Friday, 2016-04-08

konetzedThanks for the help. I think I will take a serious look at liberty release and swing in tomorrow AM to ask my question again. See if cant get some of those EU guys to answer :D00:02
*** david-lyle has quit IRC00:02
*** david-lyle has joined #openstack-ironic00:03
*** rama_y has quit IRC00:06
openstackgerritOpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements  https://review.openstack.org/30285300:10
*** sukhdev has quit IRC00:12
*** saripurigopi has joined #openstack-ironic00:12
JayFNobodyCam: lucasagomes: FWIW, I'm very okay with https://review.openstack.org/#/c/291427 going in without a bug number, despite my comment on the patch, especially given the wide consensus.00:15
*** shakamunyi has joined #openstack-ironic00:18
*** garthb has quit IRC00:21
*** rbudden has quit IRC00:22
*** rbudden has joined #openstack-ironic00:23
*** shakamunyi has quit IRC00:23
openstackgerritOpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements  https://review.openstack.org/30285100:27
openstackgerritOpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements  https://review.openstack.org/30285200:27
*** spandhe has quit IRC00:28
*** jayc has quit IRC00:29
openstackgerritOpenStack Proposal Bot proposed openstack/python-ironic-inspector-client: Updated from global requirements  https://review.openstack.org/30287900:32
openstackgerritOpenStack Proposal Bot proposed openstack/python-ironicclient: Updated from global requirements  https://review.openstack.org/30288000:32
openstackgerritMarcellin Fom Tchassem proposed openstack/ironic-inspector: Make tox respect upper-constraints.txt  https://review.openstack.org/30170000:33
*** jayc has joined #openstack-ironic00:41
*** jjohnson2 has quit IRC00:41
*** rbudden has quit IRC00:45
*** hoangcx_ has joined #openstack-ironic00:49
*** hoangcx has quit IRC00:49
*** hoangcx_ is now known as hoangcx00:49
*** cdearborn has joined #openstack-ironic00:52
*** amotoki has quit IRC00:56
*** spandhe has joined #openstack-ironic01:01
*** smoriya_ has joined #openstack-ironic01:06
*** spandhe has quit IRC01:08
*** [1]cdearborn has joined #openstack-ironic01:09
*** penick has quit IRC01:16
jlvillallintan, ping?01:17
jlvillallintan, It would be great to get your input on: https://review.openstack.org/303048    Thanks01:18
*** shakamunyi has joined #openstack-ironic01:18
*** amotoki has joined #openstack-ironic01:22
*** shakamunyi has quit IRC01:23
*** cdearborn has quit IRC01:24
*** davidlenwell has quit IRC01:28
*** amotoki has quit IRC01:29
*** [1]cdearborn has quit IRC01:30
*** davidlenwell has joined #openstack-ironic01:32
*** baoli has joined #openstack-ironic01:32
*** Haomeng|2 has joined #openstack-ironic01:35
*** saripurigopi has quit IRC01:36
*** baoli has quit IRC01:38
*** saripurigopi has joined #openstack-ironic01:40
*** shu-mutou has joined #openstack-ironic01:46
*** baoli has joined #openstack-ironic01:47
lintanOK, let me check jivillal01:48
*** baoli has quit IRC01:52
openstackgerritTan Lin proposed openstack/ironic: Add nodeinfo_list() to Node Object  https://review.openstack.org/29031302:00
*** baoli has joined #openstack-ironic02:00
*** saripurigopi has quit IRC02:03
*** baoli has quit IRC02:07
*** baoli has joined #openstack-ironic02:08
*** davideag_ has joined #openstack-ironic02:13
*** mtanino__ has joined #openstack-ironic02:15
*** suro-patz has joined #openstack-ironic02:15
*** Sukhdev_ has quit IRC02:15
*** davideagnello has quit IRC02:15
*** jtomasek has quit IRC02:15
*** mtanino has quit IRC02:15
*** baoli has quit IRC02:16
*** jaybeale has quit IRC02:17
*** baoli has joined #openstack-ironic02:17
*** amotoki has joined #openstack-ironic02:17
*** jaybeale has joined #openstack-ironic02:18
*** vishwana_ has quit IRC02:18
*** shakamunyi has joined #openstack-ironic02:19
*** openstackstatus has quit IRC02:19
*** aginwala has quit IRC02:19
*** vishwanathj has joined #openstack-ironic02:19
*** JieLi has joined #openstack-ironic02:22
*** shakamunyi has quit IRC02:24
*** baoli has quit IRC02:27
*** aginwala has joined #openstack-ironic02:29
openstackgerritMerged openstack/bifrost: Updated from global requirements  https://review.openstack.org/30284102:29
*** suro-patz has quit IRC02:35
*** jaybeale has quit IRC02:36
*** aginwala has quit IRC02:39
*** JieLi has quit IRC02:39
*** suro-patz has joined #openstack-ironic02:42
*** yuanying has quit IRC02:46
*** suro-patz has quit IRC02:50
*** aginwala has joined #openstack-ironic02:52
*** Marga_ has quit IRC02:53
*** aginwala has quit IRC02:59
*** Goneri has joined #openstack-ironic03:04
openstackgerritTan Lin proposed openstack/ironic: Make use of oslo-config-generator  https://review.openstack.org/24733103:17
*** shakamunyi has joined #openstack-ironic03:18
jlvillallintan, If you would like to make those changes and add yourself as a co-author, feel free :)03:18
jlvillalOtherwise I could work on it tomorrow.03:19
jlvillallintan, Thanks for the review!03:19
*** shakamunyi has quit IRC03:23
openstackgerritMerged openstack/ironic: Adjust ipminative._reboot to comply with pyghmi contract  https://review.openstack.org/26499003:24
openstackgerritDao Cong Tien proposed openstack/ironic-specs: Nova serial console support  https://review.openstack.org/29686903:26
*** Marga_ has joined #openstack-ironic03:27
lintanOK, I can update this :)03:31
*** Marga_ has quit IRC03:31
*** Sukhdev has joined #openstack-ironic03:34
*** Goneri has quit IRC03:37
*** Marga_ has joined #openstack-ironic03:45
openstackgerritTan Lin proposed openstack/ironic: Prepare for transition to oslo-config-generator  https://review.openstack.org/30304803:46
*** Sukhdev has quit IRC03:47
*** yuanying has joined #openstack-ironic03:49
*** links has joined #openstack-ironic03:53
*** saripurigopi has joined #openstack-ironic04:04
*** harlowja_at_home has joined #openstack-ironic04:13
*** Sukhdev has joined #openstack-ironic04:16
*** keedya has quit IRC04:16
*** garthb has joined #openstack-ironic04:37
*** harlowja_at_home has quit IRC04:39
*** aginwala has joined #openstack-ironic04:39
*** mtanino__ has quit IRC04:48
*** UForgotten has joined #openstack-ironic04:50
*** rcernin has joined #openstack-ironic04:55
*** alex_xu has quit IRC04:56
*** shu-mutou is now known as shu-mutou-AFK04:57
*** UForgotten has quit IRC04:59
*** UForgotten has joined #openstack-ironic04:59
*** stevebaker has joined #openstack-ironic05:01
*** alex_xu has joined #openstack-ironic05:01
*** garthb has quit IRC05:02
openstackgerritTan Lin proposed openstack/ironic: Prepare for transition to oslo-config-generator  https://review.openstack.org/30304805:06
*** links has quit IRC05:07
*** rbrud has quit IRC05:12
*** ChubYann has quit IRC05:18
*** shakamunyi has joined #openstack-ironic05:18
*** shakamunyi has quit IRC05:23
*** links has joined #openstack-ironic05:24
*** moshele has joined #openstack-ironic05:27
*** suro-patz has joined #openstack-ironic05:40
*** marios has joined #openstack-ironic05:46
*** aginwala has quit IRC05:53
*** Sukhdev has quit IRC05:54
*** mvpnitesh has quit IRC05:56
*** mvpnitesh has joined #openstack-ironic05:56
*** saripurigopi has quit IRC05:58
*** saripurigopi has joined #openstack-ironic05:58
*** saripurigopi has quit IRC06:02
*** links has quit IRC06:03
*** amotoki has quit IRC06:06
*** shakamunyi has joined #openstack-ironic06:18
*** links has joined #openstack-ironic06:19
*** shakamunyi has quit IRC06:23
*** ndipanov has joined #openstack-ironic06:29
*** amotoki has joined #openstack-ironic06:29
*** amotoki has quit IRC06:33
*** amotoki has joined #openstack-ironic06:33
*** fangfenghua has joined #openstack-ironic06:44
*** fangfenghua has left #openstack-ironic06:44
*** ndipanov has quit IRC06:47
*** amotoki has quit IRC06:52
*** yuanying has quit IRC06:57
*** praneshp_ has quit IRC07:01
*** mrda is now known as mrda_away07:01
*** suro-patz has quit IRC07:01
*** e0ne has joined #openstack-ironic07:01
*** jtomasek_ has joined #openstack-ironic07:03
*** jtomasek_ has quit IRC07:04
*** amotoki has joined #openstack-ironic07:05
*** dtantsur|afk is now known as dtantsur07:12
dtantsurMorning Ironic07:14
dtantsurkonetzed, hi! I work on ironic-discoverd, feel free to ask your questions :) ironic-discoverd did work in kilo for me, and I still support it internally07:14
dtantsur(not it's called ironic-inspector, but it's only compatible with liberty and newer)07:15
dtantsur* now07:15
dtantsurkonetzed, I'd also recommend considering liberty, if possible. ironic-inspector got a lot of improvements, and so did ironic.07:17
*** shakamunyi has joined #openstack-ironic07:18
*** shakamunyi has quit IRC07:23
*** ndipanov has joined #openstack-ironic07:28
*** alex_xu_ has joined #openstack-ironic07:32
*** alex_xu has quit IRC07:34
*** jtomasek has joined #openstack-ironic07:38
*** ifarkas has joined #openstack-ironic07:38
openstackgerritMerged openstack/ironic-inspector: Updated from global requirements  https://review.openstack.org/30285207:48
*** pcaruana has joined #openstack-ironic07:49
*** ohamada has joined #openstack-ironic07:52
openstackgerritMerged openstack/python-ironic-inspector-client: Updated from global requirements  https://review.openstack.org/30287907:54
openstackgerritMerged openstack/python-ironicclient: Updated from global requirements  https://review.openstack.org/30288007:54
*** ohamada has quit IRC07:54
*** ohamada has joined #openstack-ironic07:57
aarefievmorning!07:58
aarefievdtantsur: morning, do we want to run discovery test on the existing job? I think about separate one07:59
dtantsuraarefiev, hmm, yeah, maybe. still, project-config is worth fixing first07:59
aarefievok08:00
*** jistr has joined #openstack-ironic08:28
*** derekh has joined #openstack-ironic08:28
*** dtantsur is now known as dtantsur|brb08:30
*** e0ne has quit IRC08:32
*** openstackgerrit has quit IRC08:33
*** athomas has joined #openstack-ironic08:33
*** kromanenko has joined #openstack-ironic08:33
*** openstackgerrit has joined #openstack-ironic08:33
*** mgould has joined #openstack-ironic08:37
mgouldmorning Ironicers08:39
*** e0ne has joined #openstack-ironic08:40
*** vmud213 has joined #openstack-ironic08:40
openstackgerritAnton Arefiev proposed openstack/ironic-inspector: Tempest plugin initial commit  https://review.openstack.org/29392508:49
*** moshele has quit IRC08:53
vdrokmorning dtantsur|brb aarefiev and mgould09:00
aarefievvdrok: morning09:00
*** harlowja has quit IRC09:06
vmud213morning ironicers09:10
sambettsMorning all09:12
sergekMorning All09:13
mgouldvdrok aarefiev vmud213 sambetts sergek morning!09:16
sergeko/09:17
vmud213mgould o/09:17
vdrokmorning vmud213 sambetts and sergek09:17
vmud213morning vdrok09:18
*** dtantsur|brb is now known as dtantsur09:21
*** e0ne has quit IRC09:25
*** divya__ has joined #openstack-ironic09:27
divya__ i am planning to install openstack in my environment..and perfer a automated deployment..suggest me few methods09:28
divya__what are the openstack automate deployment available09:28
sambettsFor what purpose? testing or production?09:29
*** afaranha has quit IRC09:30
*** yuikotak_ has joined #openstack-ironic09:31
mgouldcan someone +W https://review.openstack.org/#/c/301225 please? Already has two +2s, CI passing09:31
divya__hi sam..09:31
divya__its for testing09:31
sambettsdevstack09:31
mgouldmorning divya__09:31
divya__good morning mgould09:31
divya__i need to form a openstack environment.. and not devstack sam09:32
divya__if devstack then easy for me as well09:32
*** yuikotakada has quit IRC09:32
mgoulddivya__, why can't you use devstack?09:32
*** alexpilotti has joined #openstack-ironic09:33
divya__my testing requirement is openstack.. they want to test it in openstack09:33
* mgould suggests https://github.com/openstack/tripleo-quickstart/ as a possible alternative09:33
mgoulddivya__, I don't follow09:34
mgoulddevstack is an installer for openstack09:34
divya__yea mgould.. we are planning to test in real time customer setup.. where all customer might not have devstack..they prefer openstack09:36
*** alexpilotti has quit IRC09:37
sambettswell then you need to find out how your customer is going to install it and use that09:37
*** subscope has joined #openstack-ironic09:38
divya__they use openstack manual deployment..but i dont prefer that09:41
divya__so checking if there is anyother way like devstack to install openstack09:42
*** subscope has quit IRC09:44
*** subscope has joined #openstack-ironic09:45
*** yuikotak_ has quit IRC09:49
mgouldwell, if the environment needs to precisely match your customer's environment, you really need to do what they do09:49
mgouldany automated installer will have the same problem as devstack09:49
*** yuanying has joined #openstack-ironic09:50
*** amotoki has quit IRC09:50
divya__oh ok.. let me create my own script rather than installting manually09:51
mgouldif you just need an OpenStack install but it doesn't need to precisely match your customer's environment, you could try https://wiki.openstack.org/wiki/Packstack or https://github.com/openstack/tripleo-quickstart/ or http://openstack.astokes.org/ maybe09:53
*** yuikotakada has joined #openstack-ironic09:53
mgould[disclaimer: I've only used tripleo-quickstart from that list]09:54
divya__ok mgould09:56
mgouldnp, good luck!09:56
*** ohamada has quit IRC09:59
*** ohamada has joined #openstack-ironic10:00
*** athomas has quit IRC10:07
*** hoangcx has quit IRC10:09
*** subscope has quit IRC10:10
*** athomas has joined #openstack-ironic10:11
*** subscope has joined #openstack-ironic10:15
*** e0ne has joined #openstack-ironic10:20
*** subscope has quit IRC10:20
openstackgerritAline Bousquet proposed openstack/ironic-lib: Add support for metrics  https://review.openstack.org/30152610:26
*** amotoki has joined #openstack-ironic10:41
*** amotoki has quit IRC10:43
sambettsTheJulia: switching to local boot has got things passing in my CI, but I'm still frustrated I don't know what was causing the pxe to fail10:45
*** aswadr_ has joined #openstack-ironic10:47
*** nathan-h has joined #openstack-ironic10:48
*** mgould has quit IRC10:50
nathan-hvdrok: Hey, with regards to all log entries having request id, the request id obviously changes a few times during provisioning, is there any plans to have a master id , that would refer to all entries of a provisioning?10:51
*** amotoki has joined #openstack-ironic10:55
*** mgould has joined #openstack-ironic11:03
vdrokhi nathan-h, lintan is working on this, I think this is the end goal11:03
*** pece has joined #openstack-ironic11:07
nathan-hvdrok: Thanks11:10
*** amotoki has quit IRC11:17
*** shakamunyi has joined #openstack-ironic11:18
*** shakamunyi has quit IRC11:23
*** amotoki has joined #openstack-ironic11:24
*** amotoki has quit IRC11:26
*** alexpilotti has joined #openstack-ironic11:27
*** keedya has joined #openstack-ironic11:28
*** piet_ has joined #openstack-ironic11:30
*** amotoki has joined #openstack-ironic11:30
mat128good morning everyone11:37
*** subscope has joined #openstack-ironic11:38
*** amotoki has quit IRC11:40
openstackgerritOpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements  https://review.openstack.org/30285111:43
divya__good morning mat12811:44
*** divya__ has quit IRC11:44
*** mkovacik has joined #openstack-ironic11:45
*** dprince has joined #openstack-ironic11:47
*** baoli has joined #openstack-ironic11:47
*** baoli_ has joined #openstack-ironic11:48
*** keedya has quit IRC11:50
*** subscope has quit IRC11:50
*** ndipanov has quit IRC11:51
*** baoli has quit IRC11:51
*** Marga_ has quit IRC11:54
*** Marga_ has joined #openstack-ironic11:54
*** sturivnyi has quit IRC11:56
*** sturivnyi has joined #openstack-ironic11:57
*** derekh has quit IRC11:58
*** e0ne has quit IRC11:59
*** ipukha has joined #openstack-ironic12:01
*** subscope has joined #openstack-ironic12:03
*** afaranha has joined #openstack-ironic12:04
TheJuliasambetts: I wouldn't call it pxe failing, the chain loader just couldn't see the disks. :\  I kind of wonder if you were to boot up ye-olde dos and if it would see the disk12:12
TheJuliaoh, and, Good morning everyone!12:12
*** ohamada_ has joined #openstack-ironic12:17
*** ohamada has quit IRC12:17
*** shakamunyi has joined #openstack-ironic12:18
*** alexpilotti has quit IRC12:19
dtantsurmorning TheJulia12:23
*** shakamunyi has quit IRC12:23
*** thrash|g0ne is now known as thrash12:23
openstackgerritAndre Aranha proposed openstack/python-oneviewclient: Added functions to translate between uuid and uri  https://review.openstack.org/30187712:24
sambettsMorning TheJulia, yeah I guess the pxe part was successful, but it was failure from the Ironic deployment perspective because syslinux couldn't boot the OS :(12:24
*** raildo-afk is now known as raildo12:26
*** Marga_ has quit IRC12:26
*** Goneri has joined #openstack-ironic12:29
*** alexpilotti has joined #openstack-ironic12:29
*** trown|outtypewww is now known as trown12:29
*** jjohnson2 has joined #openstack-ironic12:29
*** Goneri has quit IRC12:30
*** alexpilotti has quit IRC12:30
*** alexpilotti has joined #openstack-ironic12:30
*** mtanino has joined #openstack-ironic12:30
*** Goneri has joined #openstack-ironic12:33
openstackgerritAndre Aranha proposed openstack/python-oneviewclient: Added get_server_profile_by_uuid and changed model  https://review.openstack.org/30222512:33
openstackgerritMilan Kováčik proposed openstack/python-ironic-inspector-client: Introspection on stored data  https://review.openstack.org/30219512:34
*** alaski is now known as lascii12:39
vdrokmorning mat128 and TheJulia12:40
openstackgerritMerged openstack/ironic-python-agent: Set modification time in tarfile of ramdisk logs  https://review.openstack.org/30122512:41
*** ndipanov has joined #openstack-ironic12:43
*** derekh has joined #openstack-ironic12:43
TheJuliavdrok: I'm pondering the thought of returning deployfail instead...  I like the tiny option box error represents, and kind of fear that someone seeing deployfail that is working against the same ironic but not aware of nodes being imported might take the wrong action.12:47
*** piet_ has quit IRC12:47
vdrokTheJulia: hm, yep, deployfail additionaly allows transition to active12:49
mat128TheJulia: I think the rational behind it was that error is essentially deletefail12:50
vdrokotoh, if you don't plan to allow it, it does not really matter right>12:50
vdrokso deploing -> deployfail seems more logical than deploying -> erro12:51
vdrok*error12:51
vdrokhave to go, will be back in an hour12:51
*** mtanino has quit IRC12:53
*** smoriya_ has quit IRC12:54
TheJuliamat128: it is, and that is part of what has me conflicted.  A user with a node in adoptfail can only go so many  places in the state machine, a user in an old client, would pretty much be limited to node-delete unless their client decided to start speaking the newer api version :)12:57
openstackgerritAndre Aranha proposed openstack/python-oneviewclient: Extend client to add Connections  https://review.openstack.org/30236512:58
mgouldmorning mat128 TheJulia12:58
TheJuliagood morning mgould12:58
*** e0ne has joined #openstack-ironic12:58
openstackgerritDmitry Tantsur proposed openstack/ironic-inspector: [WIP] Store ramdisk logs on preprocessing failures as well  https://review.openstack.org/30001113:01
jrollmorning friends13:02
TheJuliagood morning jroll13:04
*** jcoufal has joined #openstack-ironic13:04
openstackgerritDmitry Tantsur proposed openstack/ironic-inspector: [WIP] Support Ironic node names in our API  https://review.openstack.org/27633113:07
dtantsurmorning jroll, happy Friday13:07
jrollhappy friday indeed \o/13:08
jrollI realized this morning I need to draft summit schedule stuff13:08
jrollso last call for things there :P13:08
*** saripurigopi has joined #openstack-ironic13:09
mgouldmorning jroll13:09
*** jayc has quit IRC13:09
*** jayc has joined #openstack-ironic13:10
*** xavierr has quit IRC13:10
openstackgerritMerged openstack/ironic: Updated from global requirements  https://review.openstack.org/30285113:10
jroll\o mgould13:11
*** piet has joined #openstack-ironic13:12
thiagopmorning Ironic13:13
thiagopTGIF!13:13
*** [1]cdearborn has joined #openstack-ironic13:13
TheJuliagood morning thiagop piet13:13
pietMorning13:14
mgouldmorning thiagop piet13:14
thiagophiya piet13:14
thiagopTheJulia mgould o/13:14
pietthiagop Hiya13:15
pietWakknig up13:15
jrolldtantsur: we talked about a general inspection session at the summit, you still want that? do you also want the "inspector HA" session?13:16
pietHad a "no show" for a usability this morning13:16
vmud213morning TheJulia, jroll, dtantsur13:16
dtantsurjroll, we need at least one session. we definitely need to talk about our HA, it can be a separate session or part of our only session (which is the most likely, I guess)13:17
dtantsurHA is the main topic we'll be discussing13:17
jrolldtantsur: okay, thinking one session for general, and maybe further HA stuff can be friday morning13:17
jrolldtantsur: fishbowl or workroom for the former?13:17
jrolls/former/general session13:18
* dtantsur tries to remember his plans for the summit13:18
dtantsurjroll, workroom, I guess. Our main topic will be the HA, everything else on Friday.13:19
dtantsurmkoderer, sambetts ^^^13:19
jrolldtantsur: aha, ok13:19
dtantsurjroll, also we need to sync with Fuel folks, but that goes beyond only inspector13:19
dtantsuroops, I planned to ping mkovacik, sorry mkoderer13:20
jrolldtantsur: oh yeah, fun13:20
*** rbudden has joined #openstack-ironic13:21
openstackgerritOpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements  https://review.openstack.org/30285313:22
*** lucasagomes has quit IRC13:22
* mkovacik closes a tab full of "yo dawgs...processing..." pics13:23
TheJuliavdrok: re https://review.openstack.org/#/c/275766/11/doc/source/deploy/adoption.rst comment below line 61, mentioning nova. I'm struggling to try and identify why we would mention nova, espescially since preserve_ephemeral, if an operator chose, is just an instance_info field, but I think that woudl be limited to only the most advanced operators :)13:23
mkovacikdtantsur, OK cool13:23
mat128TheJulia: any thought on managing the instance part? I know this is out of scope for your spec, just wanted to talk about it13:24
*** alexpilotti has quit IRC13:24
jrolldtantsur: one more thing, removing dib ramdisk, I thought we had a plan to drop the test on DIB master and just eat the stable branch breaks when they happen13:24
TheJuliamat128: as in creating a record in nova, someone was mentioning they were working on that idea a few months back, and to make the node visible in nova they would have just to populate ironic's instance_info field with the uuid from nova13:25
*** alexpilotti has joined #openstack-ironic13:25
TheJuliaerr, instance_uuid13:25
mat128yeah, so far (before your spec) we were going to nova boot with --az to land on a specific hypervisor, and have the "fake" driver in Ironic so nothing happens, all of this while using the old api version to end up in the active state13:26
dtantsurjroll, yeah, probably13:26
mat128TheJulia: just trying to replace the fake/old api by adopt stuff13:26
jrolldtantsur: okay, going to skip that session then if that's good with you13:26
dtantsurjroll, TheJulia, also, do we need a session on boot from volume?13:26
dtantsurjroll++13:26
jrolldtantsur: probably? I sadly have not reviewed that spec yet13:27
TheJuliadtantsur: possibly, but the lack of spec reviews for the spec and underlying spec would kind of make it difficul to have a session on it13:27
dtantsurfair13:28
TheJuliaGiven I think we'll likely (i hope) get it approved early on in the cycle, and begin landing substrate, the real discussion for remaining things would end up being at the next major gathering, whatever that is for us :)13:28
*** lucasagomes has joined #openstack-ironic13:28
dtantsursuch lucasagomes wow :)13:28
jrollsambetts: woooo https://www.openstack.org/summit/austin-2016/summit-schedule/events/9491?goback=113:29
*** moshele has joined #openstack-ironic13:29
jrollsambetts: we probably still want our own session on that?13:29
dtantsurdo we?13:30
jrollidk13:30
*** subscope has quit IRC13:30
jrollI'm sure there's plenty of details to work out13:30
dtantsurwe'll definitely won't go over all details in one or two sessions :)13:31
jrollbut maybe we can work most of it out there and follow up in gerrit13:31
jrollya13:31
TheJuliaYeah, but maybe give time for people to bring their wants/needs to the table13:31
openstackgerritMilan Kováčik proposed openstack/python-ironic-inspector-client: Introspection on stored data  https://review.openstack.org/30219513:31
jrollTheJulia: that is to say we should or should not have another session? :)13:32
TheJuliajroll: maybe one, or have a sign and a suggestion box :)13:32
* TheJulia kind of likes the suggestion box idea13:32
jrollheh13:32
*** lucasagomes has quit IRC13:33
TheJuliaThe more I think about it, that might not be a bad way to get some feedback13:33
* TheJulia may also be completely crazy13:34
*** alexpilo_ has joined #openstack-ironic13:34
*** alexpilotti has quit IRC13:35
openstackgerritMilan Kováčik proposed openstack/python-ironic-inspector-client: Introspection on stored data  https://review.openstack.org/30219513:37
*** alexpilotti has joined #openstack-ironic13:37
*** alexpilo_ has quit IRC13:39
*** dims is now known as dimsum__13:41
*** mvpnitesh has quit IRC13:41
*** lucasagomes has joined #openstack-ironic13:42
*** mtanino has joined #openstack-ironic13:42
*** dansmith is now known as superdan13:42
*** piet has quit IRC13:42
jrollTheJulia: I set up a suggestion box at https://bugs.launchpad.net/ironic13:43
jroll:P13:43
openstackgerritMilan Kováčik proposed openstack/python-ironic-inspector-client: Introspection on stored data  https://review.openstack.org/30219513:43
TheJuliajroll: :)13:43
*** mgoddard_ has joined #openstack-ironic13:45
*** mgoddard has quit IRC13:49
*** ndipanov has quit IRC13:50
*** links has quit IRC13:51
openstackgerritMerged openstack/ironic: Allow user to specify cleaning network  https://review.openstack.org/29142713:53
*** hoangcx has joined #openstack-ironic13:53
*** subscope has joined #openstack-ironic13:56
openstackgerritThiago Paiva Brito proposed openstack/ironic-specs: Dynamic allocation of nodes on OneView drivers  https://review.openstack.org/27572613:58
*** jayc has quit IRC13:58
sambettsjroll: awesome! Yeah not sure about having our own session, I'm sure if it comes to it we can always organise a coffee/beer session or something to work things out on the side14:05
sambettson another awesome note, the tinyipa post job has merged!14:06
sambettswe just need an IPA patch to go in to trigger its first build!14:07
jrollsambetts: yeah, we could, and \o/14:08
sambettsjroll: and the  patch to enable non-src tinyipa jobs has merge!14:09
sambettsjroll: Do you know were I can look at the results of the post jobs?14:09
jrollsambetts: logs.o.o/<first two characters of commit sha>/<commit sha>14:12
vdrokmorning jroll and thiagop14:15
thiagophiya vdrok14:15
jroll\o vdrok14:15
thiagopjroll: did you have a fine travel?14:16
jrollthiagop: so far, so good :)14:17
jrollcurrently in the middle of wyoming14:17
NobodyCamgood morning Ironicers14:17
NobodyCamjroll: :) good to hear travel is going well14:18
vdrokmorning NobodyCam14:18
NobodyCammorning vdrok14:18
thiagopjroll: wait, you haven't arrived yet? o.O14:18
jrollhey, good morning NobodyCam14:18
thiagopgood morning NobodyCam14:18
jrollthiagop: we only left wednesday, and taking our time to see all the things :)14:18
jrollit'll be ~38 hours of driving not including stops altogether14:18
NobodyCamthank you for hitting the button on the review14:18
thiagopjroll: woot14:18
vdrokTheJulia: yep, you're right about preserve_ephemeral, ignore that :) As for error/deployfail I think about the following. node gets to adoptfail state, older client sees ERROR state, and cannot do set-provision-state deleted or rebuild, which are allowed for a node that actually is in error state. in case if it is DEPLOYFAIL, he still can not do14:24
vdrokset-provision-state deleted, rebuild and additionally cannot do active. so both options will bring confusion14:24
sambettsjroll: this patch merged yesterday after the post job was in and I don't see any post queue logs https://review.openstack.org/#/c/299560/ http://logs.openstack.org/60/299560/2/14:24
sambettsjroll: Oh yeah I forgot you were still on the road!14:24
TheJuliavdrok: well, the only other option is to reveal new states to the client..14:25
jrollsambetts: it's the commit sha, and if there's a merge commit it's *that* one14:26
jrollso e.g. last commit is at http://logs.openstack.org/2d/2d8e139f0394cdd9b713bcf97bf300ee43165bf7/14:26
jrollseems to be still processing14:26
sambettsjroll: oooh, I see! So its under the commit not the change, I miss understood14:26
vdrokTheJulia: yep. I'm not insisting on doing either of the options, just think that if one sees deployfail after deploying in older versions, it may be more logical. let's wait what others think :)14:28
*** piet has joined #openstack-ironic14:28
jrollsambetts: yep, no worries14:28
sambettsjroll: Do you know how to get requirements added to the post job? mksquashfs is missing :'(14:30
jrolldtantsur: devananda: I've got 6/10 sessions locked in at the bottom of the pad. still need to think more about the rest, might update later, would love your thoughts, we could discuss in the meeting monday as well14:30
jrollsambetts: either install them in your post job script in the repo, or figure out how to add to base images14:30
jrollsambetts: I think there's some sort of extra-requirements.txt that is apt packages that jobs may pick up too? only heard about it briefly so I may be completely off base14:31
TheJuliavdrok: okay, all for other people's input :)14:32
sambettsjroll: I thought you had to do the same thing recently for a missing requirement for the coreos image?14:32
sambettsor was that in the built image14:32
jrollsambetts: I did, we added to the full_trusty_build.sh, which is what the post job runs14:32
jrollI believe we were also missing mksquashfs there :P14:32
sambettsjroll: Ah, I see you've add the logic to install the deps based on which system14:33
jrollsambetts: yeah, I think lucas added the fedora stuff14:34
jrollbut that's unrelated14:34
*** subscope has quit IRC14:34
*** persia has quit IRC14:34
*** daemontool has quit IRC14:35
*** jjohnson2 has quit IRC14:36
*** jjohnson2 has joined #openstack-ironic14:36
jrollalright, gotta bounce for now. catch you all on monday or something :)14:37
sambettscya jroll14:37
*** persia has joined #openstack-ironic14:37
NobodyCamtravel safe jroll14:37
*** trown is now known as trown|relocating14:37
TheJuliatrown|relocating: please tell me your not fleeing the state....14:37
trown|relocatingTheJulia: lol, no just driving downtown, but with our legislature it has crossed my mind :)14:38
TheJulia:)14:38
*** karimb has joined #openstack-ironic14:38
TheJuliatrown|relocating: sadly I know WAY too many people who have already left :(14:39
mgouldjroll, safe travels14:39
*** karimb has quit IRC14:40
*** karimb has joined #openstack-ironic14:41
*** saripurigopi has quit IRC14:42
*** piet has quit IRC14:46
NobodyCammorning mgould :)14:47
NobodyCammorning sambetts :)14:47
NobodyCammorning vdrok :)14:47
sambettso/ NobodyCam14:47
NobodyCamhehehe :)14:47
* sambetts thinks NobodyCam is scripting this stuff now 14:48
NobodyCamlol nope14:48
dtantsurmorning NobodyCam, happy Friday14:48
NobodyCamhey hey dtantsur good morning and totally TGIF!14:48
openstackgerritSam Betts proposed openstack/ironic-python-agent: Add script to install missing tinyipa dependencies  https://review.openstack.org/30346514:48
sambettsdtantsur, jroll, NobodyCam: ^ the tinyipa post job is failing until we merge that14:49
* NobodyCam looks14:50
*** mgoddard has joined #openstack-ironic14:51
*** mgoddard_ has quit IRC14:51
NobodyCamsambetts: not bug #?14:52
sambettsNobodyCam: the tinyipa post job is so new I've not got round to summiting a bug report for it, probably should though for tracking I guess14:53
*** jaybeale has joined #openstack-ironic14:55
mgouldNobodyCam, morning!14:56
openstackgerritSam Betts proposed openstack/ironic-python-agent: Add script to install missing tinyipa dependencies  https://review.openstack.org/30346514:56
sambettsNobodyCam: ^ added a bug :)14:57
NobodyCam:)14:57
NobodyCamwow same package name across all the distros... what are the chances of that,14:59
sambetts:D I know15:00
mgouldit's madness15:01
mgouldcrops won't grow15:02
sambettshahaha15:02
NobodyCamlol15:02
NobodyCam+215:02
sambetts:D thanks NobodyCam15:03
NobodyCamsambetts: you should go to vegas15:03
sambetts:-P15:03
*** lucasagomes has quit IRC15:03
*** vishwanathj has quit IRC15:04
*** subscope has joined #openstack-ironic15:04
konetzeddtantsur: can you run ironic-inspector on the same node that neutron is running on and they share the bridge interface or do you need a different interface for it to run on.  The docs for it really dont clarify that.15:07
konetzeddtantsur: basically i am tryign to run the full stack on one server15:07
* mgould is ssh'ed into a server running both ironic-inspector and neutron as we speak15:08
dtantsurkonetzed, that's what tripleo does. yes, you can. neutron uses namespaces (iirc), so their dhcp servers do not overlap15:08
sambettskonetzed: devstack single node will run inspector and neutron on the same node15:09
konetzedah yes never thought of devstack doing that. Thanks guys I will run through a liberty setup this morning and get ironic-inspector setup :D15:12
*** jcoufal has quit IRC15:12
*** lucasagomes has joined #openstack-ironic15:12
dtantsuryou're welcome :)15:12
*** moshele has quit IRC15:16
sambettsmgould: I never knew about .PHONY thats cool!15:17
mgouldsambetts, :-)15:18
*** cloudnull has joined #openstack-ironic15:18
cloudnulli have a question regarding network setups using ironic. specifically im having an issue with tftp once the node comes online. The host is getting DHCP from neutron and thall seems to work just fine, i can manually retrieve files from tftp using the cli client, but its just not working when PXE booting. Im using the ipmi agent w/ the latest mitaka release.15:21
cloudnullnothing notable in the logs15:21
cloudnullbut the  nodes gets cycled via ipmi, neutron provides dhcp, all seems well, then tftp times out15:22
cloudnullif i boot the node manually install an OS configure the interface to get DHCP it gets the address neutron provides and again all is well.15:22
cloudnulli can then  install a tftp client and retrieve files without issues.15:22
mgouldcloudnull, is the pxe rom being requested?15:23
sambettshow is you network configured? Does the the baremetal machine have a route to the tftp server?15:23
cloudnullsambetts: yes it has a route.15:23
cloudnullthe network within neutron is a flat network15:23
cloudnullon a native vlan15:23
cloudnullmgould: not that i can see15:24
cloudnullthe console shows getting dhcp15:24
cloudnullthen "tftp open timeout"15:24
sambettsif you do a tcp dump on the machine running the tftp server do you see any packets coming in requesting tftp?15:24
sambettsand is the tftp server running?15:24
cloudnulli was doing that last night, i thought i did, let me do it again. might be a min to cycle things15:24
cloudnulltftp is running a15:25
cloudnullinstalled with tftp-hpa15:25
cloudnullw/ ubuntu 14.04.415:25
*** aginwala has joined #openstack-ironic15:25
cloudnullits running the tftp-hpa service , though i have also tried doing it with xinetd15:25
mgouldcloudnull, you could also try adding log-dhcp to dnsmasq-ironic.conf and restarting it15:25
mgouldthat should log whether or not the machine being booted is trying to do the fetch15:26
cloudnulli will add that too15:26
*** garthb has joined #openstack-ironic15:26
mgouldsambetts, is there a special tftp server? I thought dnsmasq did its own tftp serving15:26
sambettsdnsmasq doesn't serve tftp, it just provides the dhcp options to point the pxe firmware to the right place15:27
cloudnullsambetts:  if i ``tftp $IPADRESS`` i can get files without issues from the ironic conductor.15:27
mgouldsambetts, thanks15:27
cloudnullso from a naive test it working15:27
jlvillaldtantsur: FYI: I replied to your comment: https://review.openstack.org/#/c/303048/5//COMMIT_MSG15:27
jlvillalthanks15:28
sambettsin the console of the bm machine, when it trying to pxe, can you see what IP address is being given for the tftp server?15:28
cloudnullno, it just times out.15:28
sambettsmy servers print it out next to its own IP it got from DHCP15:28
cloudnulli have a screen shot15:29
*** aginwala has quit IRC15:29
*** piet has joined #openstack-ironic15:29
cloudnullhttp://snag.gy/GWQJa.jpg15:29
sambettsok, if you neutron port-show the port that the machine is given, you should be able to see the dhcp options, can you see if the tftp server ip is correct15:30
cloudnullthe client ip is what neutron is providing to the node15:30
*** aginwala has joined #openstack-ironic15:30
cloudnulllet me check15:30
*** trown|relocating is now known as trown15:31
openstackgerritvinay kumar muddu proposed openstack/ironic-python-agent: Wait for all interfaces to be up before node lookup  https://review.openstack.org/30350315:33
cloudnullah so this may be my issue, my nodes have multiple nics and the {"opt_value": "$IP", "ip_version": 4, "opt_name": "tftp-server"} is not the correct network for the tftp-server, seems i need to set the my_ip value or similar15:33
*** mkovacik has quit IRC15:33
* cloudnull rereading the ironic.conf sample file15:33
sambettscloudnull: in the ironic.conf there is a tftp ip setting you can configure so you may need to update that15:34
*** saripurigopi has joined #openstack-ironic15:35
*** praneshp_ has joined #openstack-ironic15:41
cloudnullbooting a new node now and will see if that has made it happy. .15:44
cloudnullthanks mgould and sambetts for the advice15:44
sambetts:)15:44
mgouldnp, good luck!15:45
devanandamorning, all15:47
dtantsurmorning devananda15:48
dtantsurand I'm calling it a day, have a great weekend folks15:48
NobodyCamgood morning devananda :)15:49
mgouldmorning devananda15:50
mgoulddtantsur, have a good weekend!15:50
*** dtantsur is now known as dtantsur|afk15:50
*** ifarkas has quit IRC15:50
NobodyCamjust fyi: I just added multiple compute host support to the newton summit etherpad15:50
*** rama_y has joined #openstack-ironic15:51
NobodyCamhave a good weekend dtantsur|afk15:51
openstackgerritJohn L. Villalovos proposed openstack/ironic: Prepare for transition to oslo-config-generator  https://review.openstack.org/30304815:52
openstackgerritJohn L. Villalovos proposed openstack/ironic: WIP/DEMO: Make use of oslo-config-generator  https://review.openstack.org/30305715:52
jlvillaldtantsur|afk: Bye!15:53
*** aginwala has quit IRC15:53
*** aginwala has joined #openstack-ironic15:54
sambettso/ dtantsur|afk15:56
*** aginwala has quit IRC15:56
*** aginwala_ has joined #openstack-ironic15:56
*** bnemec has quit IRC15:56
sambettsNobodyCam: I think (if I remember correctly) jroll left that out because he was pretty confident that we knew what we were doing with it, although I'd confirm with him15:57
NobodyCamsambetts: ya :) I think it may have something to do with node claims stuff but I added just with: anything to discuss / go over with this?15:59
NobodyCamâ—ľorginal spec: https://review.openstack15:59
NobodyCamya15:59
NobodyCamgah15:59
*** ipukha has quit IRC16:00
TheJuliaalthough the filtering in terms of api indexing I think we bumped because there were no good ways we liked16:00
*** e0ne has quit IRC16:00
*** saripurigopi has quit IRC16:01
sambettsNobodyCam: Make sense, I'm really interested in seeing where the claims API goes, because I've got some interesting ideas for abusing it :-P16:01
NobodyCamlol16:01
TheJuliaI think everyone does :)16:01
cloudnullsambetts mgould -- http://snag.gy/MJT1x.jpg -- great success!16:03
cloudnullthanks again16:03
cloudnullnow off to solve other problems :)16:03
*** bnemec has joined #openstack-ironic16:04
*** piet has quit IRC16:05
sambettscloudnull: :D awesome!!16:07
*** kromanenko has quit IRC16:07
*** amotoki has joined #openstack-ironic16:10
cloudnullquick question, besides building images with disk image builder, is there a cache of pre-built images/kernels that I can use for deployments ?16:10
cloudnullim happy to build one, but figured it one existed id rather not go the route of NIH.16:12
sambettscloudnull: http://tarballs.openstack.org/ironic-python-agent/coreos contains prebuilt coreos IPA images, and there will soon be /tinyipa too containing tinycore linux IPA images too16:12
cloudnullexcellent16:13
mgouldcloudnull, hurrah!16:13
*** daemontool has joined #openstack-ironic16:13
*** rbrud has joined #openstack-ironic16:17
*** shakamunyi has joined #openstack-ironic16:18
*** cdearborn has joined #openstack-ironic16:18
*** aginwala_ has quit IRC16:19
*** aginwala has joined #openstack-ironic16:19
*** saripurigopi has joined #openstack-ironic16:20
*** saripurigopi has quit IRC16:21
*** nathan-h has quit IRC16:22
*** shakamunyi has quit IRC16:23
*** aginwala has quit IRC16:24
*** karimb has quit IRC16:27
*** saripurigopi has joined #openstack-ironic16:30
*** aginwala has joined #openstack-ironic16:31
*** [1]cdearborn has quit IRC16:33
*** vmud213 has quit IRC16:33
*** derekh has quit IRC16:33
*** subscope has quit IRC16:35
openstackgerritJohn L. Villalovos proposed openstack/ironic: Prepare for transition to oslo-config-generator  https://review.openstack.org/30304816:41
openstackgerritJohn L. Villalovos proposed openstack/ironic: WIP/DEMO: Make use of oslo-config-generator  https://review.openstack.org/30305716:41
*** baoli_ has quit IRC16:42
*** Marga_ has joined #openstack-ironic16:43
*** harlowja has joined #openstack-ironic16:46
*** ohamada_ has quit IRC16:46
*** pcaruana has quit IRC16:46
*** suro-patz has joined #openstack-ironic16:50
*** KennethWilke has joined #openstack-ironic16:52
*** devananda has quit IRC16:53
NobodyCambrb16:54
*** praneshp_ has quit IRC16:55
*** rcernin has quit IRC16:55
*** e0ne has joined #openstack-ironic16:56
*** e0ne has quit IRC16:57
jlvillalNobodyCam: come back!!!16:59
NobodyCamlol16:59
TheJulialol16:59
*** aginwala has quit IRC17:00
*** aginwala has joined #openstack-ironic17:00
*** aginwala has quit IRC17:01
jlvillalWhew, I missed you NobodyCam :)17:01
*** aginwala has joined #openstack-ironic17:02
jlvillal16 days until Austin....17:02
NobodyCamand here I tought you were going see if I could review something :p17:02
jlvillalNobodyCam: I'd be happy on your thoughts on: https://review.openstack.org/30304817:03
jlvillalBut that is not why I told you to come back :)17:03
SpamapSHi!17:03
SpamapSso, people, I'm playing with the patchset that adds Neutrony things to Ironic17:03
SpamapSand it does something very bad, which I'd like to fix17:04
SpamapSUUIDs do not belong in config files.17:04
SpamapSwe have names in our API's for this purpose.17:04
SpamapSanybody around that knows about the neutron stuff?17:04
jlvillalLink?17:05
TheJuliaare names globally unique in neutron?17:05
*** devananda has joined #openstack-ironic17:06
mat128TheJulia: nope17:07
*** baoli has joined #openstack-ironic17:07
mat128just created two networks with the same name, neutron accepts and lists them17:07
davidlenwellhey SpamapS17:08
mat128SpamapS: How do you disambiguate network names?17:08
sambettsSpamapS: Ironic has had cleaning network uuid in the config file long before the full neutron intergration work17:09
NobodyCamjlvillal: lol17:10
NobodyCammorning SpamapS17:12
SpamapSdavidlenwell: howdy17:13
* mgould is being kicked out of the office by the Royal Scottish Pipers Society, so must bid you all goodnight17:14
SpamapSmat128: you fail if there are two networks in neutron with the same name in your tenant. Because that is _an insane situation_17:14
NobodyCammgould: good night and good weekend... also LOL17:14
SpamapSsambetts: ahh, ok, that was also a maligned idea. :)17:14
SpamapSsambetts: not saying it does not work, but it is a lot of extra complexity for no gain.17:14
mgouldNobodyCam, the hazards of subletting office space :-)17:14
*** e0ne has joined #openstack-ironic17:15
*** mgould has quit IRC17:15
sambettsextra complexity? You mean you have to configure and start Ironic after starting neutron and setting up a network?17:15
*** aginwala has quit IRC17:16
SpamapSsambetts: not saying you shouldn't have to reconfigure it or restart it if you're making networking changes. I'm saying, having to plumb UUIDs into config files is extra complexity.17:16
TheJuliawouldn't that just be a fact in one's configuration management system?17:17
SpamapSafter you have created everything, yes17:17
SpamapSWe went through this in Heat too with tenant and domain ids. The burden you're putting on the config management system is really undue.17:18
TheJuliabut could the creation and then fact generation be completely automated?17:18
SpamapSIt already is fully automated. However, the ordering becomes rigid.17:18
mat128We've had the same issue with swift_account uuid vs account name17:19
SpamapSIt's just extra work17:19
mat128We run things from scratch all the time (automated testing) and ended up having to query the uuid for the name17:19
mat128same pattern17:19
SpamapSAnd there are a _lot_ more divergent config management systems than there are Ironics.17:19
*** aginwala has joined #openstack-ironic17:20
SpamapSalso if you don't want the "fail on two network names" you can also have the "pick random one" or "pick first".17:21
mat128Seeing how neutron just bails out on a multiple match (Multiple network matches found for name 'test', use an ID to be more specific.), that's what I would do if we were going this route17:21
sambettsNova does the same for instance names17:24
TheJuliaSo how would we acurately identify the exact same network every single time?17:24
*** alexpilotti has quit IRC17:24
mat128TheJulia: always lookup, crash on multiple results17:24
mat128you might break the service simply by creating a network though17:24
TheJuliaso then it becomes who creates it first scenario17:25
TheJuliaand what do they create17:25
*** jistr has quit IRC17:25
TheJuliaNot disagreeing, just thinking that just a name may not be the best idea, espescially if I had a custom hardwaremanager doing cleaning that needed access to some resources and my cleaning network could talk to not a single thing in the world beyond the conductors.17:26
TheJuliabut that comes down to operational needs more than anything17:26
NobodyCamjlvillal: this is a quite a ominous note to sneak in: "Its value may be silently ignored in the future".... lol ;)17:26
jlvillalNobodyCam: That is what oslo-config-generator prints.17:27
NobodyCamlol ++17:27
jlvillalBasically trying to get as much of the content to look the same as how it will look with oslo-config-generator so we can actually see what the transition to that really changes.17:27
jlvillalNobodyCam: and yes a little ominous :)17:28
jlvillalNobodyCam: I have this WIP patch where I rebased lintan's patch onto my patch: https://review.openstack.org/#/c/303057/17:28
jlvillalSo you can see what the oslo-generator patch changes to the config file more easily.17:28
*** athomas has quit IRC17:29
NobodyCamjlvillal: line 252 : default_log_levels seem strange to the space for the first one but then not for any of the others... (just looks strange to me)17:30
jlvillalNobodyCam: looking...17:32
jlvillalNobodyCam: I think the only thing that changed was the space added.17:34
NobodyCamoh y :)17:34
NobodyCamya17:34
jlvillalNobodyCam: https://review.openstack.org/#/c/303057/5/etc/ironic/ironic.conf.sample   This shows what happens after oslo-config-generator change over. So no change to that line, so it matches what ocg does.17:35
* sambetts is heading off for the weekend 17:36
NobodyCamhave a great weekend sambetts17:36
jlvillalsambetts: Enjoy!17:36
sambettsSee everyone on Monday :D17:36
*** sambetts is now known as sambetts|afk17:36
sambetts|afko/ NobodyCam jlvillal17:37
SpamapSTheJulia: I think your fears are valid, but the risks are very low.17:37
SpamapSTheJulia: administrators create networks _almost never_17:37
davidlenwellSpamapS: could you be a little more vague?17:38
SpamapSdavidlenwell: it's my speciality17:38
SpamapSsomething is17:38
SpamapSor isn't17:38
SpamapSsometimes17:38
TheJulialol17:38
davidlenwellSpamapS: https://www.youtube.com/watch?v=_P_hKcZgQhY17:38
* TheJulia giggled and was immedialy asked "what are you up to?"17:39
SpamapSTheJulia: the appropriate response is "something"17:39
davidlenwellSpamapS: im sorry the answer we were looking for was "stuff"17:40
SpamapSSo anyway, I think one thing to make clear is, the query for "give me the uuid of this netowrk" should only query the tenant that ironic's service user runs in.17:40
SpamapSproject17:40
TheJuliaSpamapS: that only causes more giggling, fwiw17:40
SpamapSit uses project or it gets the hose17:40
SpamapSdavidlenwell: maybe17:41
NobodyCamlol17:41
davidlenwelllol17:41
davidlenwellSpamapS: thats actually logical.17:41
*** garthb has quit IRC17:42
*** garthb has joined #openstack-ironic17:42
SpamapSTheJulia: anyway, I'm not saying don't _allow_ uuid btw.17:43
SpamapSJust, don't require it.17:43
jlvillalhttps://twitter.com/stevebot/status/71818566770926796917:43
TheJuliaOkay, in that case making it optional just gives more flexibility17:43
TheJuliaforcing it to be a name just makes me think that we're sitting around a table pushing bottle caps of burden back and forth around on a map17:44
TheJuliaSpamapS: I say submit patches! :)17:44
davidlenwell++ TheJulia17:45
NobodyCamTheJulia: sounds like a game of Reversi17:45
*** piet has joined #openstack-ironic17:46
*** praneshp_ has joined #openstack-ironic17:46
TheJuliaNobodyCam: Kind of yeah, but likely with at least three players... not sure what that would be :)17:47
SpamapSTheJulia: oh, I will, I just wanted to make sure I'm not _100%_ nuts on this. :)17:48
TheJuliaSpamapS: well, standard disclaimer, I don't claim to be sane17:48
*** Sukhdev has joined #openstack-ironic17:48
SpamapSYeah, it's a scale17:49
jlvillalI see what the twitter comment means, as an example: http://stackalytics.com/?module=ironic-group&user_id=cody.eilrich@ericsson.com17:51
* jlvillal now goes back to work...17:52
*** e0ne has quit IRC17:52
TheJuliaheh17:53
*** amotoki has quit IRC17:53
mat128jlvillal: can't contribute too much :P17:53
jlvillal:)17:53
*** Marga_ has quit IRC17:57
*** achanda has joined #openstack-ironic17:57
openstackgerritJulia Kreger proposed openstack/ironic: Active Node Creation via adopt state  https://review.openstack.org/27576618:06
TheJuliaJayF: you around today?18:07
*** saripurigopi has quit IRC18:07
JayFTheJulia: ayup18:08
JayFTheJulia: sup?18:08
*** saripurigopi has joined #openstack-ironic18:08
TheJulia270902, are you thinking that we should just update the catch calling _ata_erase()?18:08
JayFI think _ata_erase() should only return BlockDeviceEraseError exceptions18:09
JayFassuming you're talking about the patch I thnk you are18:09
TheJuliayes18:09
TheJuliaokay, simple enough18:09
TheJuliaJust wanted to make sure, since one line versues many others :)18:11
openstackgerritMerged openstack/ironic-lib: Updated from global requirements  https://review.openstack.org/30296918:11
*** achanda has quit IRC18:11
*** lucasagomes has quit IRC18:14
*** saripurigopi has quit IRC18:18
*** saripurigopi has joined #openstack-ironic18:18
*** e0ne has joined #openstack-ironic18:22
*** lucasagomes has joined #openstack-ironic18:22
*** rama_y has quit IRC18:24
*** aginwala has quit IRC18:24
*** alexpilotti has joined #openstack-ironic18:25
*** e0ne has quit IRC18:25
*** e0ne has joined #openstack-ironic18:26
*** e0ne has quit IRC18:26
*** Marga_ has joined #openstack-ironic18:26
*** aginwala has joined #openstack-ironic18:28
*** ChubYann has joined #openstack-ironic18:29
*** alexpilotti has quit IRC18:29
JayFjlvillal: you had mentioned yesterday you might want some help diagnosing some grenade problems, here's your offer of help before I delve into something different :)18:30
jlvillalJayF: I'm actually heading to lunch now...18:30
jlvillalWas about to walk out the door.18:31
JayFjlvillal: enjoy :)18:31
jlvillalI'll see if you are busy when I get back :)  And thanks!18:31
*** vishwanathj has joined #openstack-ironic18:49
*** aginwala has quit IRC18:49
*** aginwala has joined #openstack-ironic18:50
NobodyCamjlvillal:  +2 with a comment on 303048 :)18:53
*** dprince has quit IRC18:56
*** aginwala has quit IRC18:57
*** piet has quit IRC18:59
NobodyCamanyone have a minute to take a look at a method rename patch: https://review.openstack.org/#/c/30222419:01
*** piet has joined #openstack-ironic19:15
*** alexpilotti has joined #openstack-ironic19:26
thiagopFolks, I'm calling it a week19:27
thiagophave a nice weekend you all19:27
*** thiagop has quit IRC19:27
davidlenwellthiagop: thanks you too!19:27
*** alexpilotti has quit IRC19:31
*** daemontool has quit IRC19:33
*** karimb has joined #openstack-ironic19:34
*** daemontool has joined #openstack-ironic19:37
*** vishwanathj has quit IRC19:38
*** vishwanathj has joined #openstack-ironic19:40
*** vishwana_ has joined #openstack-ironic19:41
*** moshele has joined #openstack-ironic19:44
*** vishwanathj has quit IRC19:45
jlvillalNobodyCam: I responded to your comment19:50
NobodyCamjlvillal: totally.. I did put in the comment "I have a comment in line it's more my own ramblings, please feel free to ignore" :)19:52
NobodyCamjlvillal: totally.. I did put in the comment "I have a comment in line it's more my own ramblings, please feel free to ignore" :);p19:52
NobodyCam:p19:52
jlvillalNobodyCam: I saw :)19:52
jlvillalNobodyCam: But I had to reply!  I couldn't control myself...19:52
NobodyCamhehehe19:52
openstackgerritJulia Kreger proposed openstack/ironic-python-agent: Provide fallback from ATA erase to shredding  https://review.openstack.org/27090219:55
*** rbrud has quit IRC19:55
*** bigjools has quit IRC19:56
*** SpamapS has quit IRC19:58
jlvillalNobodyCam: TheJulia: So I'm just adding to the waffling with my comment :(  https://review.openstack.org/#/c/302224/19:58
* TheJulia goes back to demanding we pain the shed purple19:59
TheJuliajlvillal: but very good point :)19:59
TheJuliadifferent word usage I guess19:59
*** aswadr_ has quit IRC20:00
jlvillalTheJulia: I see 'assert' and I just think Python 'assert' statement.20:00
NobodyCamjlvillal: that is a much less bikeshed-y type comment :)20:00
TheJuliahow about _handle_previous_state_names ?20:01
NobodyCammaybe we should ask for a spec on this and iterate there... /me *DUCKS*20:02
*** daemontool has quit IRC20:02
jlvillalTheJulia: sounds sort-of okay. I'm not yelling out, "Yes!" on it :)20:03
* TheJulia tosses things at NobodyCam 20:03
jlvillalNaming is the hard part of computer science...20:03
TheJuliaindeed20:03
TheJulia_handle_backwards_compat_api_states ?20:04
jlvillalTheJulia: update_state_in_older_versions  ?20:04
jlvillalI'm looking at function above.20:04
jlvillalhide_fields_in_newer_versions20:04
jlvillalHow I came up with my suggestion.20:04
JayFdef _this_is_the_review_that_doesnt_end()20:05
JayFdef _yes_it_goes_on_and_on_my_friends()20:05
JayF:)20:05
jlvillalLOL for real here :)20:05
*** bigjools has joined #openstack-ironic20:06
*** bigjools has quit IRC20:06
*** bigjools has joined #openstack-ironic20:06
* jlvillal wonders if the Gerrit '/' search sucks in all browsers or just Firefox. So slow....20:06
TheJuliaJayF: lol20:09
mtreinishjlvillal: it's all browsers20:09
jlvillalmtreinish: Well that makes me feel better and sad at the same time.20:10
mtreinishwell I haven't tried it on lynx, but at least chrome and firefox are really bad with it20:10
* mtreinish tries lynx20:10
JayFwho uses lynx?20:10
JayFelinks ftw20:10
JayFgotta have that sweet, sweet js support20:10
jlvillalI really dislike that doing CTRL-F in my browser does not work reliably in Gerrit.20:10
JayFit does if you go to the unified diff20:11
JayFthat's what I started doing20:11
jlvillalAh20:11
NobodyCamI know we can have a Thunderdome session at summit "two names enter one name...."20:11
JayFnot for all reviews, but for when I'm looking for something in particular20:11
TheJuliaNobodyCam: lol20:11
jlvillalNobodyCam: Good one :)20:12
NobodyCamlol20:12
*** SpamapS has joined #openstack-ironic20:15
*** karimb has quit IRC20:17
JayFTheJulia: re +2'd that IPA patch; thanks!20:20
openstackgerritJarrod Johnson proposed openstack/pyghmi: Implement MegaRAC IPv6 SNMP destinations  https://review.openstack.org/30281720:22
JayFNobodyCam: https://review.openstack.org/#/c/275726 I just added my +2 to this (OneView spec), IDK if any other cores have interest in this, so you might wanna land it now20:23
JayFNobodyCam: I'll leave that to you or someone who is more familiar overall with the hp oneview stuff though20:24
NobodyCamI'd like lucasagomes to be able to have a look at it, as he had a +2 before it changed20:24
NobodyCamso I may hold off to monday20:25
JayFthat works for me20:25
NobodyCam:)20:25
JayFjust figured that one has been through the ringer a little bit20:25
JayFmight make someones' weekend to get it in, lol20:25
NobodyCam++++20:25
*** moshele has quit IRC20:25
NobodyCamyea, but he signed off already20:25
NobodyCamlol20:25
*** daemontool has joined #openstack-ironic20:26
TheJuliaJayF: NP20:28
openstackgerritGonĂ©ri Le Bouder proposed openstack/ironic: iscsi: wipe the disk before deployment  https://review.openstack.org/30360520:32
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Remove deprecated option names in deploy_utils.py  https://review.openstack.org/30308120:36
*** jayc has joined #openstack-ironic20:42
openstackgerritJulia Kreger proposed openstack/ironic: Rename juno name state modification method  https://review.openstack.org/30222420:44
* TheJulia expects new paint colors to be proposed :)20:44
*** baoli_ has joined #openstack-ironic20:46
*** baoli has quit IRC20:47
* NobodyCam clicks20:48
NobodyCamTheJulia: the new color is quite nice20:50
JayFTheJulia: jlvillal had requested a docstring, do you mind adding one real quick before it causes +2s to get ruined?20:50
*** piet has quit IRC20:51
TheJuliaJayF: context please20:53
JayFTheJulia: bikeshed patch20:53
JayFTheJulia: lol20:53
NobodyCami've been calling it hte purple patch20:53
TheJuliaI'm laughing and sighing at the same moment :)20:53
NobodyCamI still like the thunderdome idea20:53
cloudnullhey all, I'm seeing a connection problem with the ironic_python_agent -- http://snag.gy/A8T94.jpg -- I know its because the API is on load balancer VIP address the kicked host does not a route to it. so my question is, is there a way to configure that to be something else within the ironic.conf?20:54
TheJuliaoh, I didn't even see the mention for docstring20:54
cloudnulllooked through the sample file but im not seeing anything obvious20:54
* TheJulia updates20:54
cloudnullthough it s friday so i could be seeing though my friday glasses20:54
JayFcloudnull: configure /what/ to be something else?20:55
cloudnullthe ironic api client seems to want to talk back to the ironic api on my load balancer20:55
cloudnullthe kicked host is on a flat network via neutron20:56
JayFYou still haven't told me what setting you're trying to change, lol20:56
cloudnullJayF: idk thats what im asking .20:56
openstackgerritMario Villaplana proposed openstack/ironic: Add notifications to Ironic  https://review.openstack.org/29846120:56
JayFyou want to add a route to the booted ramdisk? Change the ironic api url it's using? What exactly20:56
cloudnullthe ironic python client is getting that address , from the agent template? is that correct?20:57
JayFAh, I can show you20:57
cloudnullshow me the way20:57
cloudnull:)20:57
JayFhttps://github.com/openstack/ironic/blob/4482eaa59b3ad4f92d7097231100260a487215f4/ironic/drivers/modules/agent_config.template20:59
JayFtrying to find where that's populated though20:59
* TheJulia feels like there is a reference to a musical, but goes back to docstring20:59
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Remove deprecated option names in deploy_utils.py  https://review.openstack.org/30308121:00
JayFcloudnull: https://github.com/openstack/ironic/blob/4482eaa59b3ad4f92d7097231100260a487215f4/ironic/drivers/modules/deploy_utils.py#L98021:00
openstackgerritJulia Kreger proposed openstack/ironic: Rename juno name state modification method  https://review.openstack.org/30222421:00
*** Sukhdev has quit IRC21:00
JayFcloudnull: alternatively, edit that agent_config.template to take whatever static value you want (it's a diff file for ipxe) if you can't/don't want to change that option21:00
cloudnullah ha!21:00
JayFcloudnull: I also could see a bug/rfe for being able to make agents see a different ipa-api-url21:00
cloudnulli can change it , i have the vi technology :)21:01
JayFlol, well it may not be great to have it be static21:01
JayFalthough I guess given it just uses that config option, already is kinda static21:01
cloudnullis it possible for the ironic python agent to configure more than one network which could used for providing a route back to the control plain?21:03
cloudnullasking because I've not tried it21:03
JayFYou can easily build custom IPA images that configure in pretty much whatever way you want21:03
cloudnullok.21:03
JayFif you're using neutron though, I bet you can configure it for that network to push down an extra route over dhcp21:03
cloudnullmakes sense21:03
JayFalthough I'm not 100% sure how you'd do that21:04
cloudnullyea i can configure the hosts, and add routes, etc, though i'd like to keep things totally separated if at all possible.21:04
JayFyep, I mean both ways work21:04
*** daemontool has quit IRC21:05
cloudnulli may go the route of custom image.21:05
JayFthe static editing of the template is likely to get you to "working" faster21:05
*** pece has quit IRC21:05
cloudnulldepends on how much hair i want to give to it21:05
JayFI mean, the custom image stuff is really super simple21:05
*** trown is now known as trown|outtypewww21:05
JayFcloudnull: if you're using coreos builder, it's all in here; https://github.com/openstack/ironic-python-agent/tree/master/imagebuild/coreos21:06
cloudnullsadly i cant use that. though i tried.21:06
JayFWhy not?21:06
cloudnulli need a super new kernel to work with the hosts we have in the OSIC21:06
cloudnullor i need to slipstream the drivers in an older kernel21:07
JayFoh, that'll be easy to fix21:07
*** raildo is now known as raildo-afk21:07
JayFcoreos is by far the quickest route to a ramdisk with shiny new kernels21:07
cloudnullyea. i went down the route of diskimage builder.21:07
JayFcloudnull: https://github.com/openstack/ironic-python-agent/blob/master/imagebuild/coreos/coreos-oem-inject.py + change COREOS_VERSION to something newer.21:07
cloudnullwhich is what i have working now.21:07
JayFcloudnull: for the DIB image, I don't know much about setting that up; we use the coreos one for OnMetal, but either way you can use the pxe template edit to get past your headache21:08
*** piet has joined #openstack-ironic21:10
* JayF goes off looking for requirements patches to blindly +1 to improve his stats /s21:12
cloudnullhahaha cheers21:12
cloudnullthanks for the assist21:12
openstackgerritJay Faulkner proposed openstack/ironic: Remove deprecated option names in deploy_utils.py  https://review.openstack.org/30308121:13
JayFjlvillal: ^ fixed your commit message nit with inline editing if you wanna revote now21:13
jlvillalJayF: Thanks!21:14
*** jjohnson2 has quit IRC21:14
openstackgerritAnup Navare proposed openstack/ironic: Add node UUID to messages/exceptions  https://review.openstack.org/30135021:15
*** karimb has joined #openstack-ironic21:23
JayFmariojv: a thought about the notifications implementation for power state changes; there are some cases where Ironic sees a power state change that it didn't initiate (i.e. if you have the option set where Ironic updates the DB instead of the power state if it polls and the state has changed)21:23
JayFmariojv: should we ensure those cases emit notifications too? i.e. "power state changed without ironic intervention" notifications in addition to "ironic changed the power state" notifications21:23
mariojvmaybe - where does that detection happen?21:24
openstackgerritJohn L. Villalovos proposed openstack/ironic-python-agent: Create helper function: create_hdparm_info()  https://review.openstack.org/30362021:25
jlvillalTheJulia: You may or may not like that patch ^^^^21:26
mariojvthat may be better in a separate notification (or we should rename this notification), on second thought, since this one is explicitly "*set*_power_state"21:27
JayFmariojv: https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L2391 the conf setting it differs on is L247121:27
JayFmariojv: I do think it should be a different notification, but I also think you need this one too or else there's going to be some really confusing results for people who run with force_power_state_during_sync on21:27
JayFmariojv: i.e. they'd see a notification that a node was powered on, never see a power off notification, then see power on again21:28
JayFmariojv: this may be stuff that was better suited to the spec, but was just thinking about how this would look as a consumer21:28
*** alexpilotti has joined #openstack-ironic21:28
mariojvi think at a minimum we definitely need a new start notification here: https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L247821:32
openstackgerritJohn L. Villalovos proposed openstack/ironic: Remove deprecated group and option names in deploy_utils.py  https://review.openstack.org/30308121:32
mariojvsince then you'll see the .end or .fail with no start during the sync21:32
*** rama_y has joined #openstack-ironic21:32
openstackgerritGonĂ©ri Le Bouder proposed openstack/ironic: iscsi: wipe the disk before deployment  https://review.openstack.org/30360521:33
*** cfarquhar_ has quit IRC21:33
*** alexpilotti has quit IRC21:33
JayFmariojv: Just was thinking if the goal of notifications is being able to track node state21:34
mariojvJayF: what do you think about having a new notification like baremetal.node.sync_power_state.{start, fail, end}?21:34
JayFmariojv: you have to track the real-world state change, not just when Ironic pushes a button and makes the change21:34
jlvillalJayF: Updated commit message again :)  ^^^21:34
TheJuliajlvillal: I kind of like it... :)21:34
*** cfarquhar has joined #openstack-ironic21:35
*** cfarquhar has quit IRC21:35
*** cfarquhar has joined #openstack-ironic21:35
jlvillalTheJulia: Thanks. When I saw all those you had to add for your patch I thought, maybe we can make it a little easier...21:35
mariojvthen it'd look something like this: sync start, set power state start, set power state end, sync power state end (in the normal case where it gets out of sync)21:35
JayFmariojv: I don't think that addresses the real thing, and I'd argue a sync_power_state inconsistency isn't DIRECTLY a failure21:35
JayFmariojv: because a tenant can power control in band without Ironic's intervention; the states getting misaligned isn't a failure case, it's just a case21:35
JayFmariojv: but that may be getting more pedantic than needed :)21:36
mariojvright, .start and .end don't indicate failure21:36
mariojv.fail indicates failure21:36
TheJuliajlvillal: there are a few minor aspects i'm letting simmer while I make dinner :)21:36
mariojv:P21:36
jlvillal:)21:36
JayFbut then I don't understand how having the sync power state notifications help then?21:36
mariojvyou can include a new field in those notifs that shows the actual state of the node ironic looked up21:37
mariojvso from a consumer perspective, you see sync_power_state.start21:37
mariojvmaybe sync_power_state.update or whatever we want to call it when we see it's changed and action is needed21:37
JayFcould you emit something like, sync_power_state.update if it updates the db21:37
mariojvthen that goes to utils.node_power_action21:37
JayFand if it just updates teh actual node power state a power action would be triggered21:37
mariojvemits set notifs, then power_sync.end21:38
JayFthe only thing that I would caution about this approach is it changes the notifications to be way more frequent21:38
JayFbecause power state doesn't change nearly as often for a node as the power state sync runs against it21:38
JayFso I guess I think only emitting in the case where the power state is updated in DB || if the power state is changed on the BMC as a result of the sync21:39
mariojvtrue21:39
JayFand AIUI the other piece, which uses power action, would emit already21:39
JayFso you're just missing like a "fake" power state update21:39
JayFyou could probably even add a field in the notification to say whether it was ironic initiated or ironic observed21:39
JayFexcept using better words than I did for that, lol21:39
mariojvmaybe have the .start and .end as DEBUG and the .update or whatever indicates that it's been changed in the hw as INFO21:39
JayFhmm. It'd still be probably one of the noisiest notifications, right?21:40
JayFhave you thought of other examples of debug level notifications?21:40
mariojvdepends on how often you sync21:40
JayFthat's basically a guarantee that for N nodes, by default at least, you'll get N notifications every 5 minutes, right?21:40
JayFpower sync defaults 5m I think21:40
JayF    cfg.IntOpt('sync_power_state_interval',21:41
JayF               default=60,21:41
JayFnope, 60s21:41
JayFso that'd be exceptionally chatty21:41
TheJuliaI was going to say I think it is 1m21:41
mariojvhmm21:41
JayFand either way, as an operator, why do I care if the sync starts?21:41
JayFI only care if it found something anomolous and takes action21:41
mariojvi think i agree21:44
mariojvi viewed DEBUG notifs originally as the spammiest thing you can have, like DEBUG logs21:44
*** piet has quit IRC21:45
JayFSo I think the easiest way to make this make sense21:45
mariojvbut maybe the sync would be too spammy even for that21:45
JayFis to have a field on the notification to say who performed the power action, or something similar21:45
JayFwhen Ironic takes the power action, it would be Ironic21:45
JayFotherwise it could be "node" or something else to indicate the power state changed outside ironic21:45
JayFso when sync_power_state finds an inconsistency, it notifies a power state change21:45
JayFthen would notify again via node_power_action when it rectified it, if you had the option to do so enabled21:46
*** piet has joined #openstack-ironic21:46
JayFmariojv: summarized all this in the review, lmk if you think it reflects things fairly21:48
mariojvi think it's a good idea21:49
JayFI'm just thinking of what we wanted downstream, a kind of "audit log" for node actions21:49
mariojvi think we also want to notify on errors in the sync, in addition to changes, like here https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L241821:49
JayFand without this, you can't really do that, because you're missing an important piece21:49
JayFmariojv: sure; generally, but that'd probably be a different patch than the initial one just dealing with power stuff?21:50
mariojvso if i.e. a BMC network becomes unreachable, you get a lot of error notifs that you can alert on21:50
JayFmariojv: unless we already send error notifications when we fail to change power state?21:50
mariojvwe do, but that piece of code is for when we retrieve the state during the sync21:50
mariojvjust extending that idea of having a full power state history of the node21:51
JayFOK, I think it makes sense all together then21:51
mariojv"I don't know what's here" is another "state"21:51
JayFyeah, exactly21:51
openstackgerritMerged openstack/ironic-python-agent: Updated from global requirements  https://review.openstack.org/30285321:51
JayF+121:51
jlvillalNobodyCam, TheJulia, JayF: What is our policy on proposal bot changes to requirements?  Is it two +2s or one +2?  I'm thinking two +2s, but wanted to be sure.21:53
JayFjlvillal: IDK what the policy is, but we rarely have trouble getting 2x+221:53
JayFwe have enough folks paying attention that they don't get lost in the shuffle21:53
jlvillalYeah. Which reminds me I wonder if the stable gate for python-ironicclient is broken. I see two approved requirements patches which have failed the gate. One for Liberty and the other for Kilo21:54
* jlvillal doesn't have the motivation to check it out on Friday afternoon with gorgeous weather outside....21:55
JayFWouldn't be too surprising, jroll had to do gymnastics to fix the "normal" gate for kilo/liberty21:55
SpamapSHey party people21:58
SpamapSso I have an Ironic running next to a regular VM based nova..21:59
SpamapStrying to use host aggs in flavors to separate things21:59
SpamapSbut the problem I'm seeing is the scheduler that we use for ironic, picks up VM scheduling requests, and fails them because of the exact match filter21:59
*** aNupoisc has joined #openstack-ironic22:00
*** piet has quit IRC22:00
*** lascii is now known as alaski22:02
*** karimb has quit IRC22:02
*** wuhg has joined #openstack-ironic22:07
*** piet has joined #openstack-ironic22:09
*** kragniz has quit IRC22:09
*** kragniz has joined #openstack-ironic22:10
*** MattMan has quit IRC22:13
*** MattMan has joined #openstack-ironic22:13
SpamapSjroll: ^^ any ieas?22:14
SpamapSideas22:14
*** suro-patz has quit IRC22:14
openstackgerritJohn L. Villalovos proposed openstack/ironic-python-agent: Switch tox unit test command to use ostestr  https://review.openstack.org/30363722:15
JayFSpamapS: he's moving right now, gone for most of next week too iirc22:16
JayFSpamapS: Sorry I don't know anything about that stuff or I'd try to help ;(22:16
mtreinishjlvillal: heh, I recognize most of that commit msg :) ^^^22:16
jlvillalmtreinish: :)  Though I have to figure out how to do the 'cover' part22:17
jlvillalI just realized22:17
mtreinishjlvillal: normally I just use setup.py for that and just have a separate command for that tox job22:17
JayFjlvillal: ...since when does IPA have working coverage jobs?22:18
jlvillalmtreinish: Okay. we currently have:  python setup.py testr --coverage {posargs:ironic_python_agent}22:18
*** baoli_ has quit IRC22:18
*** shakamunyi has joined #openstack-ironic22:18
jlvillalJayF: There is a section in there :)22:18
JayFthat has no relevency to if it works ;P22:18
JayFwe used to have a pypy section at one point too22:19
JayFlol22:19
jlvillalmtreinish: Can I just do:  s/testr/ostestr/ ??22:19
mtreinishjlvillal: yeah, that sounds right22:19
mtreinishno22:19
mtreinishbecause calling setup.py is where the --coverage flag is defined22:19
mtreinishit's a pbr thing, ostestr doesn't have an equiv. (I woudln't be opposed to adding it though)22:20
jlvillalmtreinish: Okay. Thanks. I wasn't sure since had removed 'testrepository' as a dependency.22:20
*** karimb has joined #openstack-ironic22:20
mtreinishjlvillal: https://github.com/openstack-dev/pbr/blob/master/pbr/testr_command.py#L117-L13322:21
JayFjlvillal: cover job doesn't work on master22:21
jlvillalJayF: I'm noticing :(22:21
openstackgerritMerged openstack/ironic-python-agent: Provide fallback from ATA erase to shredding  https://review.openstack.org/27090222:21
mtreinishJayF: what's wrong with it?22:21
JayFI don't know, I'm 99% sure it never was made to work22:21
JayFWhen we made IPA there was massive copying of configs, espeically around ci22:22
jlvillalmtreinish: I think the :ironic_python_agent part makes it fail22:22
JayFI suspect this is an artifiact of that22:22
jlvillalpython setup.py testr --coverage {posargs:ironic_python_agent}22:22
jlvillalIf I remove the ':ironic_python_agent' it does run22:22
mtreinishjlvillal: oh I missed that, yeah I would remove that part22:22
JayFjlvillal: that's what I'm seeing, too22:22
jlvillalRan 261 (+260) tests in 0.929s (+0.761s)22:22
jlvillalPASSED (id=132)22:22
jlvillalNo data to report.22:22
JayFyeah, it runs, but it doesn't do anything22:22
*** shakamunyi has quit IRC22:22
JayFCoverage.py warning: Module ironic-python-agent was never imported.22:23
JayFCoverage.py warning: No data was collected.22:23
JayFbefore that seems to be relevant22:23
*** Marga_ has quit IRC22:23
JayFjlvillal: python setup.py testr --coverage --coverage-package-name ironic_python_agent22:23
JayFjlvillal: seems to make it even happier22:23
jlvillalAh cool22:24
mtreinishJayF: yeah, packages where the repo name doesn't match the package name need that flag22:24
JayFyeah, that's the case here22:24
mtreinishbasically everything with dashes has to use it :)22:24
jlvillalJayF:   python setup.py testr --coverage --coverage-package-name ironic_python_agent {posargs}22:25
mtreinishjlvillal, JayF: fwiw, you can set it in a .coveragerc instead of on the cli, like: https://github.com/openstack/tempest-lib/blob/master/.coveragerc22:25
jlvillalerror: option --coverage-package-name not recognized22:26
mtreinishjlvillal: try switching testr to test22:26
JayFfwiw it wfm with testr22:26
JayFfedora 23, ipa master22:26
*** kragniz has quit IRC22:27
jlvillalHmmm. Works with 'test' on Fedora 22.22:27
*** karimb has quit IRC22:27
jlvillalI did remove the 'testrepository' package from the test-requirements.txt if that would make any difference22:27
*** karimb has joined #openstack-ironic22:27
jlvillalJayF: Does 'test' work for you too?22:28
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Centralize config options - [snmp]  https://review.openstack.org/30364422:28
JayFappears to have done the exact same thing22:28
*** Marga__ has joined #openstack-ironic22:28
*** kragniz has joined #openstack-ironic22:28
jlvillalJayF: What do you mean?  Also any interest in pushing your changes as a patch?22:29
mtreinishjlvillal: it's a pbr quirk don't remember what causes it22:29
jlvillalWith an addition to  .gitignore  for 'cover'?22:29
JayFjlvillal: I wasn't making changes, just trying to help you test since I saw the patch and that you said cover had to be taken care of :)22:29
*** alexpilotti has joined #openstack-ironic22:29
JayFjlvillal: I never let abide an opportunity to just see what happens when I do something :)22:30
jlvillalJayF: Okay. I can push a separate patch then.22:30
JayFjlvillal: unless you want me to push a patch fixing cover/22:30
JayFjlvillal: oh, if you're going to do a sep patch I'll do it real quick22:30
jlvillalJayF: If you like too.22:30
JayFsince my repo is already setup for it22:30
jlvillalsounds great :)22:30
jlvillalThe .gitignore should be updated to JayF22:30
*** davidlenwell has quit IRC22:32
openstackgerritJay Faulkner proposed openstack/ironic-python-agent: Fix coverage for IPA  https://review.openstack.org/30364622:32
JayFwow, 97% test coverage22:33
*** Marga__ has quit IRC22:33
JayFI don't know why, but I was expecting that to be much worse22:33
JayFhow can it be showing <100% coverage on *test* files?22:33
*** alexpilotti has quit IRC22:34
*** aNupoisc has quit IRC22:34
*** davidlenwell has joined #openstack-ironic22:40
*** aNupoisc has joined #openstack-ironic22:40
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ssh]  https://review.openstack.org/30364922:41
jlvillalJayF: +2  and thanks.22:42
openstackgerritAnup Navare proposed openstack/ironic: In node_power_action() add node.UUID to log message  https://review.openstack.org/30135022:48
*** kragniz has quit IRC22:49
*** kragniz has joined #openstack-ironic22:49
*** karimb has quit IRC22:53
*** karimb has joined #openstack-ironic22:54
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Centralize config options - [swift]  https://review.openstack.org/30365322:54
*** suro-patz has joined #openstack-ironic22:55
*** Marga_ has joined #openstack-ironic22:56
jlvillalhttps://twitter.com/sempf/status/514473420277694465?lang=en    Probably an old one, but I saw it for the first time this week22:59
*** suro-patz has quit IRC23:00
*** piet has quit IRC23:00
*** karimb has quit IRC23:01
*** cdearborn has quit IRC23:03
*** jayc has quit IRC23:04
openstackgerritRamamani Yeleswarapu proposed openstack/ironic: Centralize config options - [virtualbox]  https://review.openstack.org/30365523:05
*** aNupoisc has quit IRC23:09
*** aNupoisc has joined #openstack-ironic23:12
*** garthb has quit IRC23:12
*** suro-patz has joined #openstack-ironic23:14
*** aNupoisc has quit IRC23:19
*** Marga_ has quit IRC23:25
*** suro-patz has quit IRC23:25
*** suro-patz1 has joined #openstack-ironic23:25
*** Marga_ has joined #openstack-ironic23:26
*** alexpilotti has joined #openstack-ironic23:30
*** alexpilotti has quit IRC23:34
*** suro-patz1 has quit IRC23:50
*** rama_y has quit IRC23:55
*** piet has joined #openstack-ironic23:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!