*** Sukhdev has quit IRC | 00:30 | |
*** rloo has quit IRC | 00:31 | |
*** Sukhdev has joined #openstack-ironic | 00:31 | |
*** Goneri has joined #openstack-ironic | 00:32 | |
*** Sukhdev has quit IRC | 00:37 | |
*** Goneri has quit IRC | 00:39 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/313729 | 00:44 |
---|---|---|
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/314353 | 00:44 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/314354 | 00:44 |
*** mbound has joined #openstack-ironic | 00:47 | |
*** mbound has quit IRC | 00:52 | |
*** hoangcx has joined #openstack-ironic | 00:59 | |
*** wajdi has joined #openstack-ironic | 01:00 | |
*** baoli has quit IRC | 01:16 | |
*** baoli has joined #openstack-ironic | 01:16 | |
*** baoli has quit IRC | 01:22 | |
*** baoli has joined #openstack-ironic | 01:34 | |
*** sacharya has joined #openstack-ironic | 01:39 | |
*** sacharya has quit IRC | 01:44 | |
*** baoli has quit IRC | 01:48 | |
*** baoli has joined #openstack-ironic | 01:48 | |
*** sacharya has joined #openstack-ironic | 02:02 | |
*** baoli has quit IRC | 02:03 | |
*** spandhe has joined #openstack-ironic | 02:06 | |
*** ijw_ has quit IRC | 02:12 | |
*** spandhe_ has joined #openstack-ironic | 02:16 | |
*** ayoung has quit IRC | 02:17 | |
*** spandhe has quit IRC | 02:18 | |
*** spandhe_ is now known as spandhe | 02:18 | |
*** amotoki has quit IRC | 02:25 | |
*** spandhe_ has joined #openstack-ironic | 02:35 | |
*** spandhe has quit IRC | 02:35 | |
*** spandhe_ is now known as spandhe | 02:35 | |
*** rajinir has joined #openstack-ironic | 02:44 | |
*** yuanying has quit IRC | 02:51 | |
*** amotoki has joined #openstack-ironic | 02:52 | |
*** amotoki has quit IRC | 03:00 | |
*** stendulker has joined #openstack-ironic | 03:04 | |
*** Sukhdev has joined #openstack-ironic | 03:05 | |
*** amotoki has joined #openstack-ironic | 03:10 | |
*** moshele has joined #openstack-ironic | 03:26 | |
*** links has joined #openstack-ironic | 03:29 | |
*** amotoki has quit IRC | 03:31 | |
*** amotoki has joined #openstack-ironic | 03:39 | |
*** ayoung has joined #openstack-ironic | 03:42 | |
*** yuanying has joined #openstack-ironic | 03:48 | |
*** irf has joined #openstack-ironic | 03:53 | |
*** sirushti has quit IRC | 03:57 | |
*** sirushti has joined #openstack-ironic | 04:06 | |
*** appu103_ has joined #openstack-ironic | 04:08 | |
*** stendulker_ has joined #openstack-ironic | 04:19 | |
*** stendulker has quit IRC | 04:23 | |
*** stendulker has joined #openstack-ironic | 04:27 | |
*** stendulker_ has quit IRC | 04:31 | |
*** jiangfei has joined #openstack-ironic | 04:42 | |
*** rajinir has quit IRC | 04:46 | |
*** Nisha_away has joined #openstack-ironic | 04:52 | |
*** fragatina has quit IRC | 04:59 | |
*** Sukhdev has quit IRC | 05:01 | |
openstackgerrit | Shivanand Tendulker proposed openstack/proliantutils: Add RIS support for power operations https://review.openstack.org/289800 | 05:04 |
*** vmud213 has joined #openstack-ironic | 05:06 | |
*** fragatina has joined #openstack-ironic | 05:17 | |
*** fragatina has quit IRC | 05:22 | |
*** appu103_ is now known as aparnav | 05:29 | |
*** itamarl has joined #openstack-ironic | 05:31 | |
openstackgerrit | Zhenguo Niu proposed openstack/ironic-inspector: Place db and cmd files to seperate directories https://review.openstack.org/312862 | 05:32 |
*** fragatina has joined #openstack-ironic | 05:33 | |
*** fragatina has quit IRC | 05:33 | |
*** fragatina has joined #openstack-ironic | 05:34 | |
*** moshele has quit IRC | 05:39 | |
*** ChubYann has quit IRC | 05:40 | |
openstackgerrit | Aparna proposed openstack/ironic: Deployment vmedia operations to run when cleaning https://review.openstack.org/307225 | 05:43 |
*** d0ugal has joined #openstack-ironic | 05:52 | |
*** itamarl has quit IRC | 05:53 | |
*** vsaienko has quit IRC | 05:58 | |
*** spandhe has quit IRC | 06:04 | |
*** itamarl has joined #openstack-ironic | 06:06 | |
*** amotoki has quit IRC | 06:10 | |
*** Nisha_away has quit IRC | 06:11 | |
*** Nisha has joined #openstack-ironic | 06:25 | |
*** itamarl has quit IRC | 06:26 | |
*** rcernin has joined #openstack-ironic | 06:27 | |
*** vsaienko has joined #openstack-ironic | 06:36 | |
*** moshele has joined #openstack-ironic | 06:41 | |
*** moshele has quit IRC | 06:42 | |
*** ChrisAusten has quit IRC | 06:42 | |
*** moshele has joined #openstack-ironic | 06:42 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: TEST Checking Gates https://review.openstack.org/311109 | 06:43 |
*** itamarl has joined #openstack-ironic | 06:45 | |
*** sacharya has quit IRC | 06:53 | |
*** amotoki has joined #openstack-ironic | 06:59 | |
*** tesseract has joined #openstack-ironic | 07:03 | |
*** d0ugal has quit IRC | 07:23 | |
pas-ha | morning Ironicers | 07:25 |
*** lennyb has quit IRC | 07:25 | |
*** d0ugal has joined #openstack-ironic | 07:26 | |
*** d0ugal has quit IRC | 07:26 | |
*** d0ugal has joined #openstack-ironic | 07:26 | |
lazy_prince | Hi.. I am running tox -e py27 for ironic-python-agent but it keeps failing ironic_python_agent.tests.unit.test_hardware.TestGenericHardwareManager.test_get_os_install_device with error about pyudev.device._errors.DeviceNotFoundByFileError: [Errno 2] No such file or directory: '/dev/sdb' | 07:31 |
lazy_prince | is there anything that I might be missing..? | 07:32 |
*** lennyb has joined #openstack-ironic | 07:34 | |
*** irf has quit IRC | 07:40 | |
*** irf has joined #openstack-ironic | 07:40 | |
*** ifarkas has joined #openstack-ironic | 07:43 | |
*** sacharya has joined #openstack-ironic | 07:54 | |
*** sacharya has quit IRC | 07:58 | |
*** zzzeek has quit IRC | 08:00 | |
*** zzzeek has joined #openstack-ironic | 08:00 | |
*** irf has quit IRC | 08:02 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: TEST Checking Gates https://review.openstack.org/311109 | 08:11 |
*** ohamada has joined #openstack-ironic | 08:12 | |
*** irf has joined #openstack-ironic | 08:14 | |
*** athomas has joined #openstack-ironic | 08:14 | |
*** aparnav has quit IRC | 08:14 | |
*** appu103_ has joined #openstack-ironic | 08:14 | |
*** dtantsur|afk is now known as dtantsur | 08:16 | |
dtantsur | Morning Ironic | 08:16 |
dtantsur | lazy_prince, from git master? | 08:16 |
*** e0ne has joined #openstack-ironic | 08:18 | |
irf | morning Ironic ! | 08:22 |
*** lucas-dinner is now known as lucasagomes | 08:22 | |
lucasagomes | morning all | 08:22 |
*** Nisha_away has joined #openstack-ironic | 08:28 | |
*** Nisha has quit IRC | 08:28 | |
*** daemontool__ is now known as daemontool | 08:30 | |
irf | as continue to the issue, "No valid host was found", i am still facing the issue ... | 08:31 |
irf | theer are still issues nova is not communicating well with Ironic or vice versa ... | 08:31 |
irf | do we require nova.conf and nova-compute files both ? | 08:35 |
dtantsur | lucasagomes, mind looking at https://review.openstack.org/#/c/313511/ again please? now it must be less controversial (even though I like it less) | 08:36 |
lucasagomes | dtantsur, hi there, sure thing | 08:37 |
lucasagomes | 1 sec | 08:37 |
*** jistr has joined #openstack-ironic | 08:37 | |
dtantsur | no worries.. just want it landing reasonably soon, I have a guy blocked on it downstream | 08:37 |
lucasagomes | irf, you want to have 1 compute per type right? Like one for baremetal and another for virtual? | 08:37 |
lucasagomes | irf, so yes, you need another nova.conf (that's where the driver is specified, plus you need to tweak things like ram_ratio for the baremetal one) | 08:38 |
*** Nisha_away has quit IRC | 08:39 | |
*** Nisha_away has joined #openstack-ironic | 08:39 | |
irf | yes i have one compute per type ... | 08:40 |
irf | on the controller node , i have set the nova compute service for baremetal ... | 08:41 |
irf | let me paste the contents of both the file , nova.conf and nova-compute.conf | 08:42 |
*** amotoki has quit IRC | 08:42 | |
*** amotoki has joined #openstack-ironic | 08:43 | |
*** Nisha_away has quit IRC | 08:47 | |
irf | http://pastebin.com/DhEN0rn5 | 08:47 |
irf | please review ... | 08:47 |
irf | this nova compute service i am using for baremetal | 08:47 |
*** derekh has joined #openstack-ironic | 08:47 | |
stendulker | dtantsur, lucasagomes: Can you please have a look at config drive related review https://review.openstack.org/#/c/230924/ | 08:50 |
dtantsur | stendulker, it's pretty big, but I'll try to fit it into today's schedule | 08:52 |
dtantsur | thanks | 08:52 |
lucasagomes | stendulker, ++ on my list as well | 08:52 |
stendulker | dtantsur, lucasagomes: thank you | 08:52 |
odyssey4me | mrda thanks - it seems clear to me that the ironicclient needs some love then cc jroll | 08:53 |
dtantsur | +100 to ironicclient needs some love, even though I don't know the context :) | 08:53 |
lucasagomes | sambetts, re your comment https://review.openstack.org/#/c/314149/2 that's the bug, pecan/WSME gets confused when routing the request | 08:57 |
lucasagomes | i totally agree that v1/nodes/detail/detail should be fine, but it's not due the technology we use | 08:57 |
*** sacharya has joined #openstack-ironic | 09:00 | |
*** ppiela has quit IRC | 09:03 | |
*** Nisha has joined #openstack-ironic | 09:03 | |
*** ppiela has joined #openstack-ironic | 09:03 | |
sambetts | lucasagomes: woah?! really... | 09:04 |
lucasagomes | sambetts, yeah, check the bug out | 09:04 |
*** sacharya has quit IRC | 09:04 | |
lucasagomes | that's why I linked it to the bug cause there's info there that would be hard to summarize in a comment | 09:05 |
lucasagomes | sambetts, and yes, pecan/wsme kinda sucks :-) | 09:05 |
*** itamarl_ has joined #openstack-ironic | 09:06 | |
sambetts | lucasagomes: I didn't read the bug as such, I read the bug as the reason v1/nodes/detail/detail is invalid is because we've created the v1/nodes/detail endpoint and therefore underminded wsmes routing | 09:07 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc: Add unittests for the utils.py module https://review.openstack.org/314197 | 09:07 |
lucasagomes | sambetts, that's it really... Maybe I missed something in your comment on the patch | 09:08 |
* lucasagomes re-reads | 09:08 | |
*** itamarl has quit IRC | 09:08 | |
*** itamarl_ is now known as itamarl | 09:09 | |
lucasagomes | sambetts, right, I agree with the comment... but not sure what actions you want on that? To update the note I left inline about new sub-controllers? | 09:10 |
sambetts | lucasagomes: I think the comment needs to be that if we add any further actions to the nodes controller like detail then it needs to be done using query, I think new sub-controllers/resources should be fine to continue as normal unless they also expose a detail like extension then that should be a query WDYT? | 09:13 |
sambetts | e.g. If I add a new drivers resource then I should be fine to have /v1/nodes/<uuid or name>/drivers | 09:13 |
sambetts | then I need to do v1/nodes/<uuid or name>/drivers?detail=true if I want more details | 09:14 |
sambetts | and then I can have v1/nodes/<uuid or name/drivers/<uuid or name> | 09:14 |
sambetts | with no conflicts | 09:15 |
lucasagomes | right, the problem will come if someone name a node "drivers" in this case | 09:15 |
sambetts | there isn't because we don't define a v1/nodes/drivers endpoint yet | 09:15 |
sambetts | so there isn't an existing endpoint to confused wsme | 09:16 |
irf | please review | 09:16 |
lucasagomes | sambetts, right, lemme check that... cause validate is the same story right? We do have v1/nodes/<uuid>/validate | 09:16 |
lucasagomes | but not v1/nodes/validate | 09:16 |
sambetts | yeah, validate shouldn't get confused if you have a noded called validate | 09:17 |
sambetts | it'd just be v1/nodes/validate/validate | 09:17 |
lucasagomes | right on! yeah I will update that comment to reflect it | 09:20 |
lazy_prince | @dtantsur: from stable/liberty | 09:23 |
sambetts | lucasagomes: with this in mind I wonder if the current reserve words cover to many things? Do we have v1/nodes/ports? | 09:23 |
lucasagomes | sambetts, nop only nodes/<ident>/ports | 09:24 |
lucasagomes | yeah, it may be covering more than needed | 09:24 |
dtantsur | lazy_prince, hmm, maybe we didn't backport the fix... there was a breaking change in pyudev | 09:25 |
dtantsur | I'll check in a few | 09:25 |
lazy_prince | dtantsur: thanks for looking into it.. | 09:25 |
dtantsur | weird, https://review.openstack.org/#/c/254612/ looks in liberty... | 09:26 |
dtantsur | lazy_prince, does you checkout have this patch applied ^^^? | 09:27 |
sambetts | lucasagomes: looking at the ref, only detail and validate are the ones we need to watch out for | 09:27 |
*** milan has quit IRC | 09:28 | |
*** daemontool has quit IRC | 09:28 | |
lucasagomes | sambetts, well, there's a bunch of comments on that patch | 09:29 |
lucasagomes | I will have to verify each situations to see | 09:29 |
lucasagomes | apparently nodes/ports will get routed to the ports controller anyway | 09:30 |
lucasagomes | even tho we only suport nodes/<uuid>/ports | 09:30 |
*** chlong has quit IRC | 09:30 | |
sambetts | :/ sounds like we added a fall back route somewhere :/ | 09:30 |
lazy_prince | dtantsur: let me check | 09:31 |
lucasagomes | sambetts, http://paste.openstack.org/show/496542/ | 09:31 |
sambetts | hmmm ... | 09:32 |
sambetts | that sounds like a real bug in our API to me | 09:32 |
lucasagomes | it is | 09:33 |
lucasagomes | that's the thing, just the way things get routed is problematic | 09:33 |
sambetts | lucasagomes: I think if we don't support v1/nodes/ports we should fix that in the code itself | 09:34 |
*** mgould has joined #openstack-ironic | 09:34 | |
lucasagomes | sambetts, so that's the pecan/wsme thingy that we need to fix | 09:34 |
lucasagomes | but, clearly we are avoiding it :-/ | 09:35 |
irf | guys , please help ... i am tsruggling lot but no luck | 09:35 |
sambetts | lucasagomes: I'm looking at our controllers now, and I think its our fault | 09:36 |
mrda | odyssey4me: that's still my problem then :) | 09:37 |
sambetts | irf: are you passing both those files into your nova compute service? | 09:37 |
sambetts | because one says one compute_driver and one says a different compute_driver | 09:37 |
sambetts | and infact the ironic virt driver is commented out | 09:37 |
lucasagomes | sambetts, so idk, I need to check the api code and see if we can avoid that | 09:42 |
lucasagomes | perhaps using custom_actions instead of subcontrollers or things like that | 09:42 |
mgould | morning Ironic! | 09:43 |
sambetts | lucasagomes: ok so our problem is that the ports controller has been added as a sub-controller of Nodes, but the ports controller hasn't been written as a sub-controller and its a complete fudge... | 09:45 |
lucasagomes | sambetts, right, that's what we thought that was the wsme magic, making it easy, reuse the same controller | 09:46 |
lucasagomes | perhaps we should not do that :-/ | 09:47 |
sambetts | lucasagomes: so I think you can reuse the same controller, however I think there is magic that looks at the number of arguments a function can take | 09:47 |
*** Nisha has quit IRC | 09:47 | |
sambetts | lucasagomes: e.g. ports get_one current takes self, port_id and fields, but to be a sub resource it needs to take self, node_id, port_id, field=None | 09:48 |
sambetts | lucasagomes: looking at these controllers makes me shiver ... | 09:50 |
*** Nisha has joined #openstack-ironic | 09:50 | |
lucasagomes | sambetts, heh I know | 09:50 |
vdrok | morning ironic, lucasagomes dtantsur sambetts and mgould ! | 09:51 |
lucasagomes | vdrok, hi there, good morning | 09:51 |
sambetts | o/ mgould | 09:51 |
lucasagomes | sambetts, maybe we should put a patch up changing one bit of it, say ports | 09:51 |
lucasagomes | and if viable we can re-structure the api | 09:51 |
sambetts | yeah, I think that might be worth doing | 09:52 |
lucasagomes | sambetts, right, should we hold-on on the follow up aptch then you think? | 09:53 |
mgould | vdrok, lucasagomes, sambetts dtantsur morning! | 09:53 |
vdrok | another problem with controllers - ports are subcontroller for both nodes and portgroups, meaning that either node_id or port_id can get into first function argument | 09:53 |
vdrok | s/port_id/portgroup_id | 09:54 |
dtantsur | morning mgould, vdrok | 09:54 |
lucasagomes | vdrok, :-/ | 09:55 |
irf | ok.. so what is the correct value for compute_driver ? | 10:01 |
*** milan has joined #openstack-ironic | 10:01 | |
irf | compute_driver=nova.virt.ironic.IronicDriver | 10:02 |
irf | is this correct ? | 10:02 |
*** daemontool has joined #openstack-ironic | 10:08 | |
sambetts | yes | 10:08 |
irf | ok | 10:08 |
irf | now another doubt is , what is the difference between nova.conf and nova-compute.conf file ? are they both really required OR it is just simply i messed up the things ... | 10:09 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Extend IRONIC_RAMDISK_TYPE to support 'dib' https://review.openstack.org/302403 | 10:11 |
mgould | question about Nisha's patch above: if IRONIC_RAMDISK_TYPE is dib, is the ramdisk image build on demand? | 10:16 |
lucasagomes | someone wants to take a quick peek at https://review.openstack.org/#/c/314197/ ? (it's just unittests for vbmc) | 10:17 |
mgould | lucasagomes, looking now | 10:18 |
*** Nisha_away has joined #openstack-ironic | 10:18 | |
lucasagomes | mgould, thanks, it currently has no unittests what so ever, that's the first ones | 10:18 |
mgould | morning Nisha | 10:18 |
Nisha_away | mgould, morning | 10:19 |
lucasagomes | so there's some files being moved around etc | 10:19 |
vmud213 | morning all | 10:26 |
irf | i made the same entries in both the files ... | 10:26 |
irf | compute_driver=nova.virt.ironic.IronicDriver | 10:26 |
mgould | vmud213, morning | 10:26 |
vmud213 | mgould o/ | 10:27 |
Nisha | mgould, u can build on demand if u set IRONIC_BUILD_DEPLOY_RAMDISK as true | 10:27 |
Nisha | mgould, or if these variables are not set it will build by itself | 10:28 |
irf | compute_driver=nova.virt.ironic.IronicDriver | 10:28 |
mgould | Nisha, thanks | 10:28 |
Nisha | for me i have tested the second way for building thru devstack | 10:28 |
vmud213 | Hello all, Can you review https://review.openstack.org/#/c/303503/ | 10:29 |
Nisha | I dont remember using IRONIC_BUILD_DEPLOY_RAMDISK option till now in CI | 10:30 |
openstackgerrit | Merged openstack/python-ironicclient: Fix typo in docstring of assertTableHeaders method https://review.openstack.org/310017 | 10:34 |
mgould | lucasagomes, done | 10:35 |
lucasagomes | thanks a lot! | 10:36 |
*** hoangcx has quit IRC | 10:37 | |
irf | even after uncommenting compute_driver ,, error still remains same | 10:37 |
Nisha | has anyone built the ramdisk successfully using these options in DIB? | 10:37 |
Nisha | sudo -E disk-image-create ubuntu-minimal simple-init ironic-agent iso -p openssh-server -o ubuntu-minimal-ramdisk | 10:37 |
Nisha | i am trying this as part of comment on patch https://review.openstack.org/#/c/264579/ by devananda | 10:38 |
Nisha | for me it fails as "+ sudo sh -c 'http_proxy=http://web-proxy.ind.hp.com:8080 debootstrap --verbose --variant=minbase --include=python,sudo,linux-image-generic --components=main,restricted,universe --arch=amd64 trusty /tmp/image.3JCwN7HU/mnt http://archive.ubuntu.com/ubuntu ' | 10:38 |
Nisha | sh: 1: debootstrap: not found | 10:38 |
Nisha | Unmount /tmp/image.3JCwN7HU/mnt/tmp/ccache | 10:38 |
Nisha | " | 10:38 |
Nisha | lucasagomes, dtantsur mgould ^^^^ any idea on above | 10:39 |
dtantsur | Nisha, you're asking 3 Red Hat guys about how often we build images with ubuntu? ;) | 10:40 |
Nisha | dtantsur, :) i didnt know all are Redhat guys | 10:40 |
dtantsur | nothing strikes me about your command line | 10:40 |
dtantsur | you may want to ask on #tripleo | 10:40 |
dtantsur | FWIW I'd prefer we do the same as we recommend in the docs | 10:41 |
Nisha | dtantsur, i would add this on the patch and will let devananda come back as of now | 10:42 |
*** MattMan has quit IRC | 10:42 | |
*** MattMan has joined #openstack-ironic | 10:42 | |
Nisha | dtantsur, ++ to what u said | 10:42 |
openstackgerrit | Merged openstack/virtualbmc: Add unittests for the utils.py module https://review.openstack.org/314197 | 10:43 |
mgould | Nisha, I've never tried those options, sorry | 10:44 |
Nisha | mgould, thanks. | 10:44 |
mgould | Nisha: dtantsur lucasagomes ifarkas and I are all Red Hatters | 10:45 |
Nisha | mgould, i knew dtantsur lucasagomes are from redHat | 10:46 |
lucasagomes | Nisha, and milan :-) | 10:46 |
* mgould still has Ubuntu installed on most of his personal machines: I only started working at RH a few months ago :-) | 10:46 | |
Nisha | i came to know about you now | 10:46 |
mgould | ah, milan, I was wondering why mkovTABTABTAB wasn't working :-) | 10:46 |
lucasagomes | lol | 10:46 |
dtantsur | mkovTABTABTAB is a great nick fwiw | 10:46 |
mgould | :-) | 10:46 |
ifarkas | +100! | 10:47 |
mgould | Nisha, it looks like you have to install debootstrap on your host machine | 10:48 |
mgould | it's a cross-compiler | 10:48 |
mgould | [cross-builder? you know what I mean] | 10:48 |
Nisha | mgould, no ... | 10:48 |
mgould | Nisha, from elements/debootstrap/README.rst "Use of this element will also require the tool 'debootstrap' to be available on your system." | 10:49 |
mgould | https://wiki.debian.org/Debootstrap | 10:49 |
Nisha | mgould, ok i will try | 10:50 |
mgould | anyway, it's a tool for installing a Debian base system onto *another* filesystem | 10:50 |
mgould | which means it needs to exist on the machine building the image | 10:50 |
mgould | "cross-compiler" == compiler running on one architecture, building binaries for a different architecture, eg gcc running on x86 but outputting ARM binaries | 10:51 |
mgould | not a great analogy, sorry | 10:51 |
openstackgerrit | Aparna proposed openstack/ironic: Deployment vmedia operations to run when cleaning https://review.openstack.org/307225 | 10:51 |
*** Nisha_away has quit IRC | 10:54 | |
*** dtantsur is now known as dtantsur|bbl | 10:55 | |
*** Nisha has quit IRC | 10:55 | |
milan | morning guys! :) | 10:58 |
milan | mgould, yeah, I've just changed that ;) | 10:58 |
milan | lucasagomes, hi :) | 10:58 |
lucasagomes | morning | 10:58 |
sambetts | lucasagomes: I think I've got a solution :) | 10:59 |
lucasagomes | sambetts, cool | 10:59 |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Update devstack plugin with new auth options https://review.openstack.org/297312 | 10:59 |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Migrate to using keystoneauth Sessions https://review.openstack.org/236982 | 10:59 |
mgould | milan, https://www.youtube.com/watch?v=rERG3x6vsGY | 11:00 |
*** sacharya has joined #openstack-ironic | 11:01 | |
milan | mgould, lol :D | 11:01 |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Update devstack plugin with new auth options https://review.openstack.org/297312 | 11:02 |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Migrate to using keystoneauth Sessions https://review.openstack.org/236982 | 11:02 |
*** sacharya has quit IRC | 11:05 | |
*** stendulker has quit IRC | 11:06 | |
jroll | nisha needs debootstrap installed is all :| | 11:08 |
jroll | morning | 11:09 |
mgould | morning jroll | 11:09 |
* mgould does the "I was right" dance | 11:09 | |
jroll | oh, I didn't read far enough heh | 11:09 |
sambetts | morning jroll | 11:09 |
mgould | jroll, np, it's great to have my suspicions confirmed :-) | 11:09 |
* sambetts is solving all the API issues ;) | 11:10 | |
sambetts | (rm -rf ironic/api) ;) haha | 11:10 |
mgould | heh | 11:10 |
jroll | itym sed -i -e 's/pecan/flask/g' ironic/api/ | 11:17 |
sambetts | nahhh, /me likes pecan hehe, we just need to use it properly | 11:18 |
*** dims has quit IRC | 11:20 | |
*** dims has joined #openstack-ironic | 11:25 | |
*** mgould is now known as mgould|afk | 11:26 | |
openstackgerrit | Sam Betts proposed openstack/ironic: Prevent URL collisions with sub-controllers: nodes/ports https://review.openstack.org/314514 | 11:30 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc: Add unittests for the cmd/vbmc.py module https://review.openstack.org/314515 | 11:32 |
*** appu103_ has quit IRC | 11:32 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc: "vbmc list" to sort the result by domain name https://review.openstack.org/314095 | 11:34 |
TheJulia | o/ just wanted to say hi folks, trying to rest and hea | 11:34 |
lucasagomes | TheJulia, hello :-) | 11:34 |
TheJulia | rest and heal, brain hasn't had coffee in days | 11:34 |
sambetts | TheJulia: :( | 11:35 |
*** thrash|g0ne is now known as thrash | 11:35 | |
TheJulia | sambetts: smiley faces, I'm smiling, as much as I can at the moment. Anyway, I think I'm going to give up on the nurse who is snoring super loudly and go hunt down some juice... although she promised to take me to real coffee today | 11:37 |
sambetts | TheJulia: the nurse is snoring?! | 11:37 |
TheJulia | Yeah, my nurse apparently has had a few long shifts recently, I'm doing fairly good so I'm being nice to her :) | 11:38 |
sambetts | :) | 11:39 |
mat128 | morning everyone o/ | 11:50 |
*** baoli has joined #openstack-ironic | 11:50 | |
*** lucasagomes is now known as lucas-hungry | 11:52 | |
*** baoli_ has joined #openstack-ironic | 11:53 | |
vdrok | morning jroll TheJulia and mat128 | 11:53 |
vdrok | hey milan, what happened to mkovacik? | 11:53 |
milan | vdrok, morning :D | 11:54 |
milan | yeah, I thought I'd change that | 11:54 |
vdrok | :) | 11:54 |
milan | mgould|afk things milan is too sexy though ;) | 11:54 |
vdrok | "I'm too sexy for my nick" then :) | 11:55 |
milan | :D | 11:55 |
*** baoli has quit IRC | 11:56 | |
* milan was know as mkovTABTABTAB for a while too | 11:56 | |
openstackgerrit | Sam Betts proposed openstack/ironic: WIP: Prevent URL collisions with sub-controllers: nodes/ports https://review.openstack.org/314514 | 11:57 |
*** jtomasek_ has joined #openstack-ironic | 11:58 | |
irf | in the conductor logs i am getting the same error : | 12:02 |
irf | http://pastebin.com/xuU1GZU3 | 12:02 |
*** jcoufal has joined #openstack-ironic | 12:04 | |
irf | please suggest | 12:05 |
openstackgerrit | Imre Farkas proposed openstack/ironic: [WIP] DRAC RAID configuration https://review.openstack.org/176319 | 12:07 |
*** jtomasek_ has quit IRC | 12:08 | |
jroll | irf: you might try the openstack users mailing list (http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack) or https://ask.openstack.org/en/questions/ | 12:10 |
*** dprince has joined #openstack-ironic | 12:16 | |
*** kromanenko_ has joined #openstack-ironic | 12:17 | |
*** rbudden has joined #openstack-ironic | 12:18 | |
*** jjohnson2 has joined #openstack-ironic | 12:19 | |
milan | sambetts, hi, I've got a question about Inspector API mod patch, you've got a minute? https://review.openstack.org/#/c/308379/10..12/ironic_inspector/main.py | 12:20 |
sambetts | milan: sure, 1 minute | 12:20 |
*** jcoufal has quit IRC | 12:21 | |
*** jcoufal has joined #openstack-ironic | 12:22 | |
milan | sambetts, :) | 12:22 |
* milan wonders should we have a GET method when we allow POST with data here? | 12:23 | |
*** xavierr has joined #openstack-ironic | 12:23 | |
milan | the reapply part | 12:23 |
vdrok | jroll: seems that we need to use ClusteredComputeManager in grenade job to be able to run tests in 3 vms | 12:23 |
milan | * to GET the (raw) image data | 12:24 |
milan | sambetts, ^ | 12:24 |
vdrok | because otherwise hypervisor resources are not being updated in time | 12:24 |
vdrok | after instance deletion | 12:24 |
sambetts | milan: I think we discussed being able to fetch the raw data | 12:24 |
jroll | vdrok: are you sure it isn't because they're busy cleaning? | 12:24 |
vdrok | jroll: I changed that, it waits for cleaning to finish now | 12:25 |
milan | sambetts, that might be true but I've forgotten the outcome here, sry | 12:25 |
sambetts | milan: GET ..../data/unprocessed -> return raw ramdisk data | 12:25 |
vdrok | and I still see 0s in hypervisor-show | 12:25 |
sambetts | I think we said that it made sense | 12:25 |
sambetts | milan: ^ | 12:25 |
jroll | vdrok: maybe we sleep 30 seconds or something to wait for it | 12:25 |
milan | sambetts, yeah, I had the same impression | 12:26 |
jroll | vdrok: or fix the actual computemanager? CCM is bad :/ | 12:26 |
milan | sambetts, thx. I'll -1 it again | 12:26 |
vdrok | jroll: well, adding time.sleep(30) in all the tests using >1 node is worse :) | 12:26 |
jroll | vdrok: right, why don't we make the in-tree ComputeManager do the right thing? | 12:27 |
vdrok | jroll: yep, I think we should, IIRC you wanter to get rid of CCM completely? | 12:27 |
vdrok | in our tree | 12:28 |
jroll | vdrok: yes, very much so. it's a hack to make two computes look like one | 12:28 |
vdrok | jroll: I think for now I'll switch to CCM, to further do the smoke stuff, then will try to look into how to fix the compute manager | 12:30 |
jroll | vdrok: sure, I'd just rather not have that in the gate | 12:30 |
vdrok | jroll: yep, completely agree, that will be too much new stuff in one job - tinyipa, smoke and CCM :) | 12:31 |
jroll | heh | 12:31 |
xavierr | good morning ironic :) | 12:36 |
*** irf has quit IRC | 12:36 | |
*** dtantsur|bbl is now known as dtantsur | 12:37 | |
dtantsur | morning jroll, TheJulia, mat128 | 12:38 |
dtantsur | and xavierr | 12:38 |
vdrok | morning xavierr ! | 12:40 |
mat128 | morning dtantsur xavierr | 12:40 |
vsaienko | morning dtansur, jroll, vdrok and all! | 12:40 |
vdrok | vsaienko: o/ | 12:41 |
openstackgerrit | Sam Betts proposed openstack/ironic: Prevent URL collisions with sub-controllers: nodes/ports https://review.openstack.org/314514 | 12:42 |
*** wajdi has quit IRC | 12:42 | |
sambetts | lucas-hungry: try that out for size ^^ | 12:42 |
sambetts | lucas-hungry: "ironic node-create --driver fake --name ports" then "ironic node-show ports" works | 12:43 |
xavierr | morning dtantsur vdrok mat128 | 12:43 |
xavierr | :) | 12:43 |
mat128 | sambetts: that review allows for reserved names as node names? | 12:43 |
mat128 | haha clever | 12:44 |
sambetts | mat128: it allows "ports" as a node name, there are still a couple of others that need reworking and unfortunatly we can never allow "detail" as a node name :'( | 12:44 |
mat128 | I was even fine with the original limitation, so this is icing on the cake :) | 12:45 |
sambetts | mat128: but IMO the ones that were breaking like v1/nodes/ports/ports was a bug that needed squashing | 12:45 |
sambetts | if people are happy with that approach then I'll put together the rest of the patches | 12:46 |
*** lucas-hungry is now known as lucasagomes | 12:51 | |
*** mjturek1 has joined #openstack-ironic | 12:52 | |
lucasagomes | sambetts, that's really cool! | 12:52 |
sambetts | as I said pecan can do it we just have to use it right :-P | 12:53 |
*** mjturek1 has quit IRC | 12:56 | |
*** mbound has joined #openstack-ironic | 12:57 | |
*** amotoki has quit IRC | 13:01 | |
*** sacharya has joined #openstack-ironic | 13:02 | |
*** Goneri has joined #openstack-ironic | 13:02 | |
jroll | vdrok: thinking about it more, it doesn't sound insane to wait for hypervisor-stats to show >0 before proceeding | 13:02 |
*** keedya has joined #openstack-ironic | 13:02 | |
*** mbound has quit IRC | 13:02 | |
*** mgould|afk is now known as mgould | 13:03 | |
sambetts | mat128: which call broke when you tried my patch? I'll try to work out if its in the code or your machine | 13:04 |
mgould | morning TheJulia mat128 | 13:05 |
mat128 | test_nodes.TestPost.test_post_ports_subresource, test_nodes.TestPatch.test_patch_ports_subresource, test_nodes.TestListNodes.test_ports_subresource_noid, test_nodes.TestDelete.test_delete_ports_subresource | 13:05 |
mat128 | Not saying it broke any of the real api calls, just unit tests (but that probably means something broke) | 13:05 |
sambetts | oh right the tests, yeah I didn't fix any of those yet :-P | 13:06 |
mat128 | MismatchError: 403 != 400 | 13:06 |
mat128 | Ahh | 13:06 |
*** vmud213 has quit IRC | 13:06 | |
mat128 | ok good to hear, then your patch works for it's intended purpose, I changed "spam" for "ports" in a get test and it worked a1 :D | 13:06 |
sambetts | wanted consensus on the approach before I fell down the test rabbit hole | 13:06 |
*** sacharya has quit IRC | 13:06 | |
sambetts | awesome :D I've tried, ironic node-port-list ports, and it works, so it likely that it might have changed an abort code somewhere or something, because there were some conditionals that I was able to remove because of the changes I made | 13:08 |
*** mjturek1 has joined #openstack-ironic | 13:10 | |
*** mjturek1 has quit IRC | 13:10 | |
*** mjturek1 has joined #openstack-ironic | 13:11 | |
openstackgerrit | Merged openstack/ironic: Cleanup unused conf variables https://review.openstack.org/280493 | 13:15 |
*** afaranha has joined #openstack-ironic | 13:15 | |
openstackgerrit | Merged openstack/virtualbmc: Add unittests for the cmd/vbmc.py module https://review.openstack.org/314515 | 13:17 |
openstackgerrit | Merged openstack/virtualbmc: "vbmc list" to sort the result by domain name https://review.openstack.org/314095 | 13:17 |
*** daemontool_ has joined #openstack-ironic | 13:21 | |
*** links has quit IRC | 13:23 | |
openstackgerrit | Andre keedy proposed openstack/ironic-inspector: Replace node data via uuid/data/unprocessed API https://review.openstack.org/308379 | 13:23 |
*** cdearborn has joined #openstack-ironic | 13:23 | |
*** daemontool has quit IRC | 13:24 | |
*** mgould has quit IRC | 13:29 | |
*** hshiina has joined #openstack-ironic | 13:31 | |
*** mgould has joined #openstack-ironic | 13:31 | |
*** ChrisAusten has joined #openstack-ironic | 13:32 | |
*** ChrisAusten has quit IRC | 13:35 | |
*** ChrisAusten has joined #openstack-ironic | 13:35 | |
jroll | dtantsur: hey, a todo popped up on my list reminding me to send an email about dropping the bash ramdisk - would you mind doing that? | 13:38 |
*** rloo has joined #openstack-ironic | 13:43 | |
jroll | dtantsur: also, just saw the last kilo release go out :D | 13:44 |
dtantsur | jroll, \o/ | 13:44 |
dtantsur | yeah, will do, gimme a few minutes - I'm on a meeting | 13:45 |
jroll | dtantsur: awesome, thank you. I'm around if you want help with words or dates or whatever | 13:45 |
sambetts | trying to shrink tinyipa more :( 25mb of tinyipa is libicu... a unicode library requirement imported from gdisk ... and then like 49mb is python deps | 13:47 |
*** wajdi has joined #openstack-ironic | 13:50 | |
wajdi | good morning all | 13:51 |
*** thiagop has joined #openstack-ironic | 13:51 | |
thiagop | Good morning, Ironic | 13:51 |
dtantsur | morning wajdi, thiagop | 13:51 |
thiagop | dtantsur: o/ | 13:52 |
openstackgerrit | Andreas Jaeger proposed openstack/pyghmi: Remove unused pngmath Sphinx extension https://review.openstack.org/286217 | 13:56 |
*** ametts has joined #openstack-ironic | 13:58 | |
*** hoangcx_ has joined #openstack-ironic | 13:59 | |
NobodyCam | Morning ironic folks, says the man with out internets :-P | 14:01 |
sambetts | NobodyCam: Morning | 14:01 |
*** amotoki has joined #openstack-ironic | 14:01 | |
NobodyCam | Morning sambetts | 14:01 |
*** causten_ has joined #openstack-ironic | 14:01 | |
dtantsur | morning NobodyCam; internets are so-so today, no worries :D | 14:01 |
*** ChrisAusten has quit IRC | 14:02 | |
*** hoangcx_ is now known as hoangcx | 14:02 | |
NobodyCam | Lol my provider (time Warner) is down in the entire neighborhood | 14:02 |
*** rloo has quit IRC | 14:02 | |
NobodyCam | Morning dtantsur | 14:02 |
*** rloo has joined #openstack-ironic | 14:02 | |
*** causten__ has joined #openstack-ironic | 14:03 | |
NobodyCam | Kinda thinking it might be a pto day if things don't get fixed soon ish:-P | 14:03 |
jroll | NobodyCam: but what would you do with a day off and no internet? :P | 14:04 |
jroll | morning :) | 14:04 |
*** mtanino has joined #openstack-ironic | 14:04 | |
NobodyCam | Morning jroll probably go pull tvs out of the old rv and move them into the new one... lol | 14:06 |
NobodyCam | :-) | 14:06 |
*** causten_ has quit IRC | 14:06 | |
jroll | heh | 14:07 |
*** hoangcx has quit IRC | 14:08 | |
openstackgerrit | Sam Betts proposed openstack/ironic-python-agent: TinyIPA: Shave off some file size from tinyipa ramdisk https://review.openstack.org/313675 | 14:09 |
rloo | morning everyone. | 14:20 |
dtantsur | morning rloo | 14:20 |
*** rajinir has joined #openstack-ironic | 14:20 | |
dtantsur | jroll, sorry, I constantly get distracted, I'll get to the bash ramdisk email as soon as I'm finally free | 14:20 |
rloo | jroll: are we still using launchpad for the client? https://blueprints.launchpad.net/python-ironicclient | 14:20 |
jroll | dtantsur: not a huge rush, thanks | 14:21 |
jroll | rloo: not for blueprints, just bugs... | 14:21 |
*** jjohnson2 has quit IRC | 14:21 | |
rloo | jroll: guess those didn't get moved to the new rfe tags blah blah | 14:22 |
*** jjohnson2 has joined #openstack-ironic | 14:22 | |
jroll | rloo: maybe not | 14:22 |
rloo | jroll: I just updated our link to our feature tracker, from https://wiki.openstack.org/wiki/Ironic. that's when i noticed this. | 14:22 |
*** mjturek1 has quit IRC | 14:23 | |
rloo | jroll: if i get a chance, i'll update it. although if anyone else wants to volunteer... | 14:23 |
jroll | rloo: oh, fun, thank you | 14:23 |
jroll | vdrok might be interested | 14:23 |
jroll | do we really need 4 RFEs for functional testing? :) | 14:23 |
jroll | also py34 is the future, and works in the client, we should probably kill the py33 thing | 14:23 |
*** Nisha_away has joined #openstack-ironic | 14:24 | |
*** Nisha has joined #openstack-ironic | 14:24 | |
rloo | dtantsur: i'm guessing the feature tracker link for ironic-inspector is also incorrect at https://wiki.openstack.org/wiki/Ironic#ironic-inspector | 14:24 |
rloo | jroll: i didn't actually look at the proposed bps. | 14:24 |
jroll | rloo: I only looked at the names | 14:24 |
dtantsur | rloo, true, I forgot about this page | 14:25 |
rloo | jroll: you looked at more than I did :) | 14:26 |
*** mjturek1 has joined #openstack-ironic | 14:27 | |
*** keedya has quit IRC | 14:35 | |
*** keedya has joined #openstack-ironic | 14:36 | |
vdrok | morning rloo | 14:39 |
rloo | hi vdrok. Did you hear your name mentioned? :) | 14:39 |
mgould | morning rloo | 14:39 |
vdrok | yep :) | 14:40 |
rloo | hi mgould | 14:40 |
vdrok | yeah, we shouldn't use it for blueprints | 14:40 |
rloo | vdrok: if you would like to move those bps to our bug rfe, that would be great. | 14:40 |
vdrok | rloo: sure, will do | 14:40 |
rloo | vdrok: thx! | 14:40 |
*** spandhe has joined #openstack-ironic | 14:41 | |
mgould | sambetts, wow, that's a lot of Unicode library | 14:41 |
mgould | is there another fdisk implementation we could use instead? | 14:42 |
vdrok | jroll: btw, should we change the icon in https://launchpad.net/ironic to https://wiki.openstack.org/wiki/File:Bear_metal.svg ? :) | 14:43 |
jroll | vdrok: hehe, maybe, remind me later | 14:43 |
* jroll jumps into meetings | 14:43 | |
mgould | vdrok, definitely | 14:44 |
sambetts | mgould: yeah, really sucks, I've tried looking into how to get rid of it, but I cant find a slimer alternative to gdisk | 14:44 |
mgould | oh wow, gdisk is hosted on sourceforge | 14:44 |
* mgould wonders if we could compile it from source with a --no-unicode option or something | 14:45 | |
jlvillal | mgould, Is that a problem? | 14:45 |
vdrok | morning jlvillal ! | 14:45 |
jlvillal | vdrok, Good morning. Privyet | 14:45 |
mgould | jlvillal, not a problem, I'm just surprised anyone still uses sourceforge | 14:46 |
jlvillal | mgould, True. I think they are getting better since they were sold. | 14:46 |
* jlvillal kind of likes gitlab.com for free private repos, unlike github.com | 14:46 | |
sambetts | jroll, lucasagomes, mgould: I tried making a tripleOwl style sticker out of BearMetal wdyt? | 14:47 |
mgould | sambetts, excellent | 14:47 |
sambetts | http://imgur.com/wmIBNfJ | 14:47 |
jlvillal | sambetts, photos or it didn't happen? | 14:47 |
mgould | jlvillal, I haven't tried gitlab, but it looks good | 14:47 |
mgould | sambetts, :-) | 14:48 |
sambetts | still a concept that needs refining, and I also have no idea how/where you get them printed/cut :-p | 14:48 |
jroll | sambetts: ha, nice | 14:49 |
jroll | sambetts: https://twitter.com/jayofdoom/status/729676783017021440 | 14:49 |
jlvillal | sambetts, http://bfy.tw/1ZoZ | 14:50 |
vdrok | ooooo | 14:50 |
sambetts | haha awesome :D, I like the round shaped ones, because the cover the apple on my mac book | 14:50 |
sambetts | jlvillal: ;) thanks | 14:50 |
jlvillal | :D | 14:50 |
*** kcalman has joined #openstack-ironic | 14:50 | |
mariojv | hi - is there a bug # i should put for recheck for the gate failures that dtantsur fixed yesterday? | 14:52 |
JayF | mariojv: that's not a thing anymore | 14:52 |
JayF | mariojv: just do recheck | 14:52 |
mariojv | thanks | 14:52 |
*** pcaruana has joined #openstack-ironic | 14:53 | |
mgould | sambetts, I'll definitely order some of those if you get them printed | 14:54 |
sambetts | mgould: :D | 14:56 |
*** stendulker_ has joined #openstack-ironic | 14:58 | |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: DNM Remove openvswitch-datapath-dkms from debs https://review.openstack.org/314616 | 15:00 |
*** piet has joined #openstack-ironic | 15:01 | |
*** piet has quit IRC | 15:02 | |
*** piet has joined #openstack-ironic | 15:02 | |
*** milan has quit IRC | 15:04 | |
openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add tempest to the test-requirements.txt https://review.openstack.org/310179 | 15:04 |
sambetts | woah?! when running devstack just now, enroll nodes just tried to do 2 port creates with the same mac addresses... | 15:09 |
*** sacharya has joined #openstack-ironic | 15:09 | |
lucasagomes | sambetts, hah looks pretty good! | 15:10 |
lucasagomes | jroll, JayF mannn I really want one sticker like that! | 15:11 |
JayF | Why you taggin' jroll | 15:11 |
JayF | I am the keeper of the stickers! | 15:11 |
lucasagomes | :-) | 15:12 |
lucasagomes | the stickers looks pretty awesome | 15:13 |
JayF | For anyone who wants pixie boots stickers: PM me an address and a number; and try to coordinate if you work in the same place as other folks on project (aka I don't wanna send 4 envelopes to 4 people if you all work in the same office). Stickers look like -> https://twitter.com/jayofdoom/status/729676783017021440/photo/1 | 15:13 |
JayF | mariojv: ^ you mind being keepers of the stickers for Rackspace castle again? If so I'll ship you a couple dozen | 15:13 |
jlvillal | mgould, lucasagomes Wondering if any updates/news on Grenade stuff during your time zone? | 15:13 |
lucasagomes | jlvillal, not here :-( (was working on virtualbmc/pyghmi's) | 15:15 |
jlvillal | lucasagomes, Thanks | 15:15 |
sambetts | jlvillal: just in the middle of trying to run it | 15:16 |
jlvillal | sambetts, cool :) | 15:16 |
lucasagomes | JayF, btw, not sure if you've seem my ping yesterday. But tweaking the ipmi session timeout in pyghmi made the vbmcs way more stable | 15:16 |
*** causten__ has quit IRC | 15:16 | |
JayF | I saw that in scrollbacka nd have the review open | 15:16 |
JayF | but I haven't looked at it at all | 15:17 |
JayF | but more than happy to have broken it and seen it fixed; I presume you did some testing with that in the devstack I stood up | 15:17 |
lucasagomes | JayF, yup, I applied the fix locally because I still need a release from pyghmi to be able to ship it | 15:18 |
JayF | fine by me, but very glad you saw it | 15:20 |
JayF | just help with stability or did it speed it up too? | 15:20 |
lucasagomes | afaict stability, I'm still looking into it | 15:22 |
lucasagomes | but for now changing pyghmi's a bit painful since there's no unittests | 15:22 |
lucasagomes | we really need to do some work on that, cause things can break easily | 15:22 |
*** kbyrne has quit IRC | 15:22 | |
*** keedya has quit IRC | 15:23 | |
mgould | jlvillal, I got a bit further after the meeting (notes added to Etherpad), but no progress today | 15:24 |
mgould | jlvillal, how are you getting on? | 15:24 |
jlvillal | mgould, I am getting an error when creating the server | 15:26 |
*** wshao has joined #openstack-ironic | 15:26 | |
*** keedya has joined #openstack-ironic | 15:28 | |
mgould | jlvillal, me too | 15:28 |
*** sacharya_ has joined #openstack-ironic | 15:29 | |
rloo | jroll: wrt the poll on midcycle, is there a deadline? | 15:29 |
mariojv | JayF: might want to ask nate or mark, i will be moving in a few months and out of office end of this month | 15:29 |
jroll | rloo: thinking I'll just collect/announce monday mornign | 15:29 |
mgould | I noticed that the nova flavor specified cpu_arch as an extra_spec, but Ironic had it as a property rather than in properties/capabilities | 15:29 |
rloo | jroll: ok, i'll reply and indicate that. | 15:30 |
mgould | removing that extra_spec from the flavor let me get a bit further, but then I hit another error | 15:30 |
jroll | rloo: cool, thanks | 15:30 |
JayF | mariojv: oh that's right | 15:31 |
*** sacharya has quit IRC | 15:31 | |
mgould | sambetts, what version of gdisk are you using? The README says that libicu isn't required after v0.8.9 | 15:32 |
mgould | current version is 1.0.1 | 15:32 |
sambetts | mgould: just chrooted into by build dir and it says its version 1.0.0 | 15:33 |
mgould | sambetts, OK | 15:36 |
*** kbyrne has joined #openstack-ironic | 15:36 | |
sambetts | mgould: so maybe we can just nuke it but I don't know :/ | 15:36 |
mgould | sambetts, you shouldn't need libicu | 15:36 |
mgould | apparently it was only needed for UTF-16 partition name support | 15:37 |
mgould | and UTF-16 should be killed with fire anyway | 15:37 |
sambetts | hmm, I wonder if it'll be sad if I just rm the data lib :/ | 15:37 |
sambetts | libicudata.so.54 ... is the file | 15:38 |
mgould | the gdisk README has instructions for modifying the Makefile by hand to enable libicu, it's not even a ./configure option | 15:38 |
jlvillal | mgould, This is what I see http://ironicstack.com/grenade-logs/2016-05-10/grenade.sh.txt | 15:38 |
jlvillal | lucasagomes, sambetts ^^^ | 15:38 |
mgould | so unless your distro packager has gone out of their way to bloat the image, you shouldn't need it | 15:38 |
mgould | jlvillal, looking now | 15:38 |
* lucasagomes checks | 15:38 | |
jlvillal | I added --verbose --debug to the openstack server create command | 15:39 |
lucasagomes | jlvillal, that's what jroll's patch was fixing wasn't it?! | 15:39 |
jlvillal | I still need to debug | 15:39 |
*** milan has joined #openstack-ironic | 15:39 | |
lucasagomes | https://review.openstack.org/313600 | 15:39 |
mgould | jlvillal, yes, that's the same error as me | 15:39 |
jroll | lucasagomes: so actually mine wasn't intended to fix that, and likely doesn't | 15:39 |
jlvillal | lucasagomes, Maybe I screwed up and not applying it | 15:40 |
jlvillal | Looking | 15:40 |
*** kbyrne has quit IRC | 15:40 | |
* lucasagomes checks... looks very similar | 15:40 | |
jroll | lucasagomes: I got some other error running by hand which that fixes, but it seems invalid | 15:40 |
lucasagomes | oh | 15:40 |
*** itamarl has quit IRC | 15:40 | |
lucasagomes | so we should not apply that ? | 15:40 |
lucasagomes | I see your comment on the patch now | 15:40 |
jroll | yeah, probably not | 15:40 |
lucasagomes | cause after that I saw something like http://paste.openstack.org/show/496372/ | 15:40 |
jlvillal | lucasagomes, jroll's patch is after the create. | 15:41 |
dtantsur | jroll, mind taking one more look at https://review.openstack.org/313511 ? seems we've found a compromise... | 15:41 |
*** amotoki has quit IRC | 15:41 | |
lucasagomes | and worked around by creating a bunch of loopdevices | 15:41 |
lucasagomes | for ((i=0;i<200;i++)); do [ -e /dev/loop$i ] || sudo mknod -m 0600 /dev/loop$i b 7 $i; done | 15:41 |
* dtantsur gets nagged about this bug all the time... | 15:41 | |
jlvillal | lucasagomes, The create is on line 82. And the patch touches line 88 or so | 15:41 |
jlvillal | And I am applying the patch. | 15:41 |
lucasagomes | right, lemme run it again | 15:42 |
lucasagomes | I probably need a new env cause mine's all snapshotted and things may have changed | 15:42 |
jlvillal | mgould, Glad you got the same error as me :) | 15:43 |
* jlvillal needs to go to work after WFH meeting... | 15:43 | |
mgould | jlvillal, I'll keep plugging away at it for another hour and note anything I find on the etherpad | 15:43 |
jlvillal | mgould, Thank you! | 15:44 |
mgould | dtantsur, my suggestion of webcam + OpenCV was, on second thoughts, overcomplicated | 15:44 |
mgould | instead, we should use laser interferometry | 15:44 |
*** moshele has quit IRC | 15:44 | |
mgould | ops people won't mind mounting lasers on their racks, right? | 15:45 |
*** rama_y has joined #openstack-ironic | 15:45 | |
sambetts | mgould: https://review.openstack.org/#/c/314086/ my attempt to prevent the cause of inconsitant file sytems | 15:45 |
mgould | sambetts, is this an alternative to dtantsur's patch, or will it work with it? | 15:46 |
jroll | dtantsur: +2 but I'd prefer one more vote on that, I think | 15:47 |
sambetts | mgould: not sure if its an alternative really because I can't reproduce the scenario where the file sytem carrier file is unreliable, but it should make dhcp-all-interfaces finish before IPA starts hopfully preventing races | 15:48 |
jroll | dtantsur: one that hasn't been involved in this patch as much :P | 15:48 |
JayF | I'm looking | 15:50 |
JayF | jroll: dtantsur: Maybe dtantsur fix the nits real quick and then jroll and I can land it/ | 15:51 |
JayF | We can also land as-is and do a followup, just with people here and watching might be easier to just fix the patch | 15:51 |
jroll | either way | 15:51 |
*** tesseract has quit IRC | 15:52 | |
dtantsur | lemme fix it, sure thing | 15:54 |
mgould | dtantsur, I left a few nitpicks in case you had to do another patch | 15:55 |
*** Sukhdev has joined #openstack-ironic | 15:55 | |
dtantsur | I will right now | 15:55 |
mgould | dtantsur, cool | 15:56 |
mgould | dtantsur, was it your machine that had the unreliable carrier files in sysfs? | 15:56 |
* mgould is wondering if it's buggy hardware or kernel drivers | 15:56 | |
mgould | sambetts, cool, so it would make sense to land both? | 15:57 |
JayF | A buggy NIC driver? Never! /s | 15:57 |
mgould | I know, right? million-to-one chance | 15:57 |
sambetts | mgould: it needs testing it was just something I put together real quick, I'm not even sure it does what I hope it does xD my first time doing major systemd stuff | 15:58 |
dtantsur | mgould, no, not mine | 15:58 |
*** dprince has quit IRC | 15:58 | |
NobodyCam | oh looks like we may be getting some internets | 15:59 |
mat128 | sambetts: looks like icu is pulled by this https://github.com/ooyala/tinycore_gdisk/blob/master/gdisk.tcz.dep | 16:02 |
sambetts | mat128: yup | 16:03 |
JayF | sambetts: unless we had a fix that worked in all the images, that's not really a fix is it? | 16:03 |
mat128 | not sure if this is source for official package (not a TCL expert), but if it does it's getting pulled all the time | 16:03 |
JayF | sambetts: re: 314086 | 16:03 |
*** wshao has quit IRC | 16:03 | |
jroll | oh man, I thought we were about to use the TCL language | 16:03 |
mat128 | jroll: lol, not proper case for TinyCoreLinux | 16:04 |
*** keedya has quit IRC | 16:04 | |
dtantsur | sambetts, we should ignore imagebuild directory in the pep8 check, otherwise imagebuild/tinyipa/tinyipabuild gives hundreds of errors :D | 16:04 |
jroll | ah right, Tcl | 16:04 |
sambetts | dtantsur: haha | 16:04 |
jroll | dtantsur: well, we should ignore imagebuild/tinyipa | 16:04 |
jroll | the coreos builder has python :P | 16:04 |
*** penick has joined #openstack-ironic | 16:04 | |
JayF | sambetts: dtantsur: What file are you talking about? I don't see any python in that dir | 16:05 |
dtantsur | imagebuild/tinyipa/tinyipafinal/usr/local/bin/... | 16:05 |
dtantsur | ditto imagebuild/tinyipa/tinyipabuild/usr/local/lib/ | 16:06 |
JayF | sambetts: dtantsur: OH, that's like, an artifact of the build process? If so only +1 to excluding the build dirs, not to excluding the code places | 16:06 |
dtantsur | yeah, yeah, that's what I meant | 16:06 |
JayF | cool perfectg | 16:06 |
jroll | oh heh | 16:06 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [inspection] wait for the PXE DHCP by default and remove the carrier check https://review.openstack.org/313511 | 16:06 |
dtantsur | jroll, JayF ^^^ | 16:06 |
JayF | should the tinyipa readme be updated to not say it's experimental | 16:06 |
dtantsur | mgould, ^^ | 16:06 |
*** e0ne has quit IRC | 16:06 | |
JayF | given we're about to make it the gate and the devstack default? | 16:06 |
jroll | JayF: yep | 16:06 |
devananda | mornin, all | 16:07 |
wajdi | morning devananda | 16:07 |
*** appu103_ has joined #openstack-ironic | 16:07 | |
*** appu103_ is now known as aparnav | 16:08 | |
jroll | dtantsur: dare I ask for unit tests for the mac normalization thing? :) | 16:08 |
dtantsur | sigh... | 16:08 |
dtantsur | well, you definitely might :) | 16:08 |
dtantsur | morning devananda | 16:08 |
JayF | dtantsur: https://review.openstack.org/#/c/313035/ should not change inspector jobs due to https://review.openstack.org/#/c/313799/ | 16:09 |
sambetts | JayF: what questions about 314068 do you gave? | 16:09 |
sambetts | have? | 16:09 |
JayF | sambetts: I have more context now; the breakage only being in DIB image was something I was missing | 16:09 |
sambetts | JayF: ah :-P | 16:09 |
JayF | sambetts: just on patrol for having logic in only one imaeg when it's needed in three | 16:09 |
dtantsur | JayF, inspector jobs don't use this template | 16:09 |
JayF | oooh | 16:09 |
sambetts | JayF: ah makes sense | 16:09 |
JayF | dtantsur: do you want me to push project-config change to force it to ipxe off? | 16:10 |
JayF | dtantsur: er, tinyipa off | 16:10 |
JayF | dtantsur: or whatever, just prevent breakage, lol | 16:10 |
mgould | dtantsur, thanks | 16:11 |
* mgould raises his estimate of "buggy driver" | 16:12 | |
dtantsur | JayF, nevermind I guess. we should default to iPXE anyway, and looks like tinyipa works for us (judging by your patch) | 16:12 |
mgould | buggy drivers for RH-issued developer hardware would be embarrassing :-) | 16:12 |
*** sacharya_ has quit IRC | 16:12 | |
mgould | mat128, sambetts I see that .tcz file is three years old, it might well date from when gdisk did need libicu | 16:13 |
*** daemontool_ has quit IRC | 16:13 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: [inspection] wait for the PXE DHCP by default and remove the carrier check https://review.openstack.org/313511 | 16:13 |
mat128 | yeah, it also says it builds 0.8.7 which required libicu | 16:13 |
dtantsur | jroll, JayF now with moar tests ^^^ | 16:13 |
jroll | sambetts: how was it we dealt with the tempest choosing the wrong flavor thing again? | 16:13 |
jroll | dtantsur: <3 | 16:13 |
mat128 | mgould: do you know where to find the official sources for TinyCore builds? | 16:13 |
mgould | mat128, sambetts says he's getting v1.0.1 IIRC, so we're probably not using that version | 16:14 |
jroll | dtantsur: +2 | 16:14 |
mgould | mat128, no idea | 16:14 |
*** dprince has joined #openstack-ironic | 16:14 | |
sambetts | mgould, mat128: http://distro.ibiblio.org/tinycorelinux/6.x/x86_64/tcz/ | 16:15 |
jroll | sambetts: never mind, found it | 16:15 |
mat128 | sambetts: how does one go and submits a change for these things? | 16:16 |
mgould | mat128, sambetts http://tinycorelinux.net/6.x/x86_64/tcz/src/gdisk/compile_gdisk | 16:16 |
mat128 | .info says it builds 1.0.0, which corresponds to what we found | 16:17 |
mgould | so they are deliberately pulling in ICU | 16:17 |
* mgould shakes his fist at TCL | 16:17 | |
mat128 | http://distro.ibiblio.org/tinycorelinux/6.x/x86_64/tcz/gdisk.tcz.dep mentions icu | 16:17 |
mgould | sambetts, how hard would it be to build it from source? | 16:17 |
sambetts | mgould: are they actually using it though or are they jut including it in their dep for lolz | 16:17 |
jroll | dtantsur: https://review.openstack.org/#/c/314662/ | 16:18 |
mat128 | sambetts: I bet just for fun, since 0.8.9 mentions getting rid of it | 16:18 |
mat128 | as a requirement | 16:18 |
mgould | sambetts, not sure: what does the compile_gdisk file I linked to mean? | 16:18 |
openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Test post don't upvote https://review.openstack.org/311865 | 16:19 |
mgould | yep, they're deliberately patching the Makefile to drag in ICU: http://tinycorelinux.net/6.x/x86_64/tcz/src/gdisk/gptfdisk-1.0.0-convenience-1.patch | 16:19 |
sambetts | mgould: -_- | 16:19 |
sambetts | bulding from source wouldn't be hard | 16:20 |
mgould | and they statically link it in | 16:20 |
sambetts | man ... | 16:21 |
* sambetts is going to add it to the list of thing he builds from source | 16:23 | |
mat128 | sambetts, mgould: asked on #tinycorelinux, not expecting a reply but maybe | 16:24 |
mgould | mat128, thanks! | 16:24 |
*** ifarkas has quit IRC | 16:28 | |
*** Sukhdev has quit IRC | 16:29 | |
*** Sukhdev has joined #openstack-ironic | 16:30 | |
rloo | by the way, did we decide on whether to merge or not the portgroups API? https://review.openstack.org/#/c/206244/100 | 16:32 |
*** ijw has joined #openstack-ironic | 16:32 | |
rloo | it is up to 100 revisions. cool. warm. good. bad. sigh. | 16:32 |
dtantsur | folks want to wait with it... | 16:32 |
rloo | dtantsur: ok, so leave it as is. I won't review it again until we think we're close to merging the rest. | 16:33 |
dtantsur | ++ me neither | 16:33 |
rloo | what about network drivers. i heard some mention of splitting it up: https://review.openstack.org/#/c/285852/ | 16:34 |
dtantsur | I did one review round for it, I want it to come closer to the driver composition (or the opposite - folks driving the driver composition closer to it) | 16:34 |
krotscheck | jroll, betherly, sambetts: Should ironic-ui be publishing its docs, or should it be a part of the ironic docs? I'm leaning former. https://review.openstack.org/#/c/305661/ | 16:35 |
krotscheck | ppiela ^^ | 16:35 |
sambetts | krotscheck: I think it should be publishing its own docs | 16:35 |
*** electrofelix has joined #openstack-ironic | 16:36 | |
jroll | krotscheck: what docs does ironic-ui have? I also lean former | 16:36 |
sambetts | krotscheck: you've already got your own in tree, so its easy to add the publish job to project-config | 16:36 |
rloo | dtantsur: i just looked at your comment. seems reasonable. i thought someone said that they wanted that patch split up into smaller patches though. | 16:36 |
*** ohamada has quit IRC | 16:38 | |
betherly | krotscheck: jroll sambetts i think it should be the former as well | 16:38 |
betherly | on my to do list for tomorrow to look into sorting that then | 16:38 |
rloo | jroll, krotscheck: I think it makes sense to be part of the ironic docs; although I don't know what 'part of the ironic docs' means. Do you mean the docs that ironic keeps intree and that end up at http://docs.openstack.org/developer/ironic/ | 16:38 |
dtantsur | ironic-inspector has separate docs, so it might make sense for ironic-ui as well.. | 16:39 |
jroll | rloo: I guess I think it depends what the docs contain :) | 16:39 |
rloo | so i don't think i want ironic-ui documentation source in ironic tree. | 16:39 |
jroll | agree | 16:40 |
rloo | but if there were a central place for users to find documentation about ironic + friends, I think that the ironic ui docs should be available there. | 16:40 |
jroll | I think the question is should it be at /developer/ironic/ or /developer/ironic-ui/ | 16:40 |
mat128 | sambetts mgould created this on their bug tracker | 16:40 |
mat128 | http://forum.tinycorelinux.net/index.php/topic,19951.0.html | 16:40 |
rloo | jroll: i would say /developer/ironic-ui then. | 16:40 |
betherly | rloo: jroll maybe if the ironic docs has a section added to provide links to subproject docs? | 16:40 |
sambetts | jroll: I think ironic-ui, and we can add a link in the ironic docs to it | 16:40 |
mat128 | had to go through 8 impossible captchas simply to post a bug :) | 16:40 |
jroll | rloo: yeah, I agree, we can link from main docs | 16:40 |
* jroll == sambetts == betherly == rloo | 16:41 | |
betherly | lol | 16:41 |
sambetts | mat128: Nice! | 16:41 |
rloo | ++ betherly | 16:41 |
mat128 | that bug tracker was hard to find :P | 16:41 |
sambetts | hehe great minds and all that | 16:41 |
betherly | jroll: rloo ppiela krotscheck so tomorrow i will look into sorting ironic-ui docs and let you know here when thats done and ready for it to be linked from ironic docs | 16:42 |
mat128 | I have a feeling it's going to be either a) not answered, or b) be told 6.x is not supported anymore | 16:42 |
rloo | betherly: is there ironic-ui information available from https://wiki.openstack.org/wiki/Ironic? | 16:42 |
sambetts | mat128, mgould: tbh at some point we might want to move to tinycorelinux v7 instead of v6 | 16:42 |
betherly | jroll: do you want me to submit a patch to update ironic docs too with space for subproject linking? | 16:42 |
sambetts | mat128, mgould: newer kernel etc | 16:42 |
betherly | rloo: hmm doesnt seem to be | 16:43 |
jroll | betherly: it's super easy, and sure go for it | 16:43 |
mat128 | sambetts: yeah, I figured. Let me check real quick but I'm pretty sure gdisk still pulls icu | 16:43 |
jlvillal | mgould: I am bumping up the number of tinyipa VMs to 7. | 16:44 |
rloo | betherly: might be good to update that page then. you should be able to do it yourself. | 16:44 |
sambetts | mat128: yup it does | 16:44 |
jlvillal | mgould: I had switched to tinyipa but forget to increase from the 1 VM. | 16:44 |
mat128 | why are we using gdisk again? | 16:44 |
jlvillal | mgould: I also changed from 1024MB per VM to 384MB | 16:44 |
sambetts | mat128: for gpt partitioning | 16:44 |
jlvillal | mgould: Rebuilding now... | 16:45 |
betherly | jroll: rloo will do on both accoutns | 16:45 |
jroll | betherly: I think you want https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L6766 and https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/projects.yaml#L2449 | 16:46 |
jroll | but I may be wrong | 16:46 |
jroll | jlvillal: you're gonna need a mitaka patch | 16:46 |
vdrok | jlvillal: done the same for smoke, the only difference - I run with 3 vms, do we need more? all the tests do 3 vms max I think | 16:46 |
jroll | jlvillal: https://review.openstack.org/#/c/314662/ | 16:46 |
betherly | jroll: thanks :) | 16:47 |
jroll | np | 16:47 |
jlvillal | jroll: Thanks. Will check it out. | 16:47 |
*** sacharya has joined #openstack-ironic | 16:47 | |
jlvillal | vdrok: Okay. So as long as we have at least 3 VMs we should be good? | 16:48 |
jroll | vdrok: well, more VMs will help with the capacity problems you're seeing :) | 16:48 |
mat128 | sambetts: https://review.openstack.org/#/c/296466/3 makes it so that it doesnt seem to use gdisk anymore, and gdisk isn't used in the 2 Depends-On or in ironic-lib directly | 16:48 |
devananda | jroll, betherly: fwiw, I think the deployer documentation (http://docs.openstack.org/developer/ironic/deploy/install-guide.html) should mention the UI and describe how to install it | 16:48 |
jroll | devananda: I'm not opposed to that | 16:49 |
*** e0ne has joined #openstack-ironic | 16:49 | |
vdrok | I'm seeing some really weird things now - using clustered compute manager from ironic tree, after termination, used fields, like local_gb_used are set to 0, all the other fields are OK. after some time, ALL the fields are set to 0, like memory_mb, local_gb. Then after some more time, things are back to normal. Has anyone seens this? | 16:49 |
* krotscheck just -2'd his first patch! | 16:49 | |
* krotscheck feels POWERFUUUUL | 16:49 | |
devananda | krotscheck: congradulations! :) | 16:49 |
jroll | krotscheck: and here you didn't want core :D | 16:50 |
mgould | krotscheck, I hope it wasn't one of mine :-) | 16:50 |
mgould | mat128++ | 16:50 |
betherly | devananda: jroll does it need to be a repeat of the ironic ui docs or can we just link to them (once the link works lol) | 16:50 |
*** sacharya has quit IRC | 16:50 | |
krotscheck | jroll: I still don't, because I'm keenly aware of how I'd abuse it :D | 16:50 |
mat128 | vdrok: maybe this? https://bugs.launchpad.net/nova/+bug/1579213 | 16:50 |
openstack | Launchpad bug 1579213 in OpenStack Compute (nova) "ComputeFilter fails because compute node has not been heard from in a while" [Undecided,New] | 16:50 |
jroll | betherly: links would be fine | 16:50 |
*** derekh has quit IRC | 16:50 | |
devananda | betherly: links ++ | 16:50 |
*** sacharya has joined #openstack-ironic | 16:50 | |
vdrok | mat128: in this case, memory_mb is 0, so ram filter fails :( | 16:51 |
krotscheck | jroll, betherly, devananda: To be fair, it's the docs patch which tries to point ironic-ui's docs link to ironic's docs, rather than actually publishing ironic-ui's docs. | 16:51 |
betherly | krotscheck: hoping it wasn't one of mine :p | 16:51 |
devananda | betherly: also, what do ya'll plan to put in the ironic-ui docs? | 16:51 |
mgould | OK, I need to go - jlvillal, let me know how you get on (I'll leave this machine connected) | 16:51 |
betherly | devananda: installation notes and features for now | 16:51 |
mat128 | vdrok: thought 0 was the symptom of "not updated recently enough" | 16:51 |
betherly | any additions welcome :) | 16:52 |
jlvillal | mgould: Thanks! And good night | 16:52 |
jroll | krotscheck: <3 | 16:52 |
jroll | best -2 ever | 16:52 |
*** stendulker_ has quit IRC | 16:53 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for dib based agent ramdisk in lib/ironic https://review.openstack.org/264579 | 16:54 |
*** rama_y has quit IRC | 16:54 | |
*** rama_y has joined #openstack-ironic | 16:54 | |
mat128 | mgould sambetts: my bad, using your browser's find function in gerrit doesnt seem to work very well, ironic-lib starts using gdisk once the changes are applied, thus the ramdisk still needs it | 16:55 |
*** mgould has quit IRC | 16:55 | |
sambetts | mat128: yeah :( shame I think I've managed to get somewhere with manually building it instead though | 16:56 |
JayF | mat128: sambetts: switch to the unified diff | 16:56 |
JayF | mat128: sambetts: if you want inline search to work | 16:56 |
sambetts | ++ | 16:57 |
vdrok | jroll: mat128: not sure, but I definitely need to look at this resource tracking. here is a paste of what happens in my case - http://paste.openstack.org/show/496597/ | 16:57 |
vdrok | which is the reason smoke tests fail sometimes | 16:57 |
jroll | vdrok: oooooo | 16:58 |
jroll | I feel like I know the answer here | 16:58 |
jroll | does CCM just set the used fields on tear down? | 16:58 |
*** mgould has joined #openstack-ironic | 16:58 | |
*** mgould is now known as mgould|afk | 16:59 | |
vdrok | jroll: https://github.com/openstack/ironic/blob/master/ironic/nova/compute/manager.py#L83-L92 | 16:59 |
jroll | vdrok: here's the thing https://github.com/openstack/nova/blob/master/nova/virt/ironic/driver.py#L312 | 16:59 |
*** penick has quit IRC | 16:59 | |
vdrok | oh | 16:59 |
jroll | that's why it drops to 0 during cleaning | 16:59 |
vdrok | yeah, that explains things :) | 17:00 |
openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 17:00 |
vdrok | thanks jroll | 17:00 |
*** ChrisAusten has joined #openstack-ironic | 17:00 | |
jroll | vdrok: also since the resource tracker will "correct" us, that's why you see 0 used, 384 available for a short time, I think | 17:00 |
*** daemontool_ has joined #openstack-ironic | 17:00 | |
jroll | it sees there's no instance and changes it to 0 used | 17:00 |
*** causten_ has joined #openstack-ironic | 17:01 | |
vdrok | so it seems we still need to wait for actual resources set correctly, not for cleaning finish | 17:01 |
jroll | right | 17:01 |
*** aparnav has quit IRC | 17:02 | |
*** causten_ has quit IRC | 17:03 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc: Add unittests for the manager.py module https://review.openstack.org/314681 | 17:04 |
*** ChrisAusten has quit IRC | 17:04 | |
*** causten_ has joined #openstack-ironic | 17:04 | |
JayF | lucasagomes: sorta been chewing on what you said earlier. I don't think we should use virtualbmc in the gate until pyghmi starts following more ironic-style review processes (rather than the instamerge that tends to happen now) | 17:05 |
JayF | lucasagomes: and has better test coverage | 17:05 |
*** e0ne has quit IRC | 17:06 | |
lucasagomes | JayF, I agree with that, we need to check with jjohnson2 | 17:06 |
lucasagomes | JayF, I'm trying to get vbmc enough test coverage as well | 17:06 |
jjohnson2 | lucasagomes, ok | 17:06 |
jlvillal | sambetts: Do you know the status on moving the gate to use tinyipa? | 17:07 |
lucasagomes | jroll, (if jjohnson2 agrees) do you think pyghmi's could be part of the ironic umbrealla? | 17:08 |
*** sacharya has quit IRC | 17:08 | |
sambetts | jlvillal: I think JayF might have put patches together to do it | 17:08 |
*** sacharya has joined #openstack-ironic | 17:08 | |
JayF | jlvillal: I have some patches up, incomplete, on my top 3 list to finsih | 17:09 |
jlvillal | JayF: sambetts Thanks | 17:10 |
JayF | https://review.openstack.org/#/c/313799/ is the frist of them, to make ipxe config explicit so we can make ipxe/tinyipa teh default in devstack without chanign gate | 17:10 |
lucasagomes | JayF, one thing we can do as well, is to pin on specific version of pyghmi | 17:10 |
*** jaybeale has joined #openstack-ironic | 17:10 | |
JayF | https://review.openstack.org/#/c/313800/ is the second, but needs the projects.yaml updated | 17:10 |
lucasagomes | or set a upper-bound at least | 17:10 |
lucasagomes | (which is the same we do for other things such as PrettyTable etc) | 17:10 |
JayF | lucasagomes: I mean, I don't think that really gets to the heart of my objection, unless we really treat pyghmi 100% as an external lib | 17:11 |
lucasagomes | JayF, that's kinda how we treat it now, pyghmi's in the global-requirements (and has a upper-bound) | 17:12 |
lucasagomes | what I mean is, that will give virtualbmc stability and we can work on pyghmi and improve it in parallel | 17:12 |
*** e0ne has joined #openstack-ironic | 17:13 | |
jroll | lucasagomes: yeah, totally re: pyghmi | 17:13 |
jjohnson2 | lucasagomes, JayF I am missing part of the conversation... If I interpret 'ironic-style review process' right, there's two interesting points: 1) I haven't had a lot of volunteers to code review 2) there's also a lot of functionality that ironic doesn't yet touch that are probably even more boring for the general audience | 17:14 |
jroll | JayF: putting a vbmc gate on pyghmi solves a decent amount of that issue, no? | 17:14 |
jjohnson2 | I certainly wouldn't mind that | 17:15 |
JayF | jroll: if we're treating pyghmi 100% as an external lib, any concerns I have are mostly gone | 17:15 |
jroll | JayF: sure | 17:16 |
devananda | JayF jroll: at the moment, I think we should be treating pyghmi like we treat other hardware drivers / libs, eg. proliantutils | 17:16 |
lucasagomes | there's a job already, it's non-voting for now until we get some stability | 17:16 |
*** athomas has quit IRC | 17:17 | |
JayF | devananda: I agree; but I think that relationship would change if we used vbmc in the gate which is built on pyghmi, which is why I started the conversation initially | 17:17 |
devananda | JayF: ah. I agree. | 17:17 |
*** causten_ has quit IRC | 17:17 | |
jjohnson2 | lucasagomes, you might want to do a testcommit that will actualy pass the gate, since I'm bogged down from fixing the python3 commit | 17:17 |
devananda | hadn't read all the backscroll, sorry | 17:17 |
lucasagomes | jjohnson2, yup | 17:17 |
lucasagomes | there are some already | 17:17 |
lucasagomes | jjohnson2, e.g https://review.openstack.org/#/c/286217/ | 17:17 |
vdrok | jroll, jlvillal, then maybe we can do the following wrt smoke - set the vm count to 6, and during instance termination, wait for ironic node cleaning to finish. This way, we don't have to wait for resource tracker, and we won't use more memory/cpu as we will have 3 vms max running at the same time. WDYT? Or wait for actual hypervisor-show command's correct | 17:17 |
vdrok | output? | 17:17 |
jroll | vdrok: I'm fine with that, we can fit 3x 384MB VMs in the gate | 17:18 |
jroll | er. | 17:18 |
jroll | 7x | 17:18 |
jjohnson2 | lucasagomes, ah, ok | 17:18 |
vdrok | aha, great :) | 17:19 |
lucasagomes | jjohnson2, so the job is passing, since it does not block the gate yet (non-voting) we should keep an eye on it before merging any patches | 17:19 |
JayF | jroll: should I change my tinyipa in gate patch to lower ram from 512 to 384? | 17:19 |
JayF | jroll: it would mean it would fail gate if we ever made tinyipa too big, which I think could be a good thing | 17:19 |
jroll | JayF: idk if it matters much, unless we need the space | 17:19 |
jroll | could be | 17:19 |
jroll | I guess I'm not opinionated | 17:19 |
jjohnson2 | I still don't get notified on pyghmi reviews... | 17:20 |
devananda | jjohnson2: you can change your subscription settings in gerrit | 17:21 |
*** jistr has quit IRC | 17:21 | |
JayF | jroll: the question boiled down is: do we want to fail a tinyipa gate if we can't run the agent in 384mb of ram | 17:21 |
jroll | JayF: right, I'm not sure | 17:22 |
vdrok | If we're going to use 7 vms for grenade, I think we should :) | 17:24 |
*** kromanenko_ has quit IRC | 17:24 | |
*** ChubYann has joined #openstack-ironic | 17:24 | |
sambetts | JayF: won't it fail anyway? because it won't even boot | 17:24 |
*** daemontool_ has quit IRC | 17:25 | |
jroll | vdrok: yeah, that's a good point | 17:25 |
jroll | JayF: sure, do it, or at least the -src job | 17:25 |
jjohnson2 | hehdevah, thanks | 17:25 |
jjohnson2 | devananda, thanks | 17:25 |
vdrok | jroll: otoh, if it will be only 3 at the same time, we can do 512? | 17:25 |
jroll | vdrok: right, we need to decide which is best | 17:26 |
vdrok | 3 running at the same time I mean | 17:26 |
jjohnson2 | pyghmi is interesting in general, it's in openstack, but has some commitments beyond (e.g. I'm not allowed to drop python 2.6 testing) | 17:27 |
lucasagomes | oh | 17:28 |
jjohnson2 | there's some non-cloud users of RHEL6 that won't even consider EL7 at this juncture | 17:29 |
lucasagomes | I think you can keep then, I mean, not sure if OpenStack would enforce dropping 2.6 support to the libraries | 17:29 |
openstackgerrit | Merged openstack/pyghmi: Remove unused pngmath Sphinx extension https://review.openstack.org/286217 | 17:29 |
jjohnson2 | of course, those same users are served through rpms... | 17:29 |
sambetts | OMG OMG OMG OMG OMG OMG First tinyIPA boot at 256MB!!! waiting for devstack to prepare so I can tempest test it but ZOMG! | 17:30 |
*** e0ne has quit IRC | 17:30 | |
*** penick has joined #openstack-ironic | 17:30 | |
dims | sambetts : wow, someone is excited :) | 17:30 |
sambetts | Much excitment :D | 17:30 |
jjohnson2 | hmm, I've been spending some of my day working on our HPC bootstrap, it's about 17 MB filesystem, got downright bloated after adding python ;) | 17:31 |
JayF | sambetts: only if I limited it to 384mb (or maybe 256 now? lol) in teh gate | 17:31 |
JayF | sambetts: that's what I was asking: should the size be 512 or 384 | 17:31 |
sambetts | JayF: ah right, missed that bit | 17:31 |
* sambetts -> gym, then I'm going to try to push a new patch to IPA with the improvements after local testing | 17:33 | |
*** sambetts is now known as sambetts|afk | 17:33 | |
jjohnson2 | oh well, back to writing my filesystem for a while | 17:33 |
jjohnson2 | this is a nice prelude for me manifesting hardware management api as a filesystem.... it's at least fun | 17:34 |
jjohnson2 | maybe not so much the 'profit' | 17:34 |
*** hshiina has quit IRC | 17:36 | |
*** spandhe has quit IRC | 17:38 | |
dtantsur | https://review.openstack.org/#/c/313511/ passed gates, could use +W | 17:42 |
dtantsur | JayF, ^^^ | 17:43 |
lucasagomes | jjohnson2, heh which filesystem are you writing? | 17:43 |
JayF | done | 17:44 |
dtantsur | our of memory in tinyipa gate, wow: http://logs.openstack.org/62/314662/1/check/gate-tempest-dsvm-ironic-pxe_ipa-tinyipa-nv/c607a1a/logs/ironic-bm-logs/baremetalbrbm_0_console_10-05-2016-17:24:24.txt.gz | 17:45 |
dtantsur | sambetts|afk, ^^^ | 17:45 |
dtantsur | jroll, do you want to recheck https://review.openstack.org/#/c/314662/ due to ^^^? | 17:46 |
*** keedya has joined #openstack-ironic | 17:46 | |
openstackgerrit | Lucas Alvares Gomes proposed openstack/virtualbmc: Add unittests for the manager.py module https://review.openstack.org/314681 | 17:46 |
jroll | dtantsur: O_o | 17:46 |
lucasagomes | dtantsur, ouch, weird | 17:47 |
dtantsur | I suspect that low memory is great, but we should not go too low in gate :) | 17:47 |
dtantsur | still, intriguing | 17:47 |
jroll | that's with 512 | 17:47 |
jroll | also keep in mind that's stable | 17:47 |
jroll | JayF: stable branches are probably a reason to keep it at 512 for most things :) | 17:47 |
jroll | stable/mitaka anyway | 17:48 |
*** lucasagomes is now known as lucas-afk | 17:49 | |
dtantsur | g'night, see you tomorrow | 17:49 |
jroll | night! | 17:50 |
* dtantsur will be on irc for some time, but can be considered gone for a day :) | 17:50 | |
jjohnson2 | lucas-afk, right now, something to provide a utility that only understands file I/O http access, long story | 17:50 |
JayF | jroll: we force ramdisk to coreos for stable; we could force size for stable too | 17:50 |
jroll | dtantsur: don't say that, I'll keep pinging you | 17:50 |
jroll | JayF: yeah, it's just harder because project-config | 17:50 |
jroll | but it is possible | 17:50 |
JayF | jroll: honestly I kinda have an itch to refactor how it works to make the ramdisk type dictate size | 17:50 |
jjohnson2 | lucas-afk, but I also have thoughts of having a confluent client fs. Already the CLI presents things like a filesystem when at its prompt, might be fun to extend out | 17:50 |
*** absubram has joined #openstack-ironic | 17:50 | |
dtantsur | after saying that I'm free to ignore any pings :-P | 17:50 |
jroll | JayF: not a bad idea | 17:51 |
* vdrok leaves too, good night! | 17:52 | |
*** thiagop has quit IRC | 17:52 | |
*** lucas-afk is now known as lucasagomes | 17:55 | |
lucasagomes | jjohnson2, sounds kinda cool | 17:55 |
*** garthb has joined #openstack-ironic | 17:57 | |
lucasagomes | will call it a day as well, have a great evening (ugt) all! | 17:57 |
*** lucasagomes is now known as lucas-dinner | 17:57 | |
jjohnson2 | lucasagomes, yeah, I really need to make a new video covering the state of things | 17:57 |
jjohnson2 | lucas-dinner, have a good evening | 17:57 |
*** Sukhdev_ has joined #openstack-ironic | 17:57 | |
NobodyCam | have a good night lucas-dinner | 17:58 |
*** Sukhdev has quit IRC | 18:00 | |
*** Sukhdev_ has quit IRC | 18:02 | |
*** ChrisAusten has joined #openstack-ironic | 18:02 | |
*** e0ne has joined #openstack-ironic | 18:03 | |
*** rajinir has quit IRC | 18:06 | |
*** Nisha_away has quit IRC | 18:09 | |
*** Nisha has quit IRC | 18:10 | |
*** causten_ has joined #openstack-ironic | 18:10 | |
*** jaybeale has quit IRC | 18:12 | |
*** ChrisAusten has quit IRC | 18:12 | |
*** rcernin has quit IRC | 18:13 | |
openstackgerrit | Merged openstack/ironic-python-agent: [inspection] wait for the PXE DHCP by default and remove the carrier check https://review.openstack.org/313511 | 18:17 |
*** electrofelix has quit IRC | 18:19 | |
*** sacharya has quit IRC | 18:31 | |
*** baoli_ has quit IRC | 18:32 | |
dtantsur | jroll, hope you don't mind backporting ^^^ (patch https://review.openstack.org/#/c/314713/) | 18:35 |
* dtantsur disappears now | 18:36 | |
NobodyCam | night dtantsur | 18:36 |
*** dtantsur is now known as dtantsur|afk | 18:36 | |
dtantsur|afk | o/ | 18:36 |
jroll | dtantsur|afk: yeah, it's fine, remind me tomorrow if I don't do it now | 18:36 |
mat128 | dtantsur: you can always use a newer IPA with an older Ironic, that's what we do here | 18:36 |
jroll | now/today | 18:36 |
dtantsur|afk | mat128, a bit trickier for distributions like RDO which strictly follow upstream branches... | 18:37 |
mat128 | oh, right | 18:37 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/313729 | 18:37 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/314354 | 18:37 |
*** penick has quit IRC | 18:41 | |
*** penick has joined #openstack-ironic | 18:41 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/313756 | 18:42 |
*** joprovost has joined #openstack-ironic | 18:43 | |
*** rcernin has joined #openstack-ironic | 18:43 | |
*** jcoufal has quit IRC | 18:47 | |
*** thiagop has joined #openstack-ironic | 18:48 | |
openstackgerrit | Mathieu Mitchell proposed openstack/ironic: Allow configuring shred's final overwrite with zeros https://review.openstack.org/304101 | 18:53 |
*** rajinir has joined #openstack-ironic | 18:53 | |
jjohnson2 | Well, now I know for certain pyghmi can juggle 2,000 real servers from a single process | 18:55 |
jjohnson2 | though the keepalive to keep sessions alive alone are enough to take up 25% of a core... | 18:56 |
*** piet has quit IRC | 18:58 | |
*** fragatina has quit IRC | 19:02 | |
*** spandhe has joined #openstack-ironic | 19:03 | |
*** jaybeale has joined #openstack-ironic | 19:07 | |
openstackgerrit | Merged openstack/ironic-lib: Fix coverage option and execution https://review.openstack.org/305557 | 19:08 |
*** jaybeale_ has joined #openstack-ironic | 19:09 | |
*** fragatina has joined #openstack-ironic | 19:09 | |
*** jaybeale has quit IRC | 19:11 | |
cinerama | jlvillal: how far are you getting with the grenade run lately? | 19:12 |
*** pcaruana has quit IRC | 19:12 | |
*** piet has joined #openstack-ironic | 19:13 | |
jlvillal | cinerama: I'm running some tests now. My latest changes seemed to have made me take a step back. | 19:13 |
jlvillal | Trying to figure out which change it is. | 19:13 |
openstackgerrit | Mathieu Mitchell proposed openstack/ironic: Allow configuring shred's final overwrite with zeros https://review.openstack.org/304101 | 19:13 |
jlvillal | cinerama: It is either the 7 VM nodes are a patch I am merging. | 19:13 |
jlvillal | I think | 19:13 |
cinerama | mm. i am getting failures in the base smoke often, and sometimes in the next bit after that | 19:14 |
jlvillal | cinerama: My latest changes are causing failures in base smoke. | 19:17 |
jlvillal | So reverting some changes. | 19:17 |
mat128 | sambetts|afk, mgould|afk: there is hope! http://forum.tinycorelinux.net/index.php/topic,19951.0.html | 19:18 |
cinerama | jlvillal: cool, i'll try again when you land that | 19:18 |
*** sacharya has joined #openstack-ironic | 19:22 | |
*** causten_ has quit IRC | 19:24 | |
*** jaybeale_ has quit IRC | 19:29 | |
*** jaybeale has joined #openstack-ironic | 19:29 | |
openstackgerrit | Ruby Loo proposed openstack/ironic-specs: Priorities that were done in Mitaka https://review.openstack.org/314739 | 19:35 |
*** jaybeale has quit IRC | 19:36 | |
*** rcernin has quit IRC | 19:51 | |
*** piet has quit IRC | 19:51 | |
*** daemontool_ has joined #openstack-ironic | 19:51 | |
*** e0ne has quit IRC | 19:53 | |
*** rcernin has joined #openstack-ironic | 19:59 | |
*** afaranha has quit IRC | 20:00 | |
*** dprince has quit IRC | 20:05 | |
*** daemontool_ has quit IRC | 20:06 | |
*** absubram has quit IRC | 20:06 | |
sambetts|afk | mat128: thats cool! | 20:10 |
*** Goneri has quit IRC | 20:12 | |
*** sacharya_ has joined #openstack-ironic | 20:13 | |
openstackgerrit | Merged openstack/ironic-python-agent: Remove oslo.i18n https://review.openstack.org/313586 | 20:13 |
*** sacharya has quit IRC | 20:16 | |
*** rama_y has quit IRC | 20:17 | |
xavierr | guys, have you faced some problem with agent deploy images these days? | 20:18 |
jroll | xavierr: you'll need to be more specific | 20:19 |
sambetts|afk | dtantsur|afk, jroll: woah?! that log with the out of memory error is booting the coreos image!!!!! | 20:20 |
*** rama_y has joined #openstack-ironic | 20:21 | |
sambetts|afk | dtantsur|afk, jroll: even though its the tinyipa job... WTF?! | 20:21 |
xavierr | I'm trying to boot a baremetal using coreos_production_pxe.vmlinuz and coreos_production_pxe_image-oem.cpio.gz, but the machine stucks when trying to load kernel image | 20:22 |
devananda | xavierr: can you observe / record the machine's console during boot? | 20:23 |
jroll | sambetts|afk: waaaat | 20:23 |
xavierr | devananda, yeap | 20:23 |
sambetts|afk | jroll: http://logs.openstack.org/62/314662/1/check/gate-tempest-dsvm-ironic-pxe_ipa-tinyipa-nv/c607a1a/logs/ironic-bm-logs/baremetalbrbm_0_console_10-05-2016-17:24:24.txt.gz you can tell because the coreos kernel outputs [<seconds since boot>] before each line and tinyipa doesn't | 20:24 |
xavierr | devananda, let me upload the print screen | 20:24 |
sambetts|afk | jroll: thats why it out of memoryed, because coreos doesn | 20:24 |
jroll | sambetts|afk: right, looking | 20:24 |
sambetts|afk | doesn't fit in 512mb | 20:25 |
openstackgerrit | Sean Dague proposed openstack/ironic: DNM: testing paramiko bump https://review.openstack.org/314756 | 20:26 |
jroll | sambetts|afk: oh, we don't support downloading tinyipa on stable/mitaka | 20:26 |
jroll | feel free to backport that | 20:26 |
jroll | but that would be why | 20:26 |
sambetts|afk | jroll: oooo of course!!! | 20:26 |
*** fragatina has quit IRC | 20:27 | |
*** keedya has quit IRC | 20:27 | |
*** aNupoisc has joined #openstack-ironic | 20:27 | |
xavierr | devananda, https://drive.google.com/open?id=0Bx3iHZG5D0neMk5kbDAxeHd4aWc | 20:27 |
sambetts|afk | jroll: we'll have to trigger the post job on IPA stable/mitaka to provide one too | 20:28 |
xavierr | devananda, thats all | 20:28 |
*** ijw_ has joined #openstack-ironic | 20:28 | |
*** ijw has quit IRC | 20:28 | |
jroll | sambetts|afk: right | 20:29 |
xavierr | devananda, after that the machine stops and reboot | 20:29 |
jroll | sambetts|afk: something to hack on tomorrow, go beer | 20:29 |
*** joprovost has quit IRC | 20:33 | |
*** absubram has joined #openstack-ironic | 20:33 | |
*** baoli has joined #openstack-ironic | 20:36 | |
*** baoli has quit IRC | 20:41 | |
*** baoli has joined #openstack-ironic | 20:42 | |
sambetts|afk | jroll: heh :-P cya tomorrow | 20:44 |
jroll | sambetts|afk: thanks for confirming we can land this, though :D | 20:45 |
JayF | sambetts|afk: jroll: That means that my changes, which force !tinyipa is stable/mitaka or earlier, would fix this too | 20:45 |
openstackgerrit | Kurt Taylor proposed openstack/ironic-specs: Revise third party continuous integration spec for new drivers https://review.openstack.org/314768 | 20:46 |
jroll | JayF: well, how does that fix "tinyipa job doesn't actually run tinyipa on mitaka"? | 20:46 |
JayF | oh. OH | 20:46 |
jroll | :) | 20:46 |
JayF | tinyipa job shouldn't run against mitaka ever, right? | 20:46 |
JayF | because we didn't support it in mitake | 20:46 |
JayF | seems like the most anti-stable thing possible to start gating a thing we didn't support on release post-release | 20:47 |
JayF | I mean really we don't support it now because it doesn't gate anything (working on it) | 20:47 |
*** baoli has quit IRC | 20:48 | |
jroll | JayF: we did though | 20:49 |
sambetts|afk | JayF: we support tinyipa build from source in stable/mitaka | 20:49 |
*** KennethWilke has quit IRC | 20:49 | |
jroll | we just didn't support *downloading* tinyipa in mitaka | 20:49 |
sambetts|afk | just not prebuilt | 20:49 |
jroll | (apparently) | 20:49 |
JayF | I guess I don't understand then | 20:49 |
JayF | because in OpenStack it's been hammered over and over | 20:49 |
JayF | if you don't gate it, you don't support it | 20:49 |
*** baoli has joined #openstack-ironic | 20:49 | |
jroll | we supported tinyipa built from source | 20:49 |
JayF | and tinyipa has zero voting jobs; even the -src job on ipa is nonvoting | 20:49 |
jroll | well, it wasn't voting | 20:49 |
jroll | here's my actual issue | 20:50 |
*** baoli has quit IRC | 20:50 | |
jroll | grenade starts with tests on $previous_stable and upgrades to current | 20:50 |
openstackgerrit | Merged openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/314353 | 20:50 |
jroll | therefore, I need tinyipa working on $previous_stable, in this case mitaka | 20:50 |
jroll | and tinyipa does exist in mitaka, and the job was previously okay, until we made it fetch from tarballs.o.o | 20:51 |
jroll | so I want to backport the patch that allows that | 20:51 |
jroll | make sense? | 20:51 |
*** lucas-dinner has quit IRC | 20:51 | |
JayF | That makes some sense; would it not work to just make a -src job for stable/mitaka since that was supported at that point? | 20:51 |
*** baoli has joined #openstack-ironic | 20:52 | |
jroll | well, it would be more of adding the special things to the grenade job to make it a -src thing | 20:53 |
jroll | which will just make it longer and more prone to failure because internets | 20:54 |
JayF | gotcha | 20:54 |
JayF | I'm like +1 in pragmatism | 20:54 |
JayF | but it does break the pattern | 20:54 |
sambetts|afk | a merge into IPA stable/mitaka will generate a tinyipa-mitaka.tar.gz on tarballs, so we just need to backport the devstack change to make it useable :) | 20:54 |
JayF | sambetts|afk: wanna push a commit to both master and stable updating the readme to say it's not experimental anymore? | 20:55 |
sambetts|afk | sure :) | 20:55 |
*** KennethWilke has joined #openstack-ironic | 20:55 | |
*** baoli has quit IRC | 20:57 | |
*** lucasagomes has joined #openstack-ironic | 20:58 | |
*** baoli has joined #openstack-ironic | 21:00 | |
JayF | Did anyone else want a pixie boots sticker? Just need an address to mail them to | 21:00 |
JayF | I'm going downstairs for stamps and to mail off the first batch in about 20m, so lmk soon or you'll have to wait longer :) | 21:01 |
*** baoli_ has joined #openstack-ironic | 21:01 | |
devananda | JayF: o/ | 21:01 |
JayF | devananda: pm me your address + how many stickers | 21:01 |
JayF | devananda: only request is that you know the stickers will be used :) | 21:01 |
openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/313729 | 21:02 |
mat128 | JayF: I'm poking around to know who really wants one | 21:04 |
JayF | mat128: wanna hit me up in pm with an address? I mean, I'm assuming if I gave you an extra one or two someone would use it given you all use the software | 21:04 |
*** baoli has quit IRC | 21:04 | |
JayF | mat128: just have to make sure to put all the onmetal stickers I'll send you on your laptop too /s | 21:04 |
mat128 | Haha | 21:05 |
mat128 | Will do :) Walking around the office with this | 21:05 |
mat128 | Let's trade stickers for pictures of stickers :) | 21:05 |
JayF | hahaha | 21:05 |
*** rama_y has quit IRC | 21:05 | |
openstackgerrit | Merged openstack/ironic: Adds RAID interface for 'iscsi_ilo' https://review.openstack.org/314016 | 21:06 |
*** rama_y has joined #openstack-ironic | 21:06 | |
*** garthb has quit IRC | 21:09 | |
*** wajdi_ has joined #openstack-ironic | 21:11 | |
*** rama_y has quit IRC | 21:11 | |
*** wajdi has quit IRC | 21:15 | |
*** wajdi_ has quit IRC | 21:16 | |
*** ametts has quit IRC | 21:17 | |
*** garthb has joined #openstack-ironic | 21:17 | |
*** jjohnson2 has quit IRC | 21:17 | |
openstackgerrit | Merged openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/313756 | 21:20 |
*** [1]cdearborn has joined #openstack-ironic | 21:20 | |
openstackgerrit | Devananda van der Veen proposed openstack/ironic: Update the api-ref documentation for Ports https://review.openstack.org/313187 | 21:23 |
openstackgerrit | Devananda van der Veen proposed openstack/ironic: Update the api-ref documentation for Root and Nodes https://review.openstack.org/312795 | 21:23 |
*** mbound has joined #openstack-ironic | 21:30 | |
openstackgerrit | Sam Betts proposed openstack/ironic-python-agent: TinyIPA: Shave off some file size from tinyipa ramdisk https://review.openstack.org/313675 | 21:35 |
*** daemontool_ has joined #openstack-ironic | 21:35 | |
*** cdearborn has quit IRC | 21:36 | |
sambetts|afk | night all | 21:38 |
*** ijw_ has quit IRC | 21:41 | |
jroll | jlvillal: so that nw info cache thing.... | 21:42 |
jroll | isn't an issue at that bit | 21:42 |
jroll | but rather | 21:42 |
jroll | 2016-05-09 19:59:57.781 | + /opt/stack/new/grenade/projects/60_nova/resources.sh:create:82 : openstack server create --image cirros-0.3.4-x86_64-uec --flavor baremetal --security-group nova_grenade --nic net-id=cfd19d5d-7d90-4552-a40b-01e3ff74a463 nova_server1 --wait | 21:42 |
jroll | 2016-05-09 20:00:04.173 | Error creating server: nova_server1 | 21:42 |
jroll | which on my latest appears to be issues building http://logs.openstack.org/65/311865/3/experimental/gate-grenade-dsvm-ironic/e635dec/logs/old/screen-n-cpu.txt.gz?level=WARNING | 21:43 |
*** milan has quit IRC | 21:44 | |
*** milan has joined #openstack-ironic | 21:44 | |
jroll | and now that I look, grenade creates a new network for resource create | 21:46 |
jroll | which I assume we don't tap the conductor into https://github.com/openstack/ironic/blob/master/devstack/lib/ironic#L653 | 21:46 |
jroll | jlvillal: we don't have a grenade plugin yet, right? | 21:47 |
jroll | oh, we do, devstack/upgrade | 21:49 |
jroll | but we need to do this before nova resource create, after neutron resource create | 21:49 |
jroll | LE SIGH | 21:49 |
NobodyCam | :( | 21:50 |
*** ijw has joined #openstack-ironic | 21:51 | |
*** sacharya has joined #openstack-ironic | 21:52 | |
*** run_dfc has quit IRC | 21:52 | |
mrda | Morning Ironic | 21:53 |
*** amotoki has joined #openstack-ironic | 21:53 | |
NobodyCam | hey hey morning mrda | 21:53 |
jroll | hi mrda | 21:54 |
*** sacharya_ has quit IRC | 21:54 | |
*** daemontool_ has quit IRC | 21:55 | |
*** fragatina has joined #openstack-ironic | 21:57 | |
*** rama_y has joined #openstack-ironic | 21:57 | |
mrda | hey NobodyCam and jroll o/ | 21:57 |
NobodyCam | :) | 21:57 |
*** rbudden has quit IRC | 22:11 | |
*** piet has joined #openstack-ironic | 22:14 | |
*** [1]cdearborn has quit IRC | 22:15 | |
xavierr | Morning mrda | 22:17 |
*** xavierr is now known as xavierr_pizza | 22:19 | |
*** rloo has quit IRC | 22:22 | |
*** raildo is now known as raildo-afk | 22:38 | |
rama_y | JayF: Hello? | 22:50 |
rama_y | jroll: hello | 22:51 |
JayF | rama_y: Is there a reason you just tried to DCC me a file? On most public networks that's considered abusive without asking first | 22:59 |
rama_y | JayF: I wanted to chat. | 22:59 |
rama_y | JayF: to be specific, ask a question. | 23:00 |
JayF | If there's something you'd like to ask, feel free to ask here and if someone (maybe me?) knows the answer I'm sure they'll try to help | 23:00 |
rama_y | I have a question regarding the metrics patch: #/c/301526/ and the spec for adding pluggable metrics backend for ironic & ipa. | 23:00 |
rama_y | Any suggestions how OSprofiler can be used as additional backend(?) or engine? | 23:00 |
*** baoli_ has quit IRC | 23:01 | |
*** Goneri has joined #openstack-ironic | 23:03 | |
JayF | rama_y: I mean, the code is up, it is pretty straightforward to implement a backend if the correct primitives exist | 23:03 |
JayF | not sure what the question is | 23:03 |
rama_y | JayF: could you be more specific about the primitives? | 23:04 |
JayF | I mean, it's all spelled out in the spec, not sure what value I'll add telling you in IRC as well :P | 23:05 |
JayF | https://review.openstack.org/#/c/301526/13/ironic_lib/metrics.py line 94 | 23:06 |
JayF | you'd just have to implement a metricslogger class for osprofiler | 23:06 |
rama_y | JayF: thanks very much for the input. | 23:07 |
rama_y | JayF: that means, it would be part of ironic-lib, am I correct? | 23:08 |
JayF | yes | 23:08 |
rama_y | JayF: thanks! | 23:08 |
JayF | I'd suggest filing an RFE bug for it, and just having the patch depend on alineb's patch | 23:09 |
JayF | also by reviewing her patch you can help metrics as a whole get in | 23:09 |
rama_y | JayF: I filed this RFE earlier: https://bugs.launchpad.net/ironic/+bug/1560704 | 23:13 |
openstack | Launchpad bug 1560704 in Ironic "[RFE] Ironic support for OSprofiler" [Wishlist,In progress] - Assigned to Ramamani Yeleswarapu (ramamani-yeleswarapu) | 23:13 |
rama_y | JayF: thanks for the suggestions. | 23:17 |
JayF | np, good luck | 23:18 |
rama_y | Thanks! | 23:19 |
*** d0ugal has quit IRC | 23:21 | |
*** absubram has quit IRC | 23:26 | |
*** jaybeale has joined #openstack-ironic | 23:26 | |
*** rcernin has quit IRC | 23:28 | |
*** Goneri has quit IRC | 23:30 | |
*** sacharya has quit IRC | 23:34 | |
*** sacharya has joined #openstack-ironic | 23:35 | |
*** spandhe has quit IRC | 23:36 | |
*** d0ugal has joined #openstack-ironic | 23:36 | |
*** spandhe has joined #openstack-ironic | 23:38 | |
jlvillal | JayF: Do you think 'OSProfiler' can be a 'backend'? Is OSProfiler the same type of thing as 'statsd' for example? | 23:39 |
jlvillal | I'm somewhat confused. | 23:39 |
JayF | jlvillal: if it fits into the model, sure it can. That's why I caveated it with 'if it implements the proper primitives' | 23:40 |
jlvillal | JayF: I guess I'm thinking of 'OSProfiler' as maybe an 'engine' and not a 'backend'. | 23:42 |
jlvillal | From reading the code the 'backend' seems to be something that records the statistics. | 23:42 |
JayF | jlvillal: I'm not an expert on OSProfiler | 23:43 |
* jlvillal is not either :( | 23:43 | |
JayF | that's why I pointed the interested person at the code | 23:43 |
NobodyCam | I seem to recall that using functions in function definition's was a bad thing? | 23:44 |
jlvillal | NobodyCam: I'm not sure I understand. | 23:46 |
NobodyCam | def function(self, node_id, mac=utils.generate_mac_address(), params='') | 23:46 |
jlvillal | NobodyCam: Hmmm. I'm not sure. Not sure if it is like doing foo={} | 23:47 |
jlvillal | As in bad. | 23:47 |
NobodyCam | I seem to recall something about it.. but didn't find anything with a quick google | 23:48 |
*** piet has quit IRC | 23:51 | |
*** crinkle has quit IRC | 23:55 | |
*** chlong has joined #openstack-ironic | 23:56 | |
jlvillal | NobodyCam: My gut feeling is that it should be okay. But I have nothing to back it up :) | 23:58 |
*** garthb has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!