub: 155900 +2'd | 14:39 |
| devananda | stendulker: hi. why is https://review.openstack.org/#/c/154808/48/ironic/drivers/modules/pxe.py,cm changing the default bootloader? | 14:39 |
| jroll | tiagogomes_: do the same for 1024 :( | 14:39 |
| BadCub | thnx NobodyCam :-) | 14:39 |
| BadCub | Anyone available to do a second +2/+A on https://review.openstack.org/#/c/155900/ ? | 14:39 |
| devananda | lucasagomes: fwiw, it's also deleting the old and adding the new boot loader template | 14:40 |
| stendulker | devananda: the default signed bootloader shipped by linux vendors is grub2 | 14:40 |
| lucasagomes | devananda, yeah | 14:40 |
| devananda | lucasagomes: seems like it is really s/elilo/grub2/ | 14:40 |
| lucasagomes | stendulker, there's an upgrade path form Juno to Kilo for that? | 14:40 |
| stendulker | lucasgomes: yes | 14:40 |
| tiagogomes_ | jroll ah, I didn't realize that failed in the next field. Thanks! | 14:40 |
| stendulker | lucagomes: for pxe, i have handles it | 14:40 |
| lucasagomes | stendulker, right, but it still needs to update other stuff | 14:41 |
| lucasagomes | like DHCP server | 14:41 |
| jroll | tiagogomes_: np :) | 14:41 |
| devananda | something which we all should have caught a long time ago -- the commit message on that ptach doesn't say ANYTHING about what it does | 14:41 |
| lucasagomes | that before was pointing to elilo now it should point to the grub binary | 14:41 |
| lucasagomes | stendulker, how is that done? | 14:41 |
| devananda | like, seriously. a 1 line commit message for 600 lines of code | 14:41 |
| stendulker | lucasgomes: yes, all documented in the install.rst | 14:41 |
| devananda | how did none of us -1 for that? | 14:41 |
| lucasagomes | stendulker, that for new deployments | 14:41 |
| lucasagomes | it's not an upgradepath | 14:41 |
| lucasagomes | stendulker, if I had a machine deployed with Juno elilo uefi security boot | 14:42 |
| NobodyCam | devananda: which # | 14:42 |
| lucasagomes | I update it to Kilo | 14:42 |
| devananda | stendulker: lucas is asking about users who have deployed ironic's Juno release and are using this already | 14:42 |
| lucasagomes | stendulker, will that work? | 14:42 |
| devananda | NobodyCam: https://review.openstack.org/#/c/154808/48 | 14:42 |
| lucasagomes | it seems not, it seems we are breaking things here | 14:42 |
| *** ramineni1 has joined #openstack-ironic | 14:43 | |
| *** ramineni has quit IRC | 14:43 | |
| rameshg87 | devananda, lucasagomes: it will still continue boot i guess | 14:43 |
| stendulker | lucasgomes: in upgrade, only active nodes would continue to use the elilo config file as long as they are not deleted | 14:43 |
| rameshg87 | devananda: lucasagomes: the elilo config files will still exist for old instances | 14:43 |
| *** achanda has quit IRC | 14:43 | |
| devananda | rameshg87: no, it's deleted | 14:43 |
| rameshg87 | devananda: only in tear down | 14:43 |
| lucasagomes | rameshg87, stendulker right... we have to think that, take over happens | 14:43 |
| stendulker | lucasgomes: the new nodes would use grub config file | 14:43 |
| lucasagomes | a new conductor that doesn't understand old deployed nodes | 14:43 |
| devananda | rameshg87: well. the new code won't know to delete the old elilo file, so tear own is going to fail .... | 14:43 |
| lucasagomes | is also a breakage here | 14:43 |
| lucasagomes | an old elilo machien can be mapped to a new conductor that doesn't support it | 14:44 |
| stendulker | devananda: new code will delete elilo file during teardown | 14:44 |
| lucasagomes | or it will just work with grub ? | 14:44 |
| stendulker | devananda: new code will delete elilo config file during teardown | 14:44 |
| rameshg87 | devananda: it knows in tear down https://review.openstack.org/#/c/154808/48/ironic/common/pxe_utils.py L244-248 | 14:44 |
| stendulker | devananda: without any exception | 14:44 |
| rameshg87 | but takeover of active nodes is an issue | 14:45 |
| rameshg87 | but we can't take over active nodes properly anyway i think :) | 14:45 |
| lucasagomes | hmm why? | 14:45 |
| devananda | rameshg87: ah | 14:45 |
| stendulker | rameshg87: how do things happen during takeover? | 14:45 |
| rameshg87 | i mean prior to kilo active instances, we can't take over because we don't have root_uuid saved in driver internal info | 14:45 |
| rameshg87 | lucasagomes: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/pxe.py#L439-L452 | 14:46 |
| *** ramineni1 has quit IRC | 14:46 | |
| rameshg87 | stendulker: ^^ this is why | 14:47 |
| lucasagomes | rameshg87, yeah | 14:47 |
| rameshg87 | so i guess take over will fail anyway - i mean continue to work as long as node is booted, but will fail on reboot | 14:47 |
| *** ramineni has joined #openstack-ironic | 14:48 | |
| stendulker | rameshg87: my question was wrt bootloader | 14:48 |
| lucasagomes | rameshg87, right I understand that... but that's 2 different problems | 14:48 |
| rameshg87 | stendulker: yeah, even if we didn't have boot loader in place, we wouldn't be able to switch | 14:48 |
| devananda | lucasagomes: so the commit message being not helpful aside, let's land this and sort it out later, as it sounds like relatively minor issue | 14:48 |
| rameshg87 | lucasagomes: yeah these are 2 different issues :) | 14:48 |
| rameshg87 | lucasagomes: but same result :) | 14:49 |
| lucasagomes | devananda, right | 14:49 |
| lucasagomes | yeah I don't mind, but I wanted to clarify wether we are aware that this can break already running nodes | 14:49 |
| *** timbyr_ has joined #openstack-ironic | 14:49 | |
| *** timbyr_ has quit IRC | 14:49 | |
| rameshg87 | lucasagomes: i guess it doesn't break anything than what's available in ironic today | 14:50 |
| NobodyCam | devananda: your reviewing 154808 now? | 14:50 |
| lucasagomes | devananda, I'm finishing reviewing it 1 sec | 14:50 |
| rameshg87 | lucasagomes: ironic can take over active nodes prior to kilo, but will fail on reboot - this will do exactly the same :) | 14:50 |
| stendulker | lucasgomes: active nodes would continue to boot with elilo | 14:50 |
| openstackgerrit | Imre Farkas proposed openstack/ironic: DRAC: add retry capability to wsman client operations https://review.openstack.org/165860 | 14:50 |
| lucasagomes | rameshg87, yeah I get that, I understand the overall result is the node won't boot | 14:51 |
| lucasagomes | but the causes are different | 14:51 |
| stendulker | lucasgomes: new nodes would use grub2. active node teardown would be clean | 14:51 |
| rameshg87 | yeah :( | 14:51 |
| lucasagomes | one have been fixed in Kilo, the other have been introduced in kilo | 14:51 |
| lucasagomes | stendulker, sure yeah... it's a bit more complex tho. If one is using a custom template in Juno and it's elilo | 14:52 |
| lucasagomes | updating Ironic won't change that config option | 14:52 |
| lucasagomes | it will change the default to grub which is not used anymore | 14:52 |
| *** rloo_ has joined #openstack-ironic | 14:52 | |
| lucasagomes | and new nodes will fail even to boot the first time | 14:52 |
| lucasagomes | I understand upgrade is complex, it's a complex beast | 14:52 |
| lucasagomes | but we should think about it | 14:52 |
| stendulker | lucasgomes: yes | 14:52 |
| lucasagomes | tihs patch, opens many possibilities for users to not be able to upgrade from Juno to Kilo | 14:53 |
| *** rloo_ has quit IRC | 14:53 | |
| lucasagomes | that's my point here | 14:53 |
| lucasagomes | we should avoid that | 14:53 |
| devananda | lucasagomes: oh. i see | 14:53 |
| lucasagomes | or offer a document, explaning in details | 14:53 |
| lucasagomes | what you have to do | 14:53 |
| lucasagomes | for existing nodes to go from A to B | 14:53 |
| *** rloo_ has joined #openstack-ironic | 14:53 | |
| lucasagomes | we dont' wannt break users every 6 months, otherwise Openstack would be a complete failure | 14:53 |
| *** rloo has quit IRC | 14:53 | |
| lucasagomes | if we do that | 14:53 |
| stendulker | lucagomes: have captured config file changes in install.rst. Will that info be useful? | 14:53 |
| jroll | lucasagomes: openstack tends to do that :P | 14:54 |
| lucasagomes | jroll, :) you said once that it;s not because the others sucks we have to suck too | 14:54 |
| jroll | then again, deploying six months of change at once is insane :) | 14:54 |
| lucasagomes | right we gotta try our best | 14:54 |
| jroll | lucasagomes: yeah, not saying we should do it :D | 14:54 |
| jroll | be the change you want to see etc | 14:55 |
| lucasagomes | stendulker, it helps, we can even put it in a wiki or something | 14:55 |
| *** romcheg1 has joined #openstack-ironic | 14:55 | |
| lucasagomes | have a DocImpact flag there pointing to it about upgrades | 14:55 |
| *** romcheg has quit IRC | 14:55 | |
| stendulker | lucasgomes: ok. i can do that | 14:55 |
| lucasagomes | not saying we should not land this, but we need to be more aware on breakages | 14:55 |
| stendulker | lucasgomes: i understand. but we should go for grub2 as it is kind of default bootloader used by most linux distributors | 14:56 |
| stendulker | lucasgomes: i mean we should move towards grub2 | 14:56 |
| lucasagomes | right (: yeah it's fine | 14:57 |
| lucasagomes | jroll, +1 D: | 14:57 |
| lucasagomes | :D* | 14:57 |
| jroll | lucasagomes: to me, that means release more than every six months | 14:58 |
| jroll | but I might have to fight a PTL for that | 14:58 |
| devananda | jroll: not at all | 14:59 |
| devananda | jroll: at least not this one | 14:59 |
| lucasagomes | release early release often :) | 14:59 |
| jroll | devananda: maybe push one then :P | 14:59 |
| lucasagomes | but yeah you might need to fight a board for that | 14:59 |
| * dtantsur likes 3 months | 14:59 | |
| jroll | devananda: is this something we're going to talk about in canadia or jfdi | 14:59 |
| *** rloo_ is now known as rloo | 14:59 | |
| NobodyCam | dtantsur: your ok with addressing the assert_called_once comments on 161453 with a follow up? | 15:00 |
| devananda | jroll: nope. I'm calling it now. | 15:00 |
| dtantsur | NobodyCam, yes | 15:00 |
| NobodyCam | awesome ;) | 15:00 |
| jroll | devananda: dtantsur: come to think about it, I actually don't care what the cycle is, as long as there's no feature freeze. because those suck. | 15:00 |
| jroll | devananda: orly, you should email a list or something :) | 15:00 |
| dtantsur | jroll, 1-2 week of soft FF might be pretty helpful | 15:01 |
| devananda | jroll: stable branches, it turns out, are important though | 15:01 |
| jroll | stable branches are fine | 15:01 |
| rloo | jroll: please do me a favour and take a look at my comments in revision 16 of https://review.openstack.org/#/c/161453/ | 15:01 |
| devananda | and so far in openstacklandia, feature freeze has been the way the project-as-a-whole makes a stable point in time | 15:01 |
| lucasagomes | jroll, right, I also don't like FF as an engineer | 15:01 |
| dtantsur | jroll, (by soft I mean freeze core things, by less strict about vendor and low-impact stuff) | 15:01 |
| lucasagomes | but other people still depends on it | 15:01 |
| lucasagomes | doc people for example | 15:01 |
| devananda | because we couldn't do stability cyucles and no one cares enough about bugs when there are shiney features to work on | 15:01 |
| lucasagomes | or i18n | 15:01 |
| devananda | </me stops ranting> | 15:02 |
| rloo | jroll: i think they may all be nits and maybe all of it can be addressed later. but just in case | 15:02 |
| lucasagomes | and what devananda said, for stability | 15:02 |
| rloo | we're probably all in a grumpy mood but really, if there wasn't FF, when would people actually do the reviews? | 15:02 |
| devananda | so - I know everyone's been pushing really hard this week -- and thank you. I'd like us to never have to do this again, but there will eventually be another reason / deadline / critical bug / thing that makes us stressed out | 15:03 |
| devananda | feature freeze isn't what causes that in and of itself, it's how we approached it // leaving so much to the last week | 15:04 |
| jroll | rloo: looking | 15:04 |
| rloo | devananda: +2 | 15:04 |
| jroll | devananda: that's inevitable, I think | 15:05 |
| * rloo also wishes we had more non-core-reviewers, reviewing | 15:05 | |
| jroll | devananda: because we don't let people land features for 2+ months, everybody is like OMG I NEED TO GET THIS IN SOON | 15:05 |
| BadCub | lucasagomes: were you on 154808? | 15:06 |
| lucasagomes | BadCub, yeah | 15:06 |
| rloo | i think at the summit, we should have some sort of post-mortem/whatever you call it, to see about if we can improve our process | 15:06 |
| lucasagomes | BadCub, finishing | 15:06 |
| BadCub | awesome thank you :-) | 15:06 |
| devananda | rloo: yup. definitely | 15:06 |
| jroll | rloo: there's one bug that worries me there, but I think we can fix it later, it's an edge case | 15:07 |
| devananda | jroll: let's make that window smaller then | 15:07 |
| rloo | jroll: ok. i'll leave it for you guys to follow up. | 15:07 |
| BadCub | process can always be improved :-) | 15:07 |
| rloo | jroll: i guess i can +2 it then. | 15:07 |
| devananda | also, folks, we've got nothing in the gate right now, and the gate is backed up | 15:07 |
| jroll | JoshNang: you should check out rloo's commends on 161453 patchset 16 | 15:07 |
| rameshg87 | jroll: some comments on 161453 | 15:08 |
| jroll | when you're around | 15:08 |
| dtantsur | devananda, if you mean smaller feature window, that will make us work more downstream, I'm afraid... | 15:08 |
| rloo | jroll: should I +2 or wait for JoshNang to look? | 15:08 |
| rameshg87 | jroll: but nothing serious, i don't know if we have plan to merge this (if not serious issues) and then do rest in follow-up patch | 15:08 |
| jroll | rloo: nah, go ahead, I just don't want him to miss them :) | 15:09 |
| rloo | jroll: ok. | 15:09 |
| devananda | I'd like to get a vote from everyone on a few things | 15:09 |
| *** kkoski has joined #openstack-ironic | 15:09 | |
| NobodyCam | rloo: are you +a'ing too | 15:09 |
| devananda | since we've all bee looking really hard at the last remaining 4 BP (which is really 2 features) | 15:09 |
| rloo | NobodyCam: I can +A too unless I shouldn't or if you want to? | 15:10 |
| devananda | we can get a feature freeze exception, if something's going to land, say, in the next couple days, even after cutting the milestone now | 15:10 |
| jroll | rameshg87: we need to land this :P | 15:10 |
| devananda | knowing how that affects testers, i18n, doc folks, etc ... it's really a last resort | 15:10 |
| rameshg87 | jroll: done +2 | 15:11 |
| rloo | except for the main cleaning stuff, what is that important to ask for FFE? | 15:11 |
| *** Marga_ has joined #openstack-ironic | 15:11 | |
| devananda | rloo: right | 15:11 |
| devananda | rloo: there are two other BP up there still, so I want a vote on those, because cleaning is a no-brainer. We all NEED that this cycle | 15:11 |
| devananda | vote +1/-1/abstain for uefi secure boot work | 15:12 |
| BadCub | -1 | 15:12 |
| devananda | (giving 2 minutes for anyone here to cast a vote) | 15:12 |
| rameshg87 | +0 (i belong to same team) :( | 15:12 |
| rloo | devananda: maybe you need 2 core reviewers to offer to see them through? | 15:12 |
| rloo | abstain | 15:12 |
| devananda | rameshg87: that's ok - you can vote too | 15:12 |
| devananda | it's a question not of "how good" but "how ready is it" | 15:12 |
| * rloo changes her vote then. -1 | 15:13 | |
| devananda | lucasagomes jroll NobodyCam dtantsur Shrews ? | 15:13 |
| * lucasagomes reads | 15:14 | |
| * NobodyCam was looking at the patches | 15:14 | |
| * dtantsur translates "abstain| | 15:14 | |
| rameshg87 | devananda: i don't know if my vote will be counted impartial, but +1 from me because 1) it was posted long back, but no reviews 2) afaik it is tested with every rebase 3) it's only in ilo driver 4) 1 patch has already landed | 15:14 |
| * rloo is going to change again cuz not having looked at all the patches, she doesn't really know how ready it is. abstain. | 15:14 | |
| rameshg87 | devananda: so upto folks to count my vote or not (as i already said i belong to same team) | 15:14 |
| jroll | devananda: -0 | 15:14 |
| lucasagomes | devananda, +0 | 15:14 |
| devananda | rameshg87: https://review.openstack.org/#/c/154808/ affects more than just ilo driver code though | 15:15 |
| dtantsur | -1 | 15:15 |
| lucasagomes | I've voted +1 on that patch, but I have my concerns. I'm sure people will work on a document to minimize my concerns and all that | 15:15 |
| * devananda doesn't know how to interpret +0/-0 | 15:15 | |
| jroll | devananda: leaning toward -1 but not educated enough to have a real opinion | 15:16 |
| NobodyCam | I would like to have it, but believe my best vote for the project is -1 | 15:16 |
| devananda | but i've got two -1 and the rest abstain, so | 15:16 |
| lucasagomes | jroll, +1 same | 15:16 |
| jroll | (side note: I +2'd https://review.openstack.org/#/c/164646/ if anyone else wants to) | 15:16 |
| devananda | that makes 4. done | 15:16 |
| devananda | next vote | 15:16 |
| devananda | automate bios iso creation for ilo driver -- how ready is this to land? | 15:16 |
| devananda | +1 | 15:16 |
| * lucasagomes needs to look at the patch | 15:17 | |
| NobodyCam | I think that is a +1 | 15:17 |
| devananda | (i think it's simple code and contained in the driver) | 15:17 |
| BadCub | +1 | 15:17 |
| jroll | devananda: is that https://review.openstack.org/#/c/155900/ | 15:17 |
| NobodyCam | yep | 15:17 |
| devananda | yes | 15:17 |
| rloo | +1 | 15:17 |
| dtantsur | looks simple, +1 | 15:17 |
| jroll | it's got 2x+2, let's do it | 15:17 |
| devananda | ok, FFE granted | 15:17 |
| lucasagomes | heh yeah it has 2+2 so I assume it's good | 15:17 |
| NobodyCam | jroll: hit hte button | 15:17 |
| jroll | NobodyCam: I haven't reviewed it, I'm not pushing that button :) | 15:18 |
| NobodyCam | lol | 15:18 |
| NobodyCam | ;) | 15:18 |
| devananda | rather than assume, i'm going to ask | 15:18 |
| lucasagomes | yeah I reviewed that my concern was only about umounting the image | 15:18 |
| lucasagomes | which seems to be fixed | 15:18 |
| rloo | oh, devananda and NobodyCam +2'd that. I guess we want a non HP person? | 15:18 |
| lucasagomes | so +1 | 15:18 |
| NobodyCam | rloo: +++++ | 15:18 |
| rloo | NobodyCam: I can take a look at that patch but you know me, slow as molasses ;) | 15:19 |
| jroll | devananda: I'm going to assume that you're asking about cleaning things :P | 15:19 |
| devananda | next vote - agent cleaning (ready to land all the major work in next few days) +1/-1/abstain | 15:19 |
| dtantsur | devananda, +1 for cleaning :) | 15:19 |
| devananda | i'm clearly =1 on this | 15:19 |
| jroll | +1 | 15:19 |
| devananda | +1 | 15:19 |
| rloo | 1 | 15:19 |
| BadCub | +1 | 15:19 |
| NobodyCam | rloo: it takes time to be good! | 15:19 |
| jroll | last two patches just need +A | 15:19 |
| jroll | https://review.openstack.org/#/c/161453/ and https://review.openstack.org/#/c/164646/ | 15:19 |
| BadCub | 161453 got +A | 15:20 |
| lucasagomes | =1 | 15:20 |
| lucasagomes | +1* | 15:20 |
| devananda | rameshg87 dtantsur Shrews ? | 15:20 |
| devananda | dtantsur: oh nvm, see your vote now | 15:20 |
| dtantsur | :) | 15:20 |
| jroll | 164646 is pretty simple (and also a fun number to type) | 15:20 |
| NobodyCam | lol | 15:20 |
| NobodyCam | +1 on cleaning here | 15:20 |
| devananda | k, glad we're all in agreement there :) | 15:21 |
| jroll | rameshg87: any reason you're only +1 on 164646? | 15:21 |
| stendulker | devananda: regarding secure boot, can we have agent_ilo and iscsi_ilo drivers? | 15:21 |
| * NobodyCam brb | 15:21 | |
| rameshg87 | jroll: same team | 15:21 |
| devananda | next vote - iLO cleaning (ready to land all the major work along with the agent cleaning) +1/-1/abstain | 15:21 |
| rameshg87 | jroll: ramineni and i work in the same team | 15:21 |
| stendulker | devananda: they do not have any such dependency | 15:22 |
| devananda | stendulker: oh?? the patches are dependent | 15:22 |
| jroll | rameshg87: hm, ok, that's your call, but I personally don't have issues with +2'ing things from my team | 15:22 |
| devananda | stendulker: if agent-ilo and iscsi-ilo are rebased on master -- then yes. those are both very small patches | 15:22 |
| jroll | devananda: oh, I thought that was part of the last vote. is that just https://review.openstack.org/#/c/164646/ ? | 15:22 |
| BadCub | +1 | 15:22 |
| devananda | stendulker: assuming that they stay small patches and it's just a rebase, that is | 15:22 |
| dtantsur | rameshg87, FWIW I +2 things from Lucas and Imre (though try not to approve, if it's only 2 of us) | 15:22 |
| * rameshg87 upgrades the votes :) | 15:22 | |
| rloo | +1 because it is almost done | 15:22 |
| devananda | jroll: technically, ilo cleaning is a separate BP, so i'm opening a separate vote | 15:23 |
| rameshg87 | i mean that particular vote | 15:23 |
| jroll | ok | 15:23 |
| devananda | I'm still +1 because it's almost done and fairly straight forward | 15:23 |
| stendulker | devananda: ok | 15:23 |
| jroll | devananda: confirm that's 164646? | 15:23 |
| jroll | devananda: if so, +1 | 15:23 |
| devananda | jroll: correct | 15:23 |
| BadCub | yep | 15:23 |
| rameshg87 | +1 from me (fairly simple and takes most of the things from parent patch) | 15:23 |
| dtantsur | devananda, I guess we approved much part of iLO cleaning already, so +1 | 15:24 |
| dtantsur | * much = large | 15:24 |
| devananda | <end voting> | 15:24 |
| * devananda wishes there was a real bot here | 15:24 | |
| rameshg87 | :) | 15:24 |
| * devananda does his best roboto impression | 15:24 | |
| *** ChuckC has joined #openstack-ironic | 15:24 | |
| jroll | so today is now "watch for gate failure" day? | 15:25 |
| rloo | why do we need FFE? We'll get them all approved in the next 1.5 hours... | 15:25 |
| krotscheck | Domo ariwhatto? | 15:25 |
| devananda | krotscheck: :-) | 15:25 |
| devananda | rloo: if that's the case ... then yea, we don't | 15:25 |
| * rameshg87 googles what it is | 15:25 | |
| devananda | rloo: it sounds like we won't get the pxe_ilo secure boot done in tht time though | 15:25 |
| rloo | rameshg87: I *think* that was a japanese thingy. | 15:26 |
| rameshg87 | thanks a lot | 15:26 |
| lucasagomes | sounds japanese :) | 15:26 |
| rloo | rameshg87: like 'thank you' | 15:26 |
| * lucasagomes dunno the meaning tho | 15:26 | |
| stendulker | devananda: i would have take common code from ilo/deploy.py and deploy_utils.py | 15:26 |
| lucasagomes | oh thought it was arigato | 15:26 |
| jroll | lucasagomes: it is :P | 15:26 |
| lucasagomes | for thank you, anyway maybe it's just spelled diff | 15:26 |
| rloo | lucasagomes: yeah, it is. so some sort of joke ;) | 15:26 |
| lucasagomes | fair nuff | 15:26 |
| stendulker | devananda: How much time do i have to refactor these patches to support iscsi and agent_ilo? | 15:27 |
| lucasagomes | 164646 seems good I'm +a that | 15:27 |
| lucasagomes | concerns? | 15:27 |
| lucasagomes | 1 | 15:27 |
| lucasagomes | 2 | 15:27 |
| BadCub | thanks lucasagomes :-) | 15:27 |
| rameshg87 | devananda: if pxe_ilo is issue, then can we get iscsi_ilo and agent_ilo in . it doesn't have the concerns same as that of pxe_ilo and it is only in ilo driver | 15:27 |
| lucasagomes | done | 15:27 |
| BadCub | awesome! | 15:27 |
| lucasagomes | I have to jump on a call in 3 mins | 15:27 |
| lucasagomes | so brb | 15:27 |
| dtantsur | lucasagomes, 164646 -ENEEDREBASE | 15:27 |
| dtantsur | no? | 15:28 |
| dtantsur | oh, I'm going on that call too :D | 15:28 |
| lucasagomes | dtantsur, oh, yeah dependent patch seems OUTDATED | 15:28 |
| lucasagomes | if someone can rebase that and +a will be good | 15:28 |
| * lucasagomes will do it after the call if no one does | 15:28 | |
| dtantsur | lucasagomes, I'll try the rebase button :) | 15:29 |
| *** kkoski has quit IRC | 15:29 | |
| dtantsur | no, it can't be rebased automatically | 15:29 |
| dtantsur | :( | 15:29 |
| rameshg87 | lucasagomes: ok i can .. | 15:29 |
| dtantsur | then I'll leave it for later too | 15:29 |
| dtantsur | rameshg87, yes please | 15:29 |
| rameshg87 | lucasagomes: hold on rebasing | 15:29 |
| lucasagomes | rameshg87, yeah I won't now. I'm in a call | 15:29 |
| lucasagomes | rameshg87, feel free to rebase and +a it | 15:30 |
| rameshg87 | okay | 15:30 |
| devananda | stendulker: will iscsi_ilo and agent_ilo patches introduce changes to config options, new template file, etc? | 15:31 |
| stendulker | devananda: no | 15:32 |
| stendulker | devananda: DIB changes required for secure boot have laready landed | 15:33 |
| * rameshg87 runs tox on 164646 | 15:33 | |
| devananda | "largely" ? | 15:33 |
| BadCub | thnx rameshg87 :-) | 15:33 |
| stendulker | devananda: already. https://review.openstack.org/#/c/153987/ | 15:33 |
| devananda | stendulker: so https://review.openstack.org/#/c/154816/45 is already approved | 15:33 |
| rameshg87 | jroll: flake8 failure on https://review.openstack.org/#/c/161453/18/ironic/tests/dhcp/test_neutron.py L401 | 15:34 |
| jroll | WAT | 15:34 |
| rameshg87 | jroll: forgot to do api = dhcp_factory.DHCPFactory().provider | 15:34 |
| jroll | oh ffs | 15:34 |
| jroll | thanks | 15:34 |
| stendulker | devananda: so i need to refactor all the common code for ilo drivers into iscsi patch | 15:35 |
| rameshg87 | jroll: okay, i will wait | 15:35 |
| stendulker | devananda: iscsi_ilo patch https://review.openstack.org/#/c/154814/ | 15:35 |
| rameshg87 | rloo: can you remove +A from https://review.openstack.org/#/c/161453/18 | 15:35 |
| rameshg87 | rloo: there is a flake8 issue | 15:36 |
| rameshg87 | rloo: new patch will be put soon | 15:36 |
| jroll | rameshg87: it can be left | 15:36 |
| jroll | jenkins does the right thing | 15:36 |
| rameshg87 | jroll: oh okay | 15:36 |
| NobodyCam | yea it will fail and we can reapprove | 15:36 |
| rloo | rameshg87: yeah, what jroll said ;) | 15:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 15:36 |
| devananda | stendulker: I just +2'd that. rather than refactoring into the iscsi_ilo patch, you should make a fourth patch that is just the common code | 15:36 |
| jroll | pushing now | 15:36 |
| jroll | also running tests locally | 15:36 |
| devananda | stendulker: then rebase iscsi_ilo and agent_ilo on top of that common patch | 15:36 |
| devananda | stendulker: and set the pxe_ilo one aside for now | 15:37 |
| stendulker | devananda: ok. thank you for considering the request | 15:37 |
| * Shrews checks scrollback | 15:37 | |
| *** lazy_prince is now known as killer_prince | 15:37 | |
| * NobodyCam waits for retest on 161066 | 15:38 | |
| BadCub | devananda: so this would basically remove the dependency for pxe_ilo from 154814? | 15:38 |
| devananda | BadCub: that is what stendulker has claimed | 15:40 |
| BadCub | kk | 15:40 |
| *** vdrok is now known as vdrok_afk | 15:40 | |
| devananda | stendulker: honestly, i'm still somewhat uncomfortable with this -- we're talking a refactoring and new code being pushed up after FF | 15:40 |
| JoshNang | morning y'all | 15:41 |
| devananda | stendulker: but the two smaller patches (agent_ilo and iscsi_ilo) are both fairly trivial -- by themselves, i would have no problem | 15:41 |
| jroll | JoshNang: mornin' | 15:41 |
| NobodyCam | morning JoshNang :) | 15:41 |
| BadCub | JoshNang: mornin! :-) | 15:41 |
| *** stendulker_ has joined #openstack-ironic | 15:41 | |
| JoshNang | looks like everyone's been busy :) | 15:42 |
| *** ramineni has quit IRC | 15:42 | |
| jroll | nahhhhhhh | 15:42 |
| jroll | business as usual | 15:42 |
| rameshg87 | JoshNang: o/ | 15:42 |
| devananda | stendulker: so far I don't see any dependencies in those two patches to call out as problems. how much time do you need for an initial go at refactoring / rebasing? | 15:42 |
| jlvillal | Good morning all. | 15:43 |
| jroll | hey jlvillal :) | 15:43 |
| NobodyCam | morning jlvillal | 15:43 |
| BadCub | mornin jlvillal | 15:43 |
| jlvillal | Hey! :) | 15:43 |
| *** hemna has joined #openstack-ironic | 15:44 | |
| BadCub | brb | 15:44 |
| NobodyCam | no Nisha this morning? | 15:44 |
| *** stendulker has quit IRC | 15:44 | |
| dtantsur | JoshNang, morning | 15:45 |
| rameshg87 | NobodyCam: anything i can help with ? i guess she's not there | 15:45 |
| jlvillal | lucasagomes: Thanks for the tip on removing trailing whitespace. As a vim user I thank you :) | 15:45 |
| NobodyCam | rameshg87: quick question, you may know the answer to, on 165310, looks like a test was removed? | 15:46 |
| lucasagomes | jlvillal, :D vim ftw! np | 15:46 |
| * rameshg87 checks | 15:46 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver https://review.openstack.org/154808 | 15:46 |
| jlvillal | lucasagomes: Do you use vim-fugitive? My tip for vim users :) | 15:46 |
| jlvillal | lucasagomes: It is awesome! :D | 15:46 |
| lucasagomes | oh I don't I will take a look after the call :) | 15:46 |
| lucasagomes | thanks! | 15:46 |
| rameshg87 | NobodyCam: this one - https://review.openstack.org/#/c/165310/5/ironic/tests/drivers/ilo/test_inspect.py ? | 15:47 |
| rameshg87 | NobodyCam: i don't see any tests removed | 15:47 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Support agent_ilo driver to perform cleaning https://review.openstack.org/164646 | 15:47 |
| NobodyCam | rameshg87: https://review.openstack.org/#/c/165310/4..5/ironic/tests/drivers/ilo/test_inspect.py | 15:47 |
| rameshg87 | jroll: dtantsur: lucasagomes: anyone wants to +A ? | 15:48 |
| rameshg87 | ^^^^ | 15:48 |
| * dtantsur is on a call | 15:48 | |
| rameshg87 | just ironic.conf.sample rebased | 15:48 |
| jroll | rameshg87: /me looks | 15:48 |
| jroll | bam | 15:48 |
| rameshg87 | jroll: thanks :) | 15:49 |
| * rameshg87 back to NobodyCam again | 15:49 | |
| NobodyCam | for that to land we also need https://review.openstack.org/#/c/161453 | 15:49 |
| rameshg87 | NobodyCam: i haven't been following up on this | 15:50 |
| rameshg87 | NobodyCam: i will have to read through comments to figure out :( | 15:50 |
| rameshg87 | NobodyCam: but diff with Base shows one test was added | 15:50 |
| NobodyCam | rameshg87: ack... looks like tere was a test for add to the dict that get removed between ps 4 and 5 | 15:50 |
| *** Nisha has joined #openstack-ironic | 15:51 | |
| rameshg87 | Nisha: right on time :) | 15:51 |
| rameshg87 | NobodyCam: Nisha is here | 15:51 |
| NobodyCam | morning Nisha | 15:51 |
| NobodyCam | hehehe | 15:51 |
| NobodyCam | we were just looking at 161310 | 15:51 |
| jlvillal | For vim users: lucasagomes: https://github.com/tpope/vim-fugitive | 15:52 |
| Nisha | NobodyCam, ok | 15:52 |
| Nisha | NobodyCam, morning | 15:52 |
| rameshg87 | Nisha: NobodyCam was asking why a test was removed https://review.openstack.org/#/c/165310/4..5/ironic/tests/drivers/ilo/test_inspect.py | 15:52 |
| NobodyCam | :) | 15:52 |
| rameshg87 | between patches 4 and 5 | 15:52 |
| jroll | anybody going to +A https://review.openstack.org/#/c/161453 | 15:52 |
| Nisha | rameshg87, let me see. | 15:52 |
| jroll | that would be awesome | 15:52 |
| jroll | jlvillal: :Gblame is the best. | 15:52 |
| NobodyCam | looks like you kept the replace test but removed the add test | 15:52 |
| jlvillal | jroll: My favorite feature. I don't use the other features that much. But stepping back in time is great. | 15:53 |
| jroll | rameshg87: tests passed locally on 161453 for me | 15:53 |
| rameshg87 | jroll: yeah me too :) | 15:53 |
| rameshg87 | jroll: i was anyway running on top of your patch | 15:53 |
| jroll | ah ok | 15:54 |
| rameshg87 | jroll: that's why i didn't raise Exception this time :) | 15:54 |
| Nisha | oh i think it happened by mistake | 15:54 |
| jroll | rloo: are you reviewing https://review.openstack.org/#/c/155900/ ? | 15:55 |
| Nisha | NobodyCam, rameshg87 I will just upload the patch again :( | 15:55 |
| rloo | jroll: yeah. | 15:55 |
| Shrews | jroll: looking @ 161453 | 15:55 |
| jroll | ok, thanks | 15:55 |
| *** kozhukalov has joined #openstack-ironic | 15:55 | |
| NobodyCam | i see a nit on 161453 | 15:55 |
| rloo | jroll: you're welcome to review too | 15:55 |
| jroll | NobodyCam: so many nits | 15:56 |
| *** kkoski has joined #openstack-ironic | 15:56 | |
| jroll | rloo: I'm stepping away for a bit but I can when I come backif you want more eyes on it | 15:56 |
| * rameshg87 goes for dinner | 15:56 | |
| rloo | jroll: ok | 15:57 |
| *** rameshg87 is now known as rameshg87-dinner | 15:57 | |
| JoshNang | NobodyCam: feel free to nit, i'll work on the follow-on nit squashing patch today | 15:57 |
| Shrews | JoshNang: was the rollback stuff in neutron.py really necessary seeing that prepare_cleaning() always calls delete_cleaning_ports() before create_cleaning_ports()? | 15:59 |
| Shrews | JoshNang: for 161453, that is | 15:59 |
| JoshNang | Shrews: maybe, maybe not, i thought it was cleaner (heh) to clean up there, because who knows, the operator might go and do something else with the node after cleanfail, rather than going directly back into cleaning | 16:00 |
| Shrews | JoshNang: true | 16:01 |
| *** 21WABNX9V has quit IRC | 16:01 | |
| *** trown|call is now known as trown | 16:02 | |
| *** hj-hp has quit IRC | 16:03 | |
| Shrews | NobodyCam: want to give the second +2 on 161453, or should I just approve it? | 16:05 |
| BadCub | Shrews: go ahead and +A it :-) | 16:05 |
| Shrews | BadCub: as you command | 16:06 |
| BadCub | thank ya kindly sir :-) | 16:06 |
| *** jcoufal_ has quit IRC | 16:08 | |
| BadCub | BRB | 16:08 |
| *** wanyen has joined #openstack-ironic | 16:09 | |
| NobodyCam | JoshNang: commented for the follow up | 16:10 |
| JoshNang | NobodyCam: thanks! | 16:10 |
| *** rameshg87-dinner is now known as rameshg87 | 16:11 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: follow up patch for ilo capabilities https://review.openstack.org/165310 | 16:13 |
| Nisha | NobodyCam, ^^^ | 16:13 |
| Nisha | reposted the patch | 16:13 |
| Nisha | rameshg87, ^^^ | 16:13 |
| rameshg87 | Nisha: ack | 16:13 |
| dtantsur | devananda, I would argue that https://review.openstack.org/#/c/165860/ is something we can land after FF, even with hardcoding values instead of creating config options | 16:13 |
| rameshg87 | devananda: i see you are -2ing all patches which are not going in | 16:14 |
| dtantsur | looks more like bug fix | 16:14 |
| devananda | indeed | 16:14 |
| rameshg87 | devananda: do you want me to remove my -2 on https://review.openstack.org/#/c/159734/ | 16:14 |
| rameshg87 | devananda: and want to put your -2 | 16:14 |
| rameshg87 | ? | 16:14 |
| devananda | rameshg87: sure | 16:14 |
| rameshg87 | okay removing mine | 16:14 |
| rameshg87 | done | 16:14 |
| *** aswadr has quit IRC | 16:14 | |
| devananda | dtantsur: i think you could easily win that argument :) | 16:14 |
| dtantsur | :) | 16:15 |
| rameshg87 | dtantsur: exactly why we need this to be out-of-tree ? :) | 16:15 |
| devananda | dtantsur: file a bug and update the patch (or let ifarkas know, since I dont see him in IRC) and I'll happily lift that. also, no rush -- ping me in a couple days about it, too | 16:15 |
| devananda | rameshg87: ++ | 16:15 |
| rameshg87 | dtantsur: so that we can fix bugs independent of ironic feature freezes | 16:16 |
| dtantsur | I guess ifarkas is on IRC :) ifarkas? | 16:16 |
| dtantsur | rameshg87, well yeah, but it's not an easy task as well | 16:16 |
| dtantsur | definitely feature freeze is not great time for new dependencies ;) | 16:16 |
| rameshg87 | yeah | 16:16 |
| rloo | I've +2'd https://review.openstack.org/#/c/155900/ and it modifies modules/ilo/deploy.py. Don't know if there are other patches that modify that file. Should I check, or just +A? | 16:17 |
| devananda | oh! tabcomplete failed me. he is here | 16:17 |
| devananda | rameshg87: so bugs can continue to be fixed during FF | 16:17 |
| BadCub | rloo: thank you :-) | 16:17 |
| devananda | rameshg87: that is the whole point of it :) | 16:17 |
| * dtantsur starts planning feature freeze for ironic-discoverd... | 16:18 | |
| devananda | dtantsur: do it if and only if it makes sense for discoverd | 16:18 |
| devananda | dtantsur: right now, you aren't tied to the coordinated / integrated release ... thta's probably a good thing for discoverd :) | 16:19 |
| lucasagomes | jlvillal, looks awesome! (the vim git wrapper) | 16:19 |
| rloo | BadCub: I've +A'd it. Looks like there aren't any more outstanding ilo changes except for the secure boot stuff. | 16:19 |
| lucasagomes | will give it a go | 16:19 |
| devananda | dtantsur: fwiw, neither is ipa | 16:19 |
| * rameshg87 checks | 16:19 | |
| * rameshg87 reads back | 16:19 | |
| BadCub | rloo: awesomeness! Thank you!! :) | 16:19 |
| dtantsur | devananda, I actually want to slow down pushing more stuff to discoverd for some time... not sure this time is _now_ though :) | 16:19 |
| NobodyCam | rloo: I believe other patches that modify that (the Uefi patch set) got bumped | 16:19 |
| dtantsur | devananda, currently it's in the shape of "nothing done for the release" https://launchpad.net/ironic-discoverd/+milestone/1.1.0 :) | 16:20 |
| BadCub | anyone want to take a look at: https://review.openstack.org/161066 Add cleaning/zapping support to IPA ? It has two +2s on it | 16:21 |
| JayF | devananda: Should we start lining up ducks to do a first agent release? | 16:22 |
| JayF | IDK if we want to land that with/before K or if we want to just let the big release blow over then cut an IPA after | 16:22 |
| devananda | jroll: are you comfortable marking cleaning as done once https://review.openstack.org/#/c/161453/ moves through the gate? | 16:22 |
| *** hj-hp has joined #openstack-ironic | 16:22 | |
| devananda | JayF: line up those ducks | 16:22 |
| JayF | quack | 16:22 |
| *** saripurigopi has joined #openstack-ironic | 16:23 | |
| devananda | jroll: the agent patch isn't +A'd yet, fwiw | 16:23 |
| JayF | I'm working on upstream for the three hours I'm in the office today | 16:23 |
| JayF | devananda: I'm about to +A it | 16:23 |
| devananda | but as that's not part of the integrated release, i'm fine | 16:23 |
| devananda | oh good | 16:23 |
| * rameshg87 wonders what ducks has to do with ipa | 16:23 | |
| JayF | JoshNang: agent patch for IPA coming in for a landing | 16:23 |
| *** Marga_ has quit IRC | 16:23 | |
| JoshNang | JayF: \o/ thanks! | 16:24 |
| JayF | rameshg87: English idiom -> "Get your ducks in a row" means to get ready for something | 16:24 |
| JayF | rameshg87: we've never done an official ironic-python-agent release, so I have to do $things to make that possible | 16:24 |
| rameshg87 | oh okay .. | 16:24 |
| NobodyCam | as /me is now lost, whats next up for reviews | 16:25 |
| BadCub | NobodyCam: I believe we have 154814 being refactored atm. | 16:27 |
| JayF | devananda: will we need a separate launchpad for ironic-python-agent, even if we don't use the bugtracker, for release purposes? | 16:27 |
| NobodyCam | BadCub: ack | 16:28 |
| ifarkas | dtantsur, sorry, what's the question? | 16:28 |
| dtantsur | ifarkas, if you stop adding config options on https://review.openstack.org/#/c/165860/ and file a bug for it, it still can land in K | 16:28 |
| NobodyCam | BadCub: https://review.openstack.org/#/c/165310/ +2'd (follow up for iLo capabilities) | 16:29 |
| devananda | JayF: TBD | 16:29 |
| ifarkas | dtantsur, should I replace them with constants for now? | 16:29 |
| JayF | OK, so for now I'll just make the pypi then | 16:29 |
| rameshg87 | so until k3 is cut out, only approved features should land ? and nothing else ? | 16:29 |
| wanyen | devananda, secure boot is not on the kilo3 LP any more. Can we apply FFE for it? | 16:29 |
| BadCub | NobodyCam: awesome! TY :) | 16:29 |
| dtantsur | ifarkas, yes | 16:29 |
| rameshg87 | i mean other bugs shouldn't land ? | 16:29 |
| * rloo wonders if we should be +1'ing the cleaning patches in nova-land | 16:29 | |
| devananda | rameshg87: until after RC1 is cut, we should not land any more features, string changes, config changes, etc -- only bug fixes | 16:29 |
| JayF | rloo: you have a link handy? | 16:30 |
| ifarkas | dtantsur, ok, I will update the patch | 16:30 |
| devananda | rameshg87: not counting the ones that are in the gate right now | 16:30 |
| dtantsur | ifarkas, then ping devananda to lift -2 (after K-3, no hurry right now) | 16:30 |
| rloo | JayF: https://review.openstack.org/#/c/164313/ | 16:30 |
| JoshNang | rloo: ++ | 16:30 |
| rloo | JayF: https://review.openstack.org/#/c/161474/ | 16:30 |
| rameshg87 | okay. | 16:30 |
| ifarkas | dtantsur, ahh, okay. I thought it requires immediate action | 16:31 |
| rameshg87 | devananda: so something like this shouldn't land ? - https://review.openstack.org/#/c/165788/3/ironic/drivers/modules/seamicro.py | 16:31 |
| devananda | reminder for every core -- https://wiki.openstack.org/wiki/FeatureFreeze << rloo yuriyz Shrews NobodyCam dtantsur lucasagomes Haomeng|2 rameshg87 jroll | 16:31 |
| NobodyCam | rloo: +1'd :) | 16:31 |
| rameshg87 | devananda: new string for translation is there in that | 16:31 |
| rameshg87 | okay | 16:31 |
| rameshg87 | reading through - first time :) | 16:31 |
| Shrews | devananda: thx for the reminder | 16:31 |
| devananda | rameshg87: correct. though we are allowed to make exceptions ... and that's one I might make | 16:32 |
| lucasagomes | thanks | 16:32 |
| rameshg87 | okay, got it | 16:32 |
| NobodyCam | ahh good link devananda TY | 16:32 |
| rloo | thx for the reminder devananda. There's also a string freeze too. | 16:32 |
| jroll | 16:22:32 devananda | jroll: are you comfortable marking cleaning as done once https://review.openstack.org/#/c/161453/ moves through the gate? <- yes | 16:32 |
| rloo | and a dependency freeze: https://wiki.openstack.org/wiki/DepFreeze | 16:33 |
| rloo | here's the string freezy: https://wiki.openstack.org/wiki/StringFreeze | 16:33 |
| devananda | if anyone wants to know what I do during this time -- https://wiki.openstack.org/wiki/PTL_Guide#Special_actions_between_Feature_freeze_and_RC1 | 16:34 |
| devananda | rloo: also good links | 16:34 |
| dtantsur | as I understand, adding new string is not as bad as changing existing one, right? | 16:34 |
| * lucasagomes will document stuff while on ff | 16:34 | |
| devananda | jroll: great. also, JayF just approved it anyway, so :) | 16:34 |
| jroll | cool | 16:35 |
| *** hj-hp has quit IRC | 16:35 | |
| NobodyCam | We should prob add the freeze link to our wiki, next week :-p | 16:35 |
| *** ifarkas has quit IRC | 16:35 | |
| NobodyCam | s/link/links. | 16:35 |
| rloo | NobodyCam, JoshNang: wrt https://review.openstack.org/#/c/164313/8/nova/virt/ironic/ironic_states.py. I am wondering if we should keep the comment 'Prior to the Kilo release...' | 16:37 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Changes for secure boot support for iLO drivers https://review.openstack.org/165907 | 16:37 |
| *** kkoski has quit IRC | 16:38 | |
| JoshNang | rloo: ah i didn't intend to remove that | 16:38 |
| rloo | JoshNang: it occurs to me that nova folks won't look at these patches this week anyway. So we have time... | 16:39 |
| JoshNang | rloo: that's probably true, i expect they're *slightly* busy | 16:40 |
| * JoshNang fixes | 16:40 | |
| NobodyCam | JoshNang: ++ TY | 16:40 |
| BadCub | stendulker_: you are refactoring 154814 and removing the dependency for pxe_ilo on it as well correct? | 16:40 |
| stendulker_ | BadCu : yes | 16:41 |
| stendulker_ | BadCub: yes | 16:41 |
| BadCub | awesome :-) TY | 16:41 |
| BadCub | I added 165907 to the etherpad too | 16:41 |
| stendulker_ | BadCub: Thank you :) | 16:42 |
| BadCub | stendulker_: YW :-) | 16:42 |
| BadCub | anyone want to take a look at: https://review.openstack.org/#/c/165310/ Follow-up Patch for iLO Discovery ? it's a tiny patch :-) | 16:44 |
| wanyen | ramesh, stendulker, can you tell me the status of secure boot? | 16:44 |
| stendulker_ | wanyen: Core reviewers felt that pxe_ilo changes of secure boot are too risky at this time given it involves change in bootloader | 16:46 |
| stendulker_ | wanyen: they have agreed for secure boot support for iscsi_ilo and agent_ilo drivers | 16:46 |
| stendulker_ | wanyen: I'm currently refactoring code to take out pxe_ilo support of secure boot | 16:47 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: [WIP] Add DevStack plugin for ironic-discoverd https://review.openstack.org/164782 | 16:47 |
| Shrews | BadCub: +A'd 165310 | 16:48 |
| BadCub | Shrews: awesme! TY :-) | 16:48 |
| wanyen | stendulker_, so secure_boot for agent_ilo and iscsi_ilo are targeted for Kilo but pxe-ilo seucre boot is not. | 16:48 |
| stendulker_ | wanyen: yes | 16:48 |
| wanyen | stendulker_: did they allow us to apply FFE for pxe-ilo secure boot? | 16:49 |
| rameshg87 | JayF: i hope we can push defects into ipa still | 16:49 |
| jroll | rameshg87: ipa doesn't FF :) | 16:49 |
| rameshg87 | JayF: wanted to push https://review.openstack.org/#/c/164931/4 if udevsettle doesn't work :) | 16:49 |
| rameshg87 | jroll: okay | 16:49 |
| jroll | what's the difference between pxe_ilo and iscsi_ilo? | 16:49 |
| jroll | virtual media? | 16:49 |
| JayF | rameshg87: IPA is not feature frozen ever | 16:49 |
| JayF | rameshg87: we treat it more like a client | 16:50 |
| BadCub | anyone free to look at https://review.openstack.org/#/c/165907/ Changes for secure boot support for ILO drivers? I think NobodyCam is giving it a +2 now | 16:50 |
| lucasagomes | rameshg87, btw, re local boot and IPA | 16:50 |
| rameshg87 | jroll: yeah, pxe vs virtual media | 16:50 |
| rameshg87 | jroll: but both use ilo for power & management | 16:50 |
| stendulker_ | wanyen: no, pxe_ilo is bumped to L | 16:50 |
| lucasagomes | did you find the problem with the partition table? | 16:50 |
| jroll | rameshg87: got it | 16:50 |
| rameshg87 | jroll: pxe_ilo better than pxe_ipmitool for ilo drivers because pxe_ilo works for all uefi boot mode operations (things like get/set boot device doesn't work with ipmi in uefi for hp servers) | 16:51 |
| lucasagomes | oh that's the patch | 16:51 |
| lucasagomes | rameshg87, will review | 16:51 |
| rameshg87 | lucasagomes: retrying help | 16:51 |
| rameshg87 | *helps | 16:51 |
| jroll | amyep, I understand :) | 16:51 |
| jroll | tab fail | 16:51 |
| jroll | rameshg87: yep, I understand | 16:51 |
| * jroll bbiaf | 16:51 | |
| rameshg87 | lucasagomes: i added retry logic and it works | 16:51 |
| rameshg87 | lucasagomes: http://paste.openstack.org/show/192820/ | 16:52 |
| lucasagomes | rameshg87, awesome, will take a look | 16:52 |
| rameshg87 | lucasagomes: log of how it works | 16:52 |
| rameshg87 | lucasagomes: NobodyCam suggested to try running "udevadm settle" solves the problem | 16:52 |
| rameshg87 | lucasagomes: yet to try that out | 16:52 |
| rameshg87 | JayF: thanks for info :) | 16:52 |
| NobodyCam | oh rameshg87 did that work? | 16:52 |
| NobodyCam | oh :-p | 16:53 |
| Shrews | rameshg87: thx for the explanation you gave me on the retry stuff | 16:53 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for iscsi_ilo driver https://review.openstack.org/154814 | 16:53 |
| rameshg87 | NobodyCam: yet to try with that .. built an ipa ramdisk with those changes, have to try it out when i am back tomorrow | 16:53 |
| rameshg87 | Shrews: yeah even i tried with attemps=3 once to see if it works :) | 16:53 |
| NobodyCam | :) | 16:53 |
| *** athomas has quit IRC | 16:54 | |
| wanyen | ramesh87 and stendulker_: why is bootloader change for pxe-ilo too risky? Grub2 are well established bootloader. | 16:54 |
| devananda | stendulker_: rameshg87: announce posted to list re: FFE for agent_ilo iscsi_ilo | 16:54 |
| devananda | everything else should already be going through the gate and being watched by ttx | 16:54 |
| devananda | BadCub: ^ | 16:55 |
| rameshg87 | wanyen: there were concerns on changing default boot loader for even pxe driver | 16:55 |
| rameshg87 | devananda: thanks | 16:55 |
| devananda | I'm going to step away for a little while -- ya'll know how to reach me if you need me for something | 16:55 |
| stendulker_ | devananda: thanks | 16:55 |
| BadCub | kk devananda | 16:55 |
| rameshg87 | wanyen: when we change for pxe_ilo, it changes for pxe drivers as well (uefi mode) | 16:55 |
| wanyen | ramesh87: grub2 is popular and well-established though | 16:55 |
| BadCub | stendulker_: I think we need to rebase https://review.openstack.org/#/c/154816/ | 16:55 |
| *** hj-hp has joined #openstack-ironic | 16:55 | |
| * NobodyCam takes the lull time to grab more coffee and a quick smoke | 16:56 | |
| rameshg87 | wanyen: yeah but we were changing elilo to grub which folks want to take closer look at | 16:56 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for agent_ilo driver https://review.openstack.org/154816 | 16:56 |
| rameshg87 | BadCub: i think it's his next target :) | 16:56 |
| rameshg87 | BadCub: ^^ | 16:56 |
| rameshg87 | yeah :) | 16:56 |
| BadCub | awesomeness :-) | 16:56 |
| wanyen | ramesh87: yes I got that. So why not give it FFE so that folks can take a closer look at? | 16:57 |
| wanyen | ramesh87: what kind of things that core reviewers want to look at? | 16:58 |
| *** romcheg has joined #openstack-ironic | 16:59 | |
| *** hyakuhei has joined #openstack-ironic | 16:59 | |
| *** romcheg1 has quit IRC | 16:59 | |
| lucasagomes | wanyen, http://eavesdrop.openstack.org/irclogs/%23openstack-ironic/%23openstack-ironic.2015-03-19.log | 17:00 |
| lucasagomes | wanyen, starts at 2015-03-19T14:35:42 | 17:00 |
| wanyen | lucasagomes: ty. I will take a look. | 17:00 |
| lucasagomes | np | 17:01 |
| rameshg87 | devananda: i think you got [1] wrong :) | 17:01 |
| lucasagomes | I left some comments in the patch too | 17:01 |
| rameshg87 | devananda: pxe_ilo patch link | 17:01 |
| rameshg87 | i mean in the mail | 17:01 |
| *** saripurigopi has quit IRC | 17:01 | |
| rloo | rameshg87: yeah. do you know the link? just reply with the correct link. | 17:02 |
| *** derekh has quit IRC | 17:02 | |
| *** ndipanov has quit IRC | 17:02 | |
| rameshg87 | rloo: yeah will do | 17:03 |
| lucasagomes | folks I will have to go | 17:03 |
| rloo | thx rameshg87 | 17:03 |
| *** hj-hp has quit IRC | 17:03 | |
| lucasagomes | came to the office and gotta get the train back | 17:03 |
| lucasagomes | have a good night everyone | 17:03 |
| rloo | g'night lucasagomes | 17:03 |
| BadCub | g'night lucasagomes | 17:03 |
| *** lucasagomes has quit IRC | 17:03 | |
| *** jmccrory has joined #openstack-ironic | 17:04 | |
| BadCub | anyone want to put a +A back on agent_ilo https://review.openstack.org/#/c/154816/ ? | 17:04 |
| *** saripurigopi has joined #openstack-ironic | 17:04 | |
| *** athomas has joined #openstack-ironic | 17:05 | |
| NobodyCam | night lucas | 17:05 |
| *** pcaruana has quit IRC | 17:05 | |
| rloo | BadCub: done | 17:05 |
| BadCub | rloo: awesome! TY :_) | 17:06 |
| BadCub | :-) | 17:06 |
| rloo | BadCub: that was my best review ever, cuz I didn't review it :D | 17:06 |
| BadCub | We also needs reviews on iscsi_ilo https://review.openstack.org/#/c/154814/ :-) | 17:06 |
| BadCub | rloo: LOL | 17:06 |
| *** jistr has quit IRC | 17:07 | |
| NobodyCam | wanyen: did you read the scroll back from this morning? | 17:08 |
| * BadCub goes to throw waffles in toaster | 17:08 | |
| *** harlowja_away is now known as harlowja_ | 17:09 | |
| *** hj-hp has joined #openstack-ironic | 17:09 | |
| openstackgerrit | Merged openstack/ironic: Add Cleaning Operations for iLO drivers https://review.openstack.org/157715 | 17:12 |
| NobodyCam | nice! | 17:13 |
| openstackgerrit | Merged openstack/ironic: Automate uefi boot iso creation for iscsi_ilo driver https://review.openstack.org/155900 | 17:13 |
| *** Marga_ has joined #openstack-ironic | 17:13 | |
| NobodyCam | nice! | 17:13 |
| rameshg87 | Nisha: ^^^ | 17:13 |
| NobodyCam | :) | 17:13 |
| *** saripurigopi has quit IRC | 17:13 | |
| * rameshg87 goes to sleep | 17:16 | |
| rameshg87 | good night folks | 17:16 |
| NobodyCam | night rameshg87 | 17:16 |
| rameshg87 | see you tomorrow | 17:16 |
| Nisha | rameshg87, :) | 17:16 |
| NobodyCam | thank you for everything | 17:16 |
| *** rameshg87 has quit IRC | 17:16 | |
| BadCub | g'night rameshg87 thank you for everything! | 17:16 |
| *** Marga_ has quit IRC | 17:17 | |
| NobodyCam | Side note: oh wow Percona got Wozniak to give a key note for their conference | 17:19 |
| BadCub | https://review.openstack.org/#/c/165907/ Changes for secure boot support for ILO drivers & iscsi_ilo https://review.openstack.org/#/c/154814/ Both need +A :-) | 17:20 |
| rloo | BadCub: are any of those needed before FF, or are they part of FF exception? | 17:24 |
| BadCub | rloo: they should be part of the FFE :-) | 17:24 |
| *** viktors is now known as viktors|afk | 17:25 | |
| rloo | BadCub: ok, so they can be done later by some non-HP person I guess. I don't have the energy to do non-urgent reviews. | 17:25 |
| BadCub | kk :) | 17:26 |
| Shrews | BadCub: +A'd the first one | 17:27 |
| BadCub | kk :) | 17:27 |
| *** hemna has quit IRC | 17:27 | |
| *** achanda has joined #openstack-ironic | 17:28 | |
| NobodyCam | I have a question on 154814 | 17:28 |
| *** achanda has quit IRC | 17:28 | |
| openstackgerrit | Ghe Rivero proposed openstack/python-ironicclient: Use oslo.i18n lib https://review.openstack.org/162797 | 17:28 |
| *** achanda has joined #openstack-ironic | 17:28 | |
| *** romcheg has quit IRC | 17:29 | |
| *** wuhg has quit IRC | 17:29 | |
| BadCub | brb | 17:29 |
| NobodyCam | we wrap a try around _update_secure_boot_mode at line 434 but then don't at line 698, do we need a try block there too? | 17:29 |
| dtantsur | calling it a day for now, g'night | 17:29 |
| *** dtantsur is now known as dtantsur|afk | 17:30 | |
| NobodyCam | night dtantsur :) | 17:30 |
| BadCub | g'night dtantsur|afk | 17:30 |
| rloo | night dtantsur|afk | 17:30 |
| NobodyCam | stendulker_: ^^^^^ | 17:30 |
| *** hemna has joined #openstack-ironic | 17:30 | |
| stendulker_ | NobodyCam: Hi | 17:30 |
| *** andreykurilin_ has joined #openstack-ironic | 17:31 | |
| stendulker_ | NobodyCam: letme check | 17:31 |
| *** coolsvap is now known as coolsvap|afk | 17:33 | |
| stendulker_ | NobodyCam: Not required. In tear_down() we wrap it so that if the user had accidentaly set the secure_boot=true and set the node for deploy | 17:34 |
| stendulker_ | then he could set the provision state to deleted safely without any exception. | 17:35 |
| stendulker_ | but we do not want to provide that in VendorPassthru during deploy | 17:36 |
| *** ijw has quit IRC | 17:36 | |
| stendulker_ | we would let deploy fail for the misconfiguration | 17:36 |
| *** ijw has joined #openstack-ironic | 17:36 | |
| *** andreykurilin__ has joined #openstack-ironic | 17:37 | |
| *** andreykurilin_ has quit IRC | 17:37 | |
| stendulker_ | NobodyCam: If we wrap around L698 then user will never know the deployment was in secure boot or a normal bailed out by the driver. | 17:37 |
| NobodyCam | stendulker_: ahh TY | 17:39 |
| NobodyCam | BadCub: 154814 now has two +2, needs +a from non-hp'er | 17:41 |
| BadCub | NobodyCam: awesome! TY!!! :-) | 17:41 |
| BadCub | when anyone has a chance: iscsi_ilo https://review.openstack.org/#/c/154814/ Needs Non-HP /+A & We can move this BP to the Approved list :-) | 17:42 |
| *** Marga_ has joined #openstack-ironic | 17:44 | |
| * jroll looks | 17:45 | |
| BadCub | TY jroll :-) | 17:45 |
| * JayF reviewed and +1'd the chain as well | 17:46 | |
| *** andreykurilin__ has quit IRC | 17:46 | |
| BadCub | ty JayF :-) | 17:47 |
| *** Marga__ has joined #openstack-ironic | 17:47 | |
| *** Marga_ has quit IRC | 17:48 | |
| jroll | NobodyCam: btw, I only see one +2 on there ;) | 17:48 |
| NobodyCam | man we got doh forgot to actually vote :-p | 17:49 |
| NobodyCam | hehehe | 17:49 |
| *** hyakuhei has quit IRC | 17:49 | |
| NobodyCam | corrected now | 17:49 |
| *** romcheg has joined #openstack-ironic | 17:51 | |
| *** kkoski has joined #openstack-ironic | 17:53 | |
| JayF | devananda: https://pypi.python.org/pypi/ironic-python-agent is a thing now, openstackci has owner access to it as documented in infra/manual/creators.html -- is there anything else I can do to help this along? | 17:54 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Ilo drivers sets capabilities:boot_mode in node https://review.openstack.org/155731 | 17:54 |
| openstackgerrit | Ghe Rivero proposed openstack/ironic: Sync with oslo.incubator https://review.openstack.org/165940 | 17:54 |
| *** romcheg1 has joined #openstack-ironic | 17:55 | |
| JoshNang | JayF: nice! | 17:55 |
| *** romcheg has quit IRC | 17:57 | |
| * BadCub sips coffee and stares at Zuul | 17:58 | |
| jroll | BadCub: dropped the +A hammer on that | 17:58 |
| BadCub | jroll: awesomeness! TY!!! :-) | 17:59 |
| BadCub | all of our feature BPs are in flight! | 17:59 |
| jroll | woot. | 18:00 |
| rloo | Wow. Great job everyone! (Why does this one seem like the worst deadline so far?) | 18:00 |
| BadCub | thank you everyone for being awesome! | 18:00 |
| BadCub | rloo: we will work on making future deadlines more happy :-) | 18:01 |
| rloo | thx BadCub :D | 18:01 |
| *** rsalevsky has left #openstack-ironic | 18:01 | |
| BadCub | :D | 18:02 |
| JoshNang | \o/ | 18:02 |
| stendulker_ | * Thank you all for allowing secure boot related patches and quickly completing the review for the same | 18:02 |
| wanyen | Hi Devananda, NobodyCam, Lucas: I looked through the irc log for secure boot discussion. I understand there are concerns about upgrade path for the conductr take over case. However, IMO, if we need to switch to Grub2 in order to support secure boot, it's better to swith earlier than later. Switching later will cause impacts to more users. | 18:02 |
| *** enikanorov has joined #openstack-ironic | 18:03 | |
| BadCub | stendulker_: thank you for getting the refactoring done so quickly! :) | 18:03 |
| NobodyCam | Awesome job everyone! | 18:04 |
| BadCub | brb | 18:05 |
| *** enikanorov__ has quit IRC | 18:06 | |
| wanyen | Wanyen: I was referring to pxe-ilo secure boot. | 18:06 |
| jroll | BadCub: so what's next? | 18:06 |
| wanyen | NobodyCam ^^^ | 18:08 |
| jlvillal | jroll: If you got 'em smoke 'em ;) | 18:08 |
| Shrews | jroll: profit | 18:08 |
| * jroll blows a cloud | 18:08 | |
| *** coolsvap|afk has quit IRC | 18:11 | |
| *** Marga__ has quit IRC | 18:11 | |
| *** Marga_ has joined #openstack-ironic | 18:12 | |
| *** Marga_ has quit IRC | 18:12 | |
| *** Marga_ has joined #openstack-ironic | 18:13 | |
| *** Marga_ has quit IRC | 18:14 | |
| *** Marga_ has joined #openstack-ironic | 18:15 | |
| * BadCub returns from smoke break hehehe | 18:16 | |
| stendulker_ | rloo: Had to rebase the https://review.openstack.org/#/c/155731/ "Ilo drivers sets capabilities:boot_mode in node" one more time due to merge conflicts | 18:16 |
| JayF | TheJulia: talked to CoreOS upstream re: your bug and put a comment on it ... basically the kernel option or removing the units from the image is the only workaround now. | 18:17 |
| JayF | TheJulia: Likely for upstream, I'll modify the CoreOS inject script to remove the guilty units | 18:17 |
| stendulker_ | rloo: please review the same when you get time | 18:17 |
| BadCub | jroll: next = take a smoke break! lol | 18:17 |
| jroll | BadCub: it's in my hand :) | 18:18 |
| NobodyCam | lol | 18:18 |
| BadCub | nice! lol | 18:18 |
| jroll | well, I guess I set it down to type | 18:18 |
| JayF | TheJulia: longer-term Alex (abcrawf in #coreos) is working on a thing that should remove this whole class of bugs | 18:18 |
| jroll | need a vape that doesn't require using my hands, hm | 18:18 |
| BadCub | I just got new juices last night. Gotta fire my vape back to life | 18:18 |
| *** jmccrory has quit IRC | 18:19 | |
| JayF | jroll: ^ also TheJulia's bug confirms that we've gotta rollback the partprobe->partx change :( | 18:19 |
| jroll | JayF: ughhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh | 18:19 |
| BadCub | :( | 18:19 |
| jroll | JayF: we can't roll it back, it won't pass the gate LOL | 18:19 |
| JayF | jroll: partprobe || true | 18:19 |
| jroll | I know | 18:19 |
| jroll | what about partprobe || partx || true | 18:20 |
| JayF | that's what I'm going to put up, and file a bug to extract all this shit out of bash into python | 18:20 |
| JayF | so we can use utils.execute(attempts=) | 18:20 |
| jroll | +1 | 18:20 |
| jroll | might be worth putting it in ironic-lib or whatever when we do that | 18:20 |
| NobodyCam | JayF: udev settle didn't help? | 18:20 |
| TheJulia | JayF: awesome | 18:20 |
| JayF | NobodyCam: this is somewhere else | 18:20 |
| NobodyCam | oh | 18:20 |
| JayF | NobodyCam: which is basically a really similar problem | 18:21 |
| JayF | NobodyCam: which is why step 2 is "refactor this into python and share the code" | 18:21 |
| NobodyCam | :) | 18:21 |
| devananda | wanyen: as i understand it, the pxe_ilo changes from elilo to grub2 are unrelated to the work to enable UEFI for the agent_ilo and iscsi_ilo drivers | 18:22 |
| NobodyCam | so does this mean we can take the rest of the off | 18:22 |
| NobodyCam | :-p | 18:23 |
| BadCub | LOL | 18:23 |
| NobodyCam | wb devananda :) | 18:23 |
| BadCub | babysit the gate and grumble at it | 18:23 |
| devananda | wanyen: if they are unrelated, and a benefit to users, then I'm OK landing those two things. on the other hand, if you're telling me that they all need to land together, I need to bump the whole feature. | 18:23 |
| BadCub | uh... | 18:24 |
| devananda | wanyen: though it looks like agent_ilo and iscsi_ilo are already moving through the gate, so it's too late | 18:24 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Call partprobe+partx before writing configdrive https://review.openstack.org/165954 | 18:25 |
| JayF | jroll: TheJulia ^ | 18:25 |
| jroll | +2 | 18:27 |
| TheJulia | JayF: Awesome, thank you! | 18:31 |
| wanyen | Devananda, ty for approving secure boot for iscsi-ilo and agent-ilo. For pxe-ilo what can we do to address the concerns? | 18:32 |
| devananda | wanyen: we'll come back to it after feature freeze | 18:36 |
| wanyen | devananda, ok. I will explore with ilo team to see if we can come up with solution to address the upgrade issue. I just think that switching to Grub2 is needed in order to support secure boot, and it's better to switch earlier than later. | 18:39 |
| jroll | wanyen: will agent_ilo secure boot work without switching to grub2? | 18:40 |
| wanyen | jroll, as far as I know grub2+shim is the only way to support secure boot. | 18:41 |
| jroll | wanyen: will agent_ilo secure boot work without switching to grub2? | 18:41 |
| JoshNang | JayF: my bash fu is pretty weak so IMBW, but if partx doesn't work, won't it fail there and not try partprobe? | 18:41 |
| jroll | JoshNang: partx exits successfully but doesn't actually do anything | 18:42 |
| JoshNang | ohh | 18:42 |
| JayF | ^ that | 18:42 |
| wanyen | jroll: I have to check with shiv. | 18:42 |
| JayF | partprobe *errors* in devstack | 18:42 |
| *** stendulker_ has quit IRC | 18:42 | |
| jroll | ok | 18:42 |
| *** Marga_ has quit IRC | 18:42 | |
| JoshNang | ah. gotcha! +2 | 18:42 |
| *** Marga_ has joined #openstack-ironic | 18:45 | |
| *** kkoski has quit IRC | 18:48 | |
| clif_h | what does node.delete in the ironic API do exactly? Does it unregister the node from the service? | 18:59 |
| clif_h | http://docs.openstack.org/developer/ironic/webapi/v1.html#delete--v1-nodes | 19:00 |
| JoshNang | clif_h: yup, deletes the db record | 19:01 |
| clif_h | so if I wanted to put a node in the deleting state I would use http://docs.openstack.org/developer/ironic/webapi/v1.html#put--v1-nodes-(node_ident)-states-provision to put it there right? | 19:02 |
| jroll | yes | 19:06 |
| *** Marga_ has quit IRC | 19:06 | |
| jroll | send target:deleted | 19:06 |
| jroll | clif_h: you're gonna be a star next cycle with this caching stuff :D | 19:07 |
| *** pelix has left #openstack-ironic | 19:07 | |
| clif_h | jroll: is that a good thing or a bad thing? :) | 19:07 |
| JayF | Ask JoshNang how he feels about it ;) | 19:08 |
| jroll | ^ | 19:08 |
| jroll | lol | 19:08 |
| * JayF & for the day | 19:08 | |
| NobodyCam | humm something up with one of my tests. | 19:09 |
| NobodyCam | should self.config(tempdir='/fake/path') set the tempdir value in a test? | 19:10 |
| *** Nisha has quit IRC | 19:11 | |
| Shrews | NobodyCam: i believe so | 19:13 |
| jroll | NobodyCam: might need the group? | 19:14 |
| NobodyCam | I tried group='DEFAULT' but it does n't like that | 19:14 |
| NobodyCam | at all | 19:14 |
| jroll | huh | 19:14 |
| jroll | idk, that method is in our tree I believe | 19:15 |
| NobodyCam | oslo_config.cfg.NoSuchGroupError: no such group: DEFAULT | 19:15 |
| NobodyCam | also oslo_config.cfg.NoSuchGroupError: no such group: default | 19:15 |
| NobodyCam | :-p | 19:15 |
| jroll | huh | 19:15 |
| *** pas-ha has quit IRC | 19:17 | |
| rloo | shouldn't need to specify the group, but that's interesting. | 19:17 |
| *** tiagogomes_ has quit IRC | 19:20 | |
| *** ijw has quit IRC | 19:20 | |
| NobodyCam | it was the way I passed the value that was wrong | 19:23 |
| NobodyCam | that is setting the conf value for the test | 19:23 |
| *** Marga_ has joined #openstack-ironic | 19:24 | |
| *** Marga_ has quit IRC | 19:25 | |
| *** kkoski has joined #openstack-ironic | 19:25 | |
| *** kkoski has quit IRC | 19:25 | |
| *** kkoski has joined #openstack-ironic | 19:26 | |
| openstackgerrit | Merged openstack/ironic: Implement cleaning/zapping for the agent driver https://review.openstack.org/161453 | 19:27 |
| jroll | JoshNang: ^ grats | 19:28 |
| JoshNang | \o/ | 19:28 |
| rloo | yay | 19:28 |
| rloo | JoshNang: don't forget all my nits ;) | 19:28 |
| *** kkoski has quit IRC | 19:29 | |
| *** hemna has quit IRC | 19:29 | |
| JoshNang | rloo: i definitely won't :) | 19:29 |
| JoshNang | hopefully have a patch up today | 19:29 |
| rloo | JoshNang: no hurry. take a break! | 19:30 |
| NobodyCam | JoshNang: be sure to look at my comment too | 19:31 |
| NobodyCam | will affect logging | 19:31 |
| *** kkoski has joined #openstack-ironic | 19:32 | |
| *** kkoski has quit IRC | 19:33 | |
| *** kkoski has joined #openstack-ironic | 19:33 | |
| JoshNang | will do! | 19:33 |
| * BadCub bangs head against the gate | 19:41 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Check temp dir is usable for ipmitool driver https://review.openstack.org/160383 | 19:43 |
| openstackgerrit | John Trowbridge proposed stackforge/ironic-discoverd: eDeploy: Store all of the facts collected by edeploy in Ironic DB https://review.openstack.org/165538 | 19:46 |
| NobodyCam | let me know what ya think the directory_to_check if / elif block.. I kinda feel the reviews have changed that around a bit. but I think it will still do what we want. | 19:46 |
| * NobodyCam steps afk fir a few | 19:47 | |
| *** dprince has quit IRC | 19:47 | |
| *** Marga_ has joined #openstack-ironic | 19:55 | |
| openstackgerrit | Ghe Rivero proposed openstack/ironic: Sync with oslo.incubator https://review.openstack.org/165940 | 19:58 |
| adam_g | is there a specific API microversion that configdrive was introduced? | 19:59 |
| *** jmccrory has joined #openstack-ironic | 20:03 | |
| openstackgerrit | Merged openstack/ironic: Support agent_ilo driver to perform cleaning https://review.openstack.org/164646 | 20:04 |
| JoshNang | \o/ | 20:04 |
| BadCub | sweet | 20:04 |
| rloo | JoshNang: sorry, I added a comment to https://review.openstack.org/#/c/164313/ | 20:04 |
| JoshNang | rloo: don't be sorry! totally agree | 20:06 |
| *** sambetts has quit IRC | 20:06 | |
| rloo | JoshNang: I can't figure out why I didn't see that before. I must be more tired of reviewing than I thought. | 20:06 |
| *** sambetts has joined #openstack-ironic | 20:06 | |
| rloo | should we hold off making any changes that aren't for k-3, so we don't exhaust jenkins? | 20:08 |
| *** Marga_ has quit IRC | 20:08 | |
| JoshNang | rloo: no worries, you're still killin it on reviews. | 20:08 |
| JoshNang | probably a good idea, i'll push that up tomorrow/monday | 20:08 |
| rloo | JoshNang: ha ha. | 20:08 |
| *** dprince has joined #openstack-ironic | 20:10 | |
| *** hemna has joined #openstack-ironic | 20:11 | |
| openstackgerrit | Merged openstack/ironic: follow up patch for ilo capabilities https://review.openstack.org/165310 | 20:13 |
| BadCub | sweet | 20:14 |
| *** kkoski has quit IRC | 20:14 | |
| *** jmccrory has quit IRC | 20:14 | |
| BadCub | afk for a few | 20:15 |
| *** BadCub is now known as BadCub_Lurking | 20:15 | |
| rloo | oh oh, merge conflicts | 20:15 |
| rloo | https://review.openstack.org/#/c/165907/ | 20:15 |
| rloo | https://review.openstack.org/#/c/154814/ | 20:16 |
| *** mrda-away is now known as mrda | 20:17 | |
| mrda | Morning Ironic | 20:17 |
| *** kkoski has joined #openstack-ironic | 20:18 | |
| rloo | morning mrda | 20:22 |
| NobodyCam | morning mrda | 20:22 |
| NobodyCam | lol /me looks looks for a pen... lol | 20:23 |
| * NobodyCam is sitting next to 7 + computers but no pen | 20:25 | |
| NobodyCam | oh I see one *steals* from BadCub_Lurking | 20:25 |
| jroll | adam_g: I think configdrive was before microversions... so 1.1 | 20:26 |
| jroll | not 100% sure | 20:26 |
| adam_g | jroll, thats what it looks like | 20:26 |
| adam_g | trying to see what the appropriate way for nova to test if its supported | 20:27 |
| adam_g | maybe just a CONF.ironic.os_ironic_api_version that defaults to latest? | 20:27 |
| adam_g | and err's the instance if less than 1.1 | 20:27 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Changes for secure boot support for iLO drivers https://review.openstack.org/165907 | 20:28 |
| jroll | adam_g: yeah, I guess so? | 20:29 |
| * BadCub_Lurking comes back to find evidence of pen thievery lol | 20:29 | |
| *** BadCub_Lurking is now known as BadCub | 20:29 | |
| * NobodyCam *ducks under desk* | 20:30 | |
| BadCub | lol | 20:30 |
| adam_g | or just keep it as is and release note that ironic needs upgrading before nova | 20:30 |
| jroll | adam_g: well. | 20:30 |
| jroll | release note that ironic needs upgrading before using configdrive | 20:30 |
| * jlvillal might get to meet adam_g tonight at the meetup :) | 20:31 | |
| adam_g | jroll, i think you can try to use config drive with juno it just wont do anything with it? whereas now you'll get an errored instance | 20:31 |
| jroll | true | 20:32 |
| jlvillal | NobodyCam: I left you some feedback on the directory_check if/else block. some ideas. | 20:32 |
| adam_g | jlvillal, you're in portland? i didn't realize the developer meetup is next week @ PSU, might skip tonight in favor of that. only enough free time | 20:32 |
| jlvillal | adam_g: I am. Well technically Hillsboro :) I'm going to try to go to both. | 20:32 |
| jlvillal | adam_g: Plus my boss wants me to go to the hackathon. So that will be work related. | 20:33 |
| jlvillal | adam_g: Hopefully I get a chance to meet you at one or the other. | 20:33 |
| adam_g | jlvillal, ive never been the thing happening tonight. it always seems very product oriented, whereas next week will just be the usuals sitting around drinking bee^Wcoffee | 20:33 |
| adam_g | *and hacking | 20:33 |
| jlvillal | adam_g: I'm still in the learning phase. So I'm hoping tonight will help me learn more. But the hackathon sounds more interactive and more fun | 20:34 |
| openstackgerrit | Ghe Rivero proposed openstack/ironic: Use oslo_log lib https://review.openstack.org/157602 | 20:36 |
| openstackgerrit | Ghe Rivero proposed openstack/ironic: Sync with oslo.incubator https://review.openstack.org/165940 | 20:36 |
| mrda | rloo, NobodyCam o/ | 20:37 |
| *** Marga_ has joined #openstack-ironic | 20:38 | |
| rloo | hi, can someone help out. I rebased https://review.openstack.org/#/c/165907 | 20:39 |
| rloo | but am having trouble rebasing https://review.openstack.org/#/c/154814/ | 20:40 |
| rloo | BadCub: ^^. | 20:40 |
| * BadCub grumbles at Jenkins | 20:40 | |
| NobodyCam | rloo: whats up? | 20:42 |
| rloo | NobodyCam: would you please rebase https://review.openstack.org/#/c/154814/ | 20:42 |
| NobodyCam | I can try | 20:42 |
| rloo | NobodyCam: thx. I can try again but I don't feel like it. At least I got the first one done. | 20:43 |
| * BadCub growls at Github for crashing constantly | 20:44 | |
| rloo | NobodyCam: oh geez. I didn't realize there were three patches to this. So this needs to be rebased too: https://review.openstack.org/#/c/165907/ | 20:46 |
| BadCub | rloo: I think you got 165907 to go through :-) | 20:47 |
| rloo | BadCub: No, I did 154816.oh hmm. i feel like i'm going around in circles. | 20:48 |
| rloo | BadCub: whatever I did, please check it ;) | 20:48 |
| BadCub | Will scan through them again now :-) | 20:48 |
| rloo | BadCub: ok, sorry. i had assumed that the patches in the etherpad were ordered in the order they are in, but they aren't. | 20:49 |
| BadCub | yip... only need 154814 & 154816 :-) | 20:49 |
| BadCub | rloo: no worries :-) y'all have been busting serious brain the past few days! | 20:50 |
| rloo | NobodyCam: ^^ those two, if you don't mind. | 20:50 |
| *** Marga_ has quit IRC | 20:52 | |
| *** hj-hp has quit IRC | 20:52 | |
| *** Marga_ has joined #openstack-ironic | 20:52 | |
| NobodyCam | i'm trying 154814 now but getting no changes ??? | 20:53 |
| NobodyCam | strange.. is that what you get rloo ? | 20:53 |
| *** Marga_ has quit IRC | 20:54 | |
| *** Marga_ has joined #openstack-ironic | 20:54 | |
| *** trown is now known as trown|outttypeww | 20:54 | |
| rloo | NobodyCam: yeah, that's what I got. I think. | 20:55 |
| BadCub | hmmm | 20:55 |
| rloo | NobodyCam: No changes between prior commit f1e6bce and new commit 78c9d4a | 20:55 |
| rloo | NobodyCam: I'm going to go outside to get some fresh air. Maybe that'll help... | 20:56 |
| *** absubram has quit IRC | 20:59 | |
| *** kozhukalov has quit IRC | 21:00 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Secure boot support for iscsi_ilo driver https://review.openstack.org/154814 | 21:00 |
| NobodyCam | rloo: ^^^ thats got it I think | 21:00 |
| *** andreykurilin_ has joined #openstack-ironic | 21:12 | |
| *** Marga_ has quit IRC | 21:14 | |
| *** Marga_ has joined #openstack-ironic | 21:15 | |
| *** andreykurilin_ has quit IRC | 21:20 | |
| rloo | thx NobodyCam. Looks good. Do you want to approve it? | 21:20 |
| *** andreykurilin_ has joined #openstack-ironic | 21:20 | |
| *** davideagnello has joined #openstack-ironic | 21:21 | |
| rloo | NobodyCam: looks like k-3 has been cut | 21:21 |
| NobodyCam | still on 154816 | 21:22 |
| rloo | NobodyCam: it isn't urgent to get those rebased then. since they missed the k-3 cutoff. | 21:24 |
| *** davideagnello has quit IRC | 21:24 | |
| *** davideagnello has joined #openstack-ironic | 21:25 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Secure boot support for agent_ilo driver https://review.openstack.org/154816 | 21:25 |
| NobodyCam | I think that it | 21:26 |
| NobodyCam | thats even | 21:26 |
| BadCub | we'll need to +A those two again :( | 21:26 |
| openstackgerrit | Merged openstack/ironic-python-agent: Call partprobe+partx before writing configdrive https://review.openstack.org/165954 | 21:27 |
| *** achanda has quit IRC | 21:28 | |
| rloo | BadCub: the +A again is easy | 21:29 |
| BadCub | rloo: yeah :-) | 21:29 |
| *** andreykurilin_ has quit IRC | 21:30 | |
| NobodyCam | si I just did them both | 21:30 |
| BadCub | too bad they got borked before k3 got cut :( | 21:30 |
| *** andreykurilin_ has joined #openstack-ironic | 21:30 | |
| rloo | BadCub: that's ok, they'll still get into k release, which is really what matters (I think). | 21:30 |
| * NobodyCam smokes | 21:30 | |
| rloo | thx NobodyCam. And HP thx you too :-) | 21:31 |
| BadCub | yeah, I think so too | 21:31 |
| * BadCub thnx everyone for burning so much brain power the past few days :) | 21:31 | |
| BadCub | brb | 21:31 |
| devananda | BadCub: how's things? | 21:32 |
| *** andreykurilin_ has quit IRC | 21:35 | |
| *** andreykurilin_ has joined #openstack-ironic | 21:36 | |
| *** ChuckC has quit IRC | 21:36 | |
| *** achanda has joined #openstack-ironic | 21:37 | |
| NobodyCam | hey devananda :) | 21:37 |
| BadCub | devananda: not too bad.. :) | 21:37 |
| devananda | BadCub: whiteboard up to date and stuff? | 21:38 |
| BadCub | Yup | 21:38 |
| devananda | i see a bunch not landed yet // still in the gate | 21:38 |
| NobodyCam | looks like a could of patches missed the cutoff | 21:38 |
| NobodyCam | *couple even | 21:38 |
| BadCub | yeah UEFI got missed and so did 161006 | 21:38 |
| devananda | yup | 21:38 |
| devananda | https://launchpad.net/ironic/+milestone/kilo-3 is finalized | 21:38 |
| devananda | also - HOLY S*** we did a lot of work | 21:38 |
| jroll | ^ | 21:39 |
| devananda | can't say it enough -- everyone here rocks | 21:39 |
| rloo | 161006 is the agent so we're good there. | 21:39 |
| BadCub | YIP!!! | 21:39 |
| *** kkoski has quit IRC | 21:39 | |
| jroll | 161006 is openstackclient | 21:39 |
| *** alexpilotti has joined #openstack-ironic | 21:39 | |
| rloo | jroll: openstackclient? | 21:39 |
| * BadCub thinks ironic-y folk are totally awesome | 21:39 | |
| jroll | https://review.openstack.org/#/c/161006/ | 21:40 |
| NobodyCam | devananda: can you dbl check my rebase on 154816? | 21:40 |
| NobodyCam | :/ | 21:40 |
| rloo | jroll: oh, i think he meant https://review.openstack.org/#/c/161066/ | 21:40 |
| jroll | ok, thanks | 21:40 |
| BadCub | ack.. Typo | 21:40 |
| BadCub | yes I did mean 161066 *grumbles* | 21:41 |
| devananda | so folks, I'd like to distribute the work of creating our Kilo release notes | 21:46 |
| devananda | as there's far more than in previous cycles (and even then I didn't do the greatest job recording it all) | 21:46 |
| NobodyCam | jlvillal: around? | 21:46 |
| jlvillal | NobodyCam: Here! | 21:46 |
| devananda | wanyen: when you have a chance (like next week) could you update https://wiki.openstack.org/wiki/Ironic/Drivers for all the ilo work that has gone in during Kilo? | 21:47 |
| NobodyCam | hey hey jlvillal I see your comemnts on 160383, If I do that I couldn't pass in a directory to check | 21:47 |
| devananda | jroll: ditto for the agent drivers (update https://wiki.openstack.org/wiki/Ironic/Drivers next week) | 21:47 |
| NobodyCam | need to be able to pass other then temp dir to that | 21:47 |
| jlvillal | NobodyCam: Looking. But I thought I was only doing it if directory_to_check was None | 21:48 |
| NobodyCam | oh in side the first if | 21:48 |
| NobodyCam | not removing it | 21:48 |
| NobodyCam | nm | 21:48 |
| *** mtanino has quit IRC | 21:48 | |
| NobodyCam | :P | 21:49 |
| devananda | BadCub: and could you (next week) start tracking down all the drivers which aren't listed on https://wiki.openstack.org/wiki/Ironic/Drivers and getting folks to update it? | 21:49 |
| jroll | devananda: I don't want to write release notes, sorry | 21:49 |
| jlvillal | NobodyCam: yeah that :) I gave three different options. And also okay with current way too. | 21:49 |
| jlvillal | NobodyCam: Whatever you want to do. I had seen you asking for opinions on IRC, so I gave you some :D | 21:49 |
| jroll | devananda: I kid, will do | 21:49 |
| devananda | jroll: ok, don't. but could you make sure the current state of the agent* drivers are accurately represented on the wiki? (again, not right now, but some time maybe next week) | 21:49 |
| jroll | yeah | 21:50 |
| NobodyCam | lol I hadn't even looked at your third option yet | 21:50 |
| devananda | jroll: heh. thanks :) | 21:50 |
| BadCub | devananda: sure can | 21:50 |
| wanyen | devananda, yes sure we will update iLo driver document to reflect Kilo features. | 21:51 |
| *** athomas has quit IRC | 21:52 | |
| devananda | wanyen: cheers :) | 21:52 |
| *** achanda has quit IRC | 21:53 | |
| devananda | BadCub: similarly, as we approach the final release of Kilo, we'll need to make sure http://stackalytics.com/report/driverlog is an accurate | 21:54 |
| *** andreykurilin_ has quit IRC | 21:55 | |
| *** andreykurilin_ has joined #openstack-ironic | 21:55 | |
| BadCub | will do! | 21:55 |
| jroll | devananda: heh, that's so wrong | 21:55 |
| devananda | jroll: i know | 21:55 |
| devananda | jroll: but people actually look at that | 21:56 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Check temp dir is usable for ipmitool driver https://review.openstack.org/160383 | 21:56 |
| jroll | devananda: I know | 21:56 |
| jlvillal | NobodyCam: Curious, why didn't you like having directory_to_check=CONF.tempdir in the function declaration? | 21:56 |
| *** ijw has joined #openstack-ironic | 21:56 | |
| *** absubram has joined #openstack-ironic | 21:57 | |
| NobodyCam | it wasn't working for testing. when I set the tempdir conf option in tests it wasn't passing it in right. I was still getting none | 21:57 |
| jroll | you can't put CONF options in method definitions | 21:57 |
| NobodyCam | so the test was failing | 21:57 |
| NobodyCam | jroll: ++ ikr TIL | 21:58 |
| jlvillal | NobodyCam: Oh. Good answer :) | 21:58 |
| jroll | because those could resolve differently at runtime and definition time | 21:58 |
| NobodyCam | and they in fact do! | 21:58 |
| NobodyCam | :-p | 21:58 |
| jroll | :P | 21:58 |
| * jlvillal learned something new... | 21:59 | |
| *** oomichi has joined #openstack-ironic | 21:59 | |
| *** achanda has joined #openstack-ironic | 22:01 | |
| *** jamielennox|away is now known as jamielennox | 22:09 | |
| devananda | jlvillal: NobodyCam you can not put an oslo CONF option as the default value in a function definition | 22:13 |
| devananda | as jroll said | 22:13 |
| * devananda reads scrollback before typing any more | 22:13 | |
| NobodyCam | devananda: yep TIL today | 22:13 |
| BadCub | devananda: do you want BPs that got bumped listed on the "stuff that isn't ready for merging yet" section of the pad? | 22:16 |
| devananda | BadCub: yah. those should get priority attention during L-1 cycle | 22:17 |
| BadCub | kk | 22:17 |
| NobodyCam | do we need to re-approve them for L (/me hopes they get automatic approval) | 22:18 |
| jroll | I think re-approving would be silly | 22:19 |
| NobodyCam | ++ | 22:19 |
| jroll | one +2/+A to move them to the L folder | 22:19 |
| mrda | I don't think we've done that previously... | 22:19 |
| NobodyCam | mrda: this is only our second cycle with specs | 22:19 |
| NobodyCam | :-p | 22:19 |
| jroll | mrda: done re-approval? | 22:19 |
| jroll | or skipper re-approval | 22:20 |
| jroll | skipped* | 22:20 |
| jroll | also, morning :) | 22:20 |
| mrda | reapproval, total. I think we've just moved them over, right? | 22:20 |
| mrda | jroll: morning right back at you! | 22:20 |
| NobodyCam | humm skipper approval... does that mean I have to hop on one foot while approving patches? | 22:21 |
| NobodyCam | lol | 22:21 |
| mrda | But if we need a sanity to make sure they are still appropriate, sure | 22:21 |
| devananda | one thing we'll have to do -- look at all our specs, and sort out the ones we did NOT finish this cycle | 22:21 |
| BadCub | we have 4 BPs tagged for L already | 22:21 |
| jroll | I don't think we moved them before | 22:21 |
| devananda | then move them to another dir | 22:21 |
| *** Marga_ has quit IRC | 22:21 | |
| devananda | we did not have any incompleted specs in Juno. we have a lot this cycle | 22:21 |
| * mrda is wary of the term skipper approval due to Australian political baggage | 22:21 | |
| *** Marga_ has joined #openstack-ironic | 22:21 | |
| NobodyCam | lol | 22:21 |
| * NobodyCam googles | 22:21 | |
| devananda | also I think we should do a re-review of all the moved specs at the start of L | 22:22 |
| jroll | hm, I thought we had some incomplete specs in juno, maybe not | 22:22 |
| devananda | it's possible some of them won't make sense any more | 22:22 |
| jroll | devananda: why | 22:22 |
| devananda | but probably, it'll be a quick "yup, still good, +A" | 22:22 |
| jroll | why did they make sense a month ago and not now? | 22:22 |
| NobodyCam | oh cricket | 22:22 |
| mrda | devananda: +1 | 22:22 |
| devananda | jroll: because we learn things | 22:22 |
| jroll | hm | 22:22 |
| *** dprince has quit IRC | 22:23 | |
| devananda | i expressly do not want spec authors to have to go through a re-proposal process | 22:23 |
| devananda | (though that is what many other projects do) | 22:23 |
| jroll | oh, I see | 22:23 |
| devananda | i want the spec review team to take, like, a couple days and review them | 22:23 |
| devananda | and say "yes" or "no" | 22:23 |
| jroll | just make sure we actually realize what we're moving | 22:23 |
| jroll | yeah | 22:23 |
| devananda | right | 22:23 |
| *** ChuckC has joined #openstack-ironic | 22:24 | |
| devananda | if the spec team says "no, that doesn't make sense now" or "well, almost, but we changed this part of things" then we need to work with the author to make sure that they understand that, are OK with it, etc | 22:24 |
| jroll | yep | 22:24 |
| * devananda thinks that should be written down somewhere besides IRC | 22:24 | |
| * jroll calls not it | 22:25 | |
| devananda | :) | 22:25 |
| jroll | put it on the meeting agenda under announcements? | 22:25 |
| NobodyCam | can we have feature based spec's that once approved cover all drivers.. to avoid the massive onslought of per driver specs? | 22:26 |
| jroll | yes please | 22:26 |
| jroll | well, we do that now | 22:26 |
| jroll | and then people pile driver-specific things on top | 22:26 |
| devananda | yes, we do that | 22:26 |
| devananda | but then ^ | 22:26 |
| jroll | because hardware is dumb | 22:27 |
| devananda | also | 22:27 |
| NobodyCam | we had a lot of add x to driver y specs this cycle | 22:27 |
| devananda | when a single vendor proposes a feature that they want to put in the common driver API | 22:27 |
| devananda | it affects all drivers | 22:27 |
| devananda | so I've required them to have a common implementation as part of that -- because sanity | 22:27 |
| jroll | NobodyCam: yeah, thing is, every driver implements stuff differently | 22:28 |
| devananda | and that multiple driver maintainers then go review that common spec and code | 22:28 |
| NobodyCam | I kinda got lost with the bp / spec tracking at the end... (thank gwad for BadCub keeping me inline) | 22:28 |
| devananda | because we need a common abstraction that drivers agree on | 22:28 |
| BadCub | huh? what'd I do now? | 22:28 |
| devananda | take RAID as an example -- the common spec ended up being /totally/ different from the initial single-vendor proposal(s) | 22:28 |
| * devananda might be exxagerating, but hopes hte point is not lost | 22:29 | |
| *** mtanino has joined #openstack-ironic | 22:29 | |
| jroll | devananda: somewhat off topic: when can we drop the bash ramdisk support? :) | 22:29 |
| jroll | devananda: (which will allow agent/pxe drivers to be way more common) | 22:29 |
| devananda | jroll: not today :) | 22:29 |
| devananda | maybe next week? | 22:30 |
| devananda | (kidding) | 22:30 |
| jroll | thinking about e.g. image caching -- having the bash ramdisk means we maybe can't do it in pxe driver | 22:30 |
| jroll | I mean, after kilo or after liberty | 22:30 |
| devananda | oh | 22:30 |
| * BadCub is now very happy to hear his car will be ready for pickup from dealer tomorrow morning after 6 weeks | 22:30 | |
| jroll | and I guess we could differentiate between "drop support" and "stop adding features" | 22:30 |
| devananda | jroll: indeed | 22:30 |
| devananda | jroll: I'm all for EOL'ing it and declaring a deprecation path | 22:30 |
| jroll | if we can stop adding features today then I'm happy (and don't really care much about actually dropping support( | 22:31 |
| NobodyCam | devananda: i got that point. but wee started to push for less and less implatiomation details in spec so are we over concerned with how they do a feature, as long as it dose what the approved spec says it should? | 22:31 |
| devananda | NobodyCam: huh? I want to see implementation details in the spec | 22:32 |
| jroll | devananda: it just occurred to me that most contributors are either deploy-driver-centric or hardware-driver-centric | 22:32 |
| jroll | NobodyCam: unfortunately those implementation details often matter architecturally | 22:32 |
| jroll | NobodyCam: the hardware driver that uses NFS comes to mind | 22:32 |
| NobodyCam | oh that I will give in to.. ya | 22:33 |
| devananda | jroll: indeed. also changes to the REST API that merely specify what the CLI will look like | 22:33 |
| jroll | devananda: ya. | 22:33 |
| BadCub | jsut a quick (and probably stupid) question. We currently have 9 approved BPs slated for L and/or Futire combined... Would it not be wise to consider those the "chunk" for L? | 22:33 |
| jroll | jeez | 22:33 |
| devananda | or something that doesn't describe (in detail) how a conductor- failover is going to be handled, or an upgrade from juno to kilo | 22:33 |
| jroll | BadCub: as in the main things we want to accomplish or ? | 22:33 |
| *** mjturek1 has quit IRC | 22:34 | |
| jroll | BadCub: might be good to know what those are :P | 22:34 |
| devananda | BadCub: nope. not wise. we're going to have a bunch more too | 22:34 |
| jroll | when does open season for specs start? | 22:34 |
| devananda | BadCub: we should probably untarget all of those. then look at two lists | 22:34 |
| BadCub | I can put up a pad listing the ones we have already approved | 22:34 |
| devananda | - BPs that were almost done in kilo, but bumped (these should be prioritized for l1) | 22:34 |
| jroll | BadCub: I really hope rackspace alone has 9 specs we complete | 22:34 |
| NobodyCam | sold! but maybe we can the submission process for such "per driver" feature implementations (thou I'm not sure how atm) | 22:35 |
| devananda | - specs that were approved in kilo, but not close to completion (these should be reviewed and possibly prioritized) | 22:35 |
| devananda | and then we have another set of refactoring things that I know we want | 22:35 |
| devananda | like splitting boot & deploy interfaces | 22:35 |
| jroll | this sounds suspiciously like an etherpad | 22:35 |
| devananda | ++ | 22:35 |
| NobodyCam | gah... we can IMPROVE the submission... | 22:35 |
| devananda | someone quick! copy/paste it somewhere | 22:35 |
| jroll | NobodyCam: you're always accidentally words | 22:35 |
| BadCub | I believe the 4 tagged for L are the ones that got bumped from K. | 22:35 |
| jroll | :) | 22:35 |
| NobodyCam | yes... I blame my keyboard | 22:36 |
| NobodyCam | :) | 22:36 |
| jroll | lol | 22:36 |
| jroll | https://etherpad.openstack.org/p/ironic-liberty-priorities | 22:38 |
| jroll | devananda: BadCub ^ | 22:38 |
| devananda | oh btw folks, i will miss the meeting on monday | 22:38 |
| devananda | going to be on a speaker panel at that time, sorry | 22:38 |
| NobodyCam | :) | 22:38 |
| NobodyCam | thats the AM meeting for us | 22:39 |
| NobodyCam | i'll be there:) | 22:39 |
| devananda | yah. i'll be in grenoble | 22:39 |
| NobodyCam | gah | 22:39 |
| jroll | cool! | 22:39 |
| NobodyCam | so night for you :( | 22:39 |
| NobodyCam | when do you fly out? | 22:39 |
| devananda | sunday | 22:40 |
| NobodyCam | have a safe trip! | 22:40 |
| mrda | Have fun! | 22:40 |
| jroll | have some good wine! | 22:41 |
| *** kkoski has joined #openstack-ironic | 22:41 | |
| *** rloo_ has joined #openstack-ironic | 22:44 | |
| *** rloo has quit IRC | 22:44 | |
| NobodyCam | so tomorrow bugs and documentation ! | 22:46 |
| devananda | continue tracking the remainder of kilo work on reviewday or on whiteboard? | 22:46 |
| jroll | BadCub: I added all the stuff we have downstream or are working on or want to work on | 22:47 |
| NobodyCam | at least at this point Id say there still enough for a pad of its own | 22:47 |
| BadCub | jroll: awesome! ty :) | 22:48 |
| BadCub | I added three otehrs we bumped from K as well | 22:49 |
| BadCub | * others even *frowns* | 22:49 |
| jroll | NobodyCam is rubbing off on you :P | 22:50 |
| NobodyCam | hehehehe | 22:50 |
| BadCub | jroll: lol | 22:50 |
| jroll | BadCub: pxe_ilo secure boot stuff should be there too no? | 22:50 |
| BadCub | yup, grabbing it now | 22:51 |
| jroll | cool | 22:51 |
| BadCub | Been keeping a spreadsheet of things dropped :) | 22:52 |
| *** kkoski has quit IRC | 22:53 | |
| jroll | look at you all fancy with your spreadsheets and bullet points and links | 22:54 |
| NobodyCam | vhd? | 22:54 |
| BadCub | hehehe | 22:54 |
| jroll | NobodyCam: http://en.wikipedia.org/wiki/VHD_%28file_format%29 | 22:54 |
| jroll | NobodyCam: to be fair, I'm not sure if it's good, or if openstack cares about it | 22:55 |
| *** kkoski has joined #openstack-ironic | 22:55 | |
| *** hemna has quit IRC | 22:56 | |
| NobodyCam | support as is in can be used for deployment.. dose qemu-img support it | 22:57 |
| jroll | mmm | 22:57 |
| jroll | I don't remember offhand | 22:57 |
| jroll | I think comstud had a fork of qemu tools that does it | 22:57 |
| jroll | and while looking for it I found this piece of awesome I forgot about https://github.com/comstud/xensa108 | 22:58 |
| jroll | NobodyCam: https://github.com/comstud/libvhd-builder | 22:59 |
| NobodyCam | I just fear the "oh can you support my virtual box image, oh and my paralles and ... and ... | 22:59 |
| jroll | lol | 22:59 |
| jroll | xen uses vhd | 22:59 |
| jroll | as does hyper-v I guess | 22:59 |
| jroll | so some providers already use vhd | 22:59 |
| jroll | http://docs.openstack.org/image-guide/content/ch_converting.html | 23:00 |
| *** vipul is now known as vipuls | 23:00 | |
| BadCub | probability that tomorrow I will be taking a journey to pick up my long missing auto :) | 23:00 |
| jroll | NobodyCam: I see your concerns, lol | 23:01 |
| NobodyCam | jroll: ya.. my quick check seems to indacate it was a M$ thing | 23:01 |
| jroll | however it's something we're going to do eventually | 23:01 |
| *** marios has quit IRC | 23:01 | |
| jroll | actually it's partially done :P https://review.openstack.org/#/c/129070/ | 23:01 |
| NobodyCam | I would like to have to ablity to support different input formats | 23:01 |
| *** ijw has quit IRC | 23:01 | |
| *** oomichi has quit IRC | 23:02 | |
| jroll | so VHD is basically a tarball with one or more layers | 23:02 |
| jroll | which does nice things for snapshots etc AIUI | 23:02 |
| *** andreykurilin_ has quit IRC | 23:05 | |
| *** kkoski has quit IRC | 23:05 | |
| devananda | gotta run - appointment for a haircut, then dinner plans. have a good night, eveyrone o/ | 23:08 |
| jroll | you too :) | 23:08 |
| NobodyCam | night devananda | 23:10 |
| BadCub | g'night devananda | 23:10 |
| * BadCub is going away for the evening as well. Thanks again to everyone for being awesome!!! :) | 23:10 | |
| *** BadCub is now known as BadCub_Hiding | 23:10 | |
| NobodyCam | yep been ats it sense 5 am so its night for /me too | 23:14 |
| jroll | night y'all | 23:15 |
| NobodyCam | :) | 23:16 |
| *** rloo_ has quit IRC | 23:21 | |
| *** rloo has joined #openstack-ironic | 23:22 | |
| *** hemna has joined #openstack-ironic | 23:24 | |
| *** chlong has joined #openstack-ironic | 23:27 | |
| *** ijw has joined #openstack-ironic | 23:28 | |
| *** ijw has quit IRC | 23:29 | |
| *** ijw has joined #openstack-ironic | 23:30 | |
| *** Marga_ has quit IRC | 23:36 | |
| *** Marga_ has joined #openstack-ironic | 23:36 | |
| clif_h | has anyone ever tried to change the default intervals for periodic tasks in the ironic conductor? | 23:37 |
| jroll | I have | 23:38 |
| clif_h | and how'd it go? | 23:38 |
| jroll | it worked | 23:38 |
| clif_h | hrm | 23:38 |
| jroll | it's in production | 23:38 |
| jroll | :P | 23:38 |
| jroll | why what's up | 23:38 |
| clif_h | it worked, as in you watching it interval at the one you specified and not the default value? | 23:39 |
| clif_h | watched* | 23:39 |
| jroll | yes | 23:39 |
| jroll | afaik | 23:39 |
| *** yuanying has joined #openstack-ironic | 23:39 | |
| jroll | we run the power state loop not at the default in prod | 23:40 |
| clif_h | well, I'm seeing behavior in arsenal that leads me to believe that the values the periodic tasks are getting are the default ones and not the ones I put in the configuration file I'm testing with | 23:40 |
| clif_h | and I'm using periodic_task in the same way as ironic ifaict | 23:41 |
| *** yuanying has quit IRC | 23:41 | |
| clif_h | so it must be that the configuration file is not being loaded before the module is loaded | 23:41 |
| *** yuanying has joined #openstack-ironic | 23:41 | |
| clif_h | in arsenal | 23:41 |
| clif_h | but in ironic I assume the configuration file is getting loaded before the conductor module | 23:42 |
| clif_h | if its working correctly there | 23:42 |
| jroll | oh, I see | 23:42 |
| jroll | clif_h: I think you may need to import the module containing the periodic task at startup | 23:42 |
| jroll | or something | 23:42 |
| *** ijw has quit IRC | 23:44 | |
| *** stendulker has joined #openstack-ironic | 23:44 | |
| *** derekh has joined #openstack-ironic | 23:50 | |
| *** stendulker has quit IRC | 23:52 | |
| *** derekh has quit IRC | 23:52 | |
| *** mtanino has quit IRC | 23:54 | |
| clif_h | yep, it has to do when the class definition is processed | 23:58 |
| clif_h | I delayed the import until after I knew the configuration file was loaded, and now its working properly | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!