Tuesday, 2015-07-28

*** penick has quit IRC00:01
*** Sukhdev has quit IRC00:01
*** vishwanathj has quit IRC00:05
*** penick has joined #openstack-ironic00:06
openstackgerritJosh Gachnang proposed openstack/ironic-python-agent: Add command metrics to IPA API  https://review.openstack.org/11998100:07
openstackgerritJosh Gachnang proposed openstack/ironic-python-agent: Allow hardware managers to override clean step priority  https://review.openstack.org/18996900:15
*** penick has quit IRC00:19
openstackgerritMerged openstack/bifrost: Documentation reformatting  https://review.openstack.org/20586300:22
*** mtanino_ has joined #openstack-ironic00:25
*** mtanino has quit IRC00:26
SpamapS+ sudo sh -c 'http_proxy=http://127.0.0.1:3128/ debootstrap --verbose         --variant=minbase         --include=python,sudo,linux-image-amd64         --components=main         --arch=amd64                           unstable         /tmp/image.Ufnu3M5k/mnt         http://httpredir.debian.org/debian         '00:27
SpamapSsh: 1: debootstrap: not found00:27
SpamapSTheJulia: looks like Bifrost using ubuntu-minimal means it is missing some package deps in the env setup00:28
TheJuliathat would be the one case I didn't test earlier00:30
SpamapS:)00:30
TheJuliaseems like that is kind of a fundimental requriement though...00:30
SpamapSTheJulia: perhaps actually a diskimage-builder role could handle that.00:31
TheJuliaI thought debootstrap was burned in that way00:32
*** naohirot has joined #openstack-ironic00:34
SpamapSTheJulia: It's one of those weird things that we don't have a good way to express in things like dib upstream00:35
SpamapSI believe 'bindep' was trying to address that.00:35
* TheJulia ponders00:36
*** praneshp has quit IRC00:43
openstackgerritMerged openstack/bifrost: Documentation reformatting  https://review.openstack.org/20620100:46
*** openstackgerrit has quit IRC00:46
*** openstackgerrit has joined #openstack-ironic00:47
*** Haomeng has quit IRC00:52
openstackgerritMerged openstack/bifrost: Documentation reformatting  https://review.openstack.org/20620800:53
openstackgerritMerged openstack/bifrost: Corrected spelling  https://review.openstack.org/20621200:53
*** kevinbenton has quit IRC00:54
*** kevinbenton has joined #openstack-ironic00:57
*** puranamr has quit IRC00:57
openstackgerritJosh Gachnang proposed openstack/ironic-python-agent: Improve IPA logging and exception handling  https://review.openstack.org/19027900:59
*** Sukhdev has joined #openstack-ironic01:14
*** absubram has quit IRC01:18
*** pal has joined #openstack-ironic01:23
*** meghal has quit IRC01:24
*** Sukhdev has quit IRC01:25
*** natorious is now known as zz_natorious01:25
*** thrash is now known as thrash|g0ne01:29
openstackgerritNaohiro Tamura proposed openstack/ironic: Add hardware inspection module for iRMC driver  https://review.openstack.org/19648001:35
*** Haomeng has joined #openstack-ironic01:35
*** davideagnello has quit IRC01:36
*** rloo has quit IRC01:51
*** jlvillal has quit IRC01:58
*** jlvillal has joined #openstack-ironic01:59
*** mtanino_ has quit IRC02:06
*** eglute has quit IRC02:07
*** eglute has joined #openstack-ironic02:09
*** eglute has quit IRC02:09
*** eglute has joined #openstack-ironic02:10
*** pal has quit IRC02:16
*** pal has joined #openstack-ironic02:17
*** zhenguo has joined #openstack-ironic02:19
*** davideagnello has joined #openstack-ironic02:42
*** hakimo_ has joined #openstack-ironic02:52
*** hakimo has quit IRC02:54
*** pal has quit IRC03:03
*** ramineni has joined #openstack-ironic03:04
*** praneshp has joined #openstack-ironic03:08
*** Sukhdev has joined #openstack-ironic03:21
*** ramineni1 has joined #openstack-ironic03:21
*** zz_natorious is now known as natorious03:22
*** ramineni has quit IRC03:24
*** praneshp_ has joined #openstack-ironic03:24
*** praneshp has quit IRC03:26
*** praneshp_ is now known as praneshp03:26
*** coolsvap|away is now known as coolsvap03:32
*** Nisha has joined #openstack-ironic03:37
*** saripurigopi has joined #openstack-ironic03:40
*** davideagnello has quit IRC03:40
openstackgerritSinval Vieira Mendes Neto proposed openstack/ironic: Add port creation passing the name of the node instead of the UUID of the node  https://review.openstack.org/19343903:49
*** meghal has joined #openstack-ironic03:53
*** pal has joined #openstack-ironic03:58
openstackgerritCameron.C proposed openstack/ironic: Allow inspector driver to work in standalone mode  https://review.openstack.org/20243503:59
openstackgerritCameron.C proposed openstack/ironic: Make inspector driver test correctly  https://review.openstack.org/20497303:59
openstackgerritSinval Vieira Mendes Neto proposed openstack/ironic: Add more info level log to deploy_utils.work_on_disk() method  https://review.openstack.org/20538704:02
*** Nisha_away has joined #openstack-ironic04:02
*** Nisha has quit IRC04:03
*** sinval_ has quit IRC04:04
*** Haomeng has quit IRC04:04
*** Haomeng has joined #openstack-ironic04:04
*** Nisha has joined #openstack-ironic04:06
*** Nisha_away has quit IRC04:08
*** meghal has quit IRC04:14
*** Nisha_away has joined #openstack-ironic04:42
*** Nisha has quit IRC04:42
saripurigopimorning Ironic04:46
*** Nisha_away has quit IRC04:46
Haomengsaripurigopi: good morning:)05:00
saripurigopiHaomeng: o/05:02
Haomengsaripurigopi: :)05:02
*** rameshg87 has joined #openstack-ironic05:05
saripurigopiHaomeng: could you have a look at https://review.openstack.org/#/c/192142/ ?05:06
Haomengsaripurigopi: sure05:10
saripurigopiHaomeng: thank you05:11
Haomengsaripurigopi: yw, my pleasure, and I am not familiar with Cisco Hardware but will review and give your comments on common level:)05:11
Haomengsaripurigopi: :)05:11
saripurigopiHaomeng: sure05:12
Haomengsaripurigopi: :)05:12
*** e0ne has joined #openstack-ironic05:14
*** Sukhdev has quit IRC05:20
*** yog__ has joined #openstack-ironic05:32
openstackgerritchenglch proposed openstack/ironic-specs: Add petitboot boot driver support  https://review.openstack.org/17764105:38
*** natorious is now known as zz_natorious05:42
*** e0ne has quit IRC05:42
*** e0ne has joined #openstack-ironic05:44
*** e0ne has quit IRC05:51
*** ukalifon has joined #openstack-ironic05:52
*** jamielennox is now known as jamielennox|away05:52
saripurigopirameshg87: to write custom deploy driver05:52
saripurigopirameshg87: what are the guide lines to follow.05:53
SpamapShm not bad. 25 hosts enrolled, all but two answering IPMI05:53
*** e0ne has joined #openstack-ironic05:55
*** jamielennox|away is now known as jamielennox05:56
rameshg87saripurigopi: there are no guidelines as such05:58
rameshg87saripurigopi: if you are planning to write a new one05:58
saripurigopirameshg87: yeah, planning to write one based on vMedia and support secure boot.05:59
rameshg87saripurigopi: just check if you actually need a different boot mechanism and deploy mechanism05:59
rameshg87saripurigopi: you rather need a separate boot mechanism05:59
rameshg87https://github.com/openstack/ironic-specs/blob/master/specs/liberty/new-boot-interface.rst05:59
rameshg87saripurigopi: the first patch for it is almost done - https://review.openstack.org/#/c/166513/06:00
rameshg87saripurigopi: you can implement a cisco virtual media boot interface rather06:00
saripurigopirameshg87: yeah, let me look at this.06:03
saripurigopirameshg87: thank you06:04
saripurigopiHaomeng: thank you for the comments.06:04
Haomengsaripurigopi: welcome, just left some nits and concern, because I am not familiar with the Cisco Hardware systems:)06:04
Haomengsaripurigopi: :)06:04
saripurigopiHaomeng: sure, will submit new patch with corrections.06:05
Haomengsaripurigopi: suggest to wait more comments to cover with new patch:)06:05
Haomengsaripurigopi: :)06:05
saripurigopiHaomeng: :)06:06
SpamapShttp://paste.openstack.org/show/406007/ 23 nodes all turned on at once.. I hope the rack doesn't pop a breaker. ;)06:10
*** e0ne has quit IRC06:13
*** chlong has quit IRC06:17
rameshg87naohirot: fyi, https://review.openstack.org/#/c/205998/06:18
rameshg87naohirot: agent_irmc should be broken in the master, and should be fixed with the above change (it's applicable for both agent_ilo and agent_irmc driver)06:19
naohirotrameshg87: Hi06:20
rameshg87naohirot: hi06:20
naohirotrameshg87: let me read06:21
naohirotrameshg87: now I understood the situation06:24
naohirotrameshg87: so after Jul. 23rd both tuned not to work. right?06:25
rameshg87naohirot: yes06:25
naohirotrameshg87: how did you detect it?06:25
naohirotrameshg87: do you remember last time i asked core os configdrive error message?06:27
rameshg87naohirot: we have a nightly jenkins running06:27
rameshg87naohirot: yes06:27
naohirotrameshg87: that was my netowrk environment problem.06:27
naohirotrameshg87: after that, I didn't run agent_irmc.06:28
rameshg87naohirot: oh okay06:28
*** vishwanathj has joined #openstack-ironic06:28
naohirotrameshg87: anyway thanks for letting me know :)06:29
rameshg87naohirot: wc06:32
SpamapShm06:34
SpamapSansible-playbook -e path/to/override.yaml doesn't seem to work. Bummer.06:35
-openstackstatus- NOTICE: zuul is stuck and about to undergo an emergency restart, please be patient as job results may take a long time06:46
*** ChanServ changes topic to "zuul is stuck and about to undergo an emergency restart, please be patient as job results may take a long time"06:46
*** romainh has joined #openstack-ironic06:50
*** praneshp has quit IRC07:02
*** achanda has quit IRC07:11
*** hblixt has quit IRC07:15
*** achanda has joined #openstack-ironic07:18
SpamapSTheJulia: for when you wake up.. something is wonky in the test to see if an image exists yet. install.yaml rebuilds the deploy image every time.07:21
*** ishant__ has joined #openstack-ironic07:21
SpamapSah, I think because it's using dib_imagename as the path, but it should be {{ dib_imagename }}.qcow207:22
openstackgerritClint 'SpamapS' Byrum proposed openstack/bifrost: Fix detection of existence of image  https://review.openstack.org/20633707:24
*** achanda has quit IRC07:35
*** e0ne has joined #openstack-ironic07:37
*** dtantsur|afk is now known as dtantsur07:41
dtantsurMorning Ironic07:41
openstackgerritMerged openstack/ironic-inspector: Fix functional test invocation  https://review.openstack.org/20507007:42
*** athomas has quit IRC07:44
*** athomas has joined #openstack-ironic07:45
*** rameshg87 is now known as rameshg87-lunch07:45
*** achanda has joined #openstack-ironic07:46
*** athomas has quit IRC07:49
*** Marga_ has quit IRC07:49
*** Marga_ has joined #openstack-ironic07:50
*** coolsvap is now known as coolsvap|away07:57
* SpamapS fat fingered the list of IPs and re-deploys all 22 working nodes again... :-P07:57
SpamapStie to sleep07:57
SpamapStime to sleep rather. :-P07:57
*** Marga_ has quit IRC07:57
*** Marga_ has joined #openstack-ironic07:58
SpamapSTheJulia: I wonder if we could build in an option to turn the dynamic inventory into auth DNS and/or static DHCP records for dnsmasq.07:58
SpamapSoh also looks like bifrost's gate is broken by the debootstrap issue..07:59
*** achanda has quit IRC08:00
*** achanda has joined #openstack-ironic08:01
*** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic"08:02
-openstackstatus- NOTICE: zuul has been restarted and queues restored. It may take some time to work through the backlog.08:02
*** jistr has joined #openstack-ironic08:03
*** achanda has quit IRC08:04
*** lucasagomes has joined #openstack-ironic08:08
*** stendulker has joined #openstack-ironic08:08
*** karimb has joined #openstack-ironic08:11
*** achanda has joined #openstack-ironic08:11
*** ndipanov has quit IRC08:12
sambettsMorning all o/08:13
sambettsdtantsur o/08:13
dtantsursambetts, o/08:13
*** ramineni1 has quit IRC08:15
*** ramineni1 has joined #openstack-ironic08:20
openstackgerritMerged openstack/ironic: Add CLEANWAIT state  https://review.openstack.org/20015208:21
openstackgerritMerged openstack/ironic: Updated from global requirements  https://review.openstack.org/20607508:22
*** MattMan has quit IRC08:23
lucasagomessambetts, dtantsur morning! How it's going?08:27
*** vishwanathj has quit IRC08:27
ramineni1Haomeng, hi08:27
sambettsHey lucasagomes, I'm good thanks, you?08:28
lucasagomessambetts, all is well (-:08:28
*** vishwanathj has joined #openstack-ironic08:28
sambetts:D08:29
*** achanda has quit IRC08:32
*** alexpilotti has joined #openstack-ironic08:34
dtantsurlucasagomes, morning! sleepy and under weather08:36
dtantsurlike, well, previous 3 weeks :)08:36
*** chenke has joined #openstack-ironic08:36
*** vishwanathj has quit IRC08:39
lucasagomesdtantsur, sad to hear that :-(08:43
chenkehi08:44
sambettsman the only conversations that happen on the ML are about API versioning :-P08:44
Haomengramineni1: pong08:45
chenkegot it08:46
chenkebut where can i raise a question? sorry about that, this is my 1st time to use IRC.08:47
Haomengchenke: welcome, what question, for ironic?08:49
lucasagomessambetts, heh yeah08:49
Haomengchenke: yes, you can raise ironic question here:)08:50
lucasagomesgotta sort it out08:50
*** achanda has joined #openstack-ironic08:51
Haomenglucasagomes: good morning:)08:52
lucasagomesHaomeng, hello there! good ugt morning!08:52
chenkeHaomeng: let me summarise it and then throw it out08:52
Haomenglucasagomes: :)08:52
Haomengchenke: ok08:52
Haomengchenke: you can paste them via - paste.openstack.org08:53
*** vishwanathj has joined #openstack-ironic08:53
Haomengchenke: sorry, I will be away to go home for a while and get back to you from home:)08:54
chenkehaomeng, it's ok! we can left it here and deal with it later.08:55
*** coolsvap|away is now known as coolsvap08:56
ramineni1Haomeng, we can catch up later too :)08:57
lucasagomeschenke, just ask, someone in the channel might know and will answer you08:57
chenketku all.  you are all really helpful@@08:58
*** achanda has quit IRC09:00
chenkei have create a baremetal image by diskimage-builder. and then i deploy the image on a physical machine09:07
chenkei install some software and do a few configuration for produce environment.09:07
chenkenow i need to make image of this physical machine, so that i can update it to glance for deploying them to many other physical machine.09:07
chenkei guess my-image.vmlinuz and my-image.initrd can be reused, so do you know how to make image of physical machine that can be used by baremetal?09:07
*** vishwanathj has quit IRC09:10
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Follow-on to b6ed09e297 to fix docstrings/comments  https://review.openstack.org/20643309:10
*** boris-42 has quit IRC09:10
chenkemaybe it will be readable..   http://paste.openstack.org/show/406031/09:12
*** vishwanathj has joined #openstack-ironic09:12
*** ndipanov has joined #openstack-ironic09:12
*** Nisha_away has joined #openstack-ironic09:13
lucasagomeschenke, so you've deployed an image to a machine so you could customize it? Now you want to get this customized image upload to glance and deploy it onto N machines?09:14
sambettschenke: Is it linux?09:14
chenkeyes.  it's centos709:14
lucasagomeschenke, my suggestion would be to you to try to customize the image without deploying it first, you can do that creating elements for disk image builder. But since you've done the customization already09:15
sambettsdd if=/dev/sda of=/path/to/images/sda.img bs=102409:15
lucasagomeschenke, just boot another ramdisk or a live cd on that machine you customized09:15
lucasagomesand dd the root partition to a file09:15
lucasagomesthat will be ur image, you can reuse the kernel and ramdisk from the previous deployment09:16
lucasagomessambetts, I think he wants a partiton image, this dd will copy the full disk09:16
*** achanda has joined #openstack-ironic09:17
openstackgerritMerged openstack/ironic-inspector: Implement optional API versioning  https://review.openstack.org/20507109:20
chenkelucasagomes, i think elements is not enough, eg. some device drivers.09:21
chenkeand dd the root partition to a file?   could u detail it?09:21
lucasagomeschenke, just like sambetts did, but instead of copying the full disk you just copy the root partition09:22
lucasagomesbecause Ironic can create the rest of the partitions (swap, configdrive etc) if you need09:22
lucasagomesaccording to the nov aflavor09:22
lucasagomesdd if=/dev/sda1 of=<path to the file> bs=1024 (bs == block size)09:23
sambettshaha you beat me to it :-P09:23
openstackgerritGopi Krishna S proposed openstack/ironic-specs: Add out-of-band RAID configuration support for Cisco Hardware  https://review.openstack.org/19214209:24
chenkei think it's a sharp way. tks sambetts09:24
chenkesambetts, should i stop some service or process for avoiding some data sync isse?09:26
sambettschenke: you'll need to shut off the machine and boot it using a live disk09:27
rameshg87-lunchpersia: hi09:29
rameshg87-lunchpersia: I can confirm that proliantutils works with ​ProLiant DL380 G7. I ran the tempest test suite with that machine.09:29
rameshg87-lunchtoday :)09:29
chenkesambetts, boot it using a live disk and then run dd mentioned to copy the full sda109:30
chenkeright?09:30
lucasagomeschenke, yes09:30
*** MattMan has joined #openstack-ironic09:30
sambettschenke: yup09:30
chenkeok. i will try it.  but how can it update it to glance by a .img file?09:31
*** achanda has quit IRC09:33
lucasagomeschenke, you can just upload the new image to glance and use it to provision the new machines09:35
*** romainh has quit IRC09:36
lucasagomeschenke, http://docs.openstack.org/developer/ironic/deploy/install-guide.html#image-requirements09:36
lucasagomesthere's a section there "Add the user images to glance"09:37
openstackgerritGopi Krishna S proposed openstack/ironic-specs: Out-of-band hardware introspection support for Cisco B/C/M-series servers  https://review.openstack.org/20473309:37
*** pal has quit IRC09:38
chenkelucasagomes, yeah got it.  i will try it.  tku09:39
*** romainh has joined #openstack-ironic09:41
openstackgerritDmitry Tantsur proposed openstack/ironic: Only take exclusive lock in sync_power_state if node is updated  https://review.openstack.org/20256209:43
*** pal has joined #openstack-ironic09:43
rameshg87-lunchlucasagomes: hi there, can you please have a look at the patch https://review.openstack.org/#/c/166513/09:48
rameshg87-lunchit has 2x+2s09:49
lucasagomesrameshg87-lunch, ++09:49
*** rameshg87-lunch is now known as rameshg8709:49
lucasagomesrameshg87, will do it now09:49
rameshg87thanks09:49
openstackgerritDmitry Tantsur proposed openstack/ironic: Don't prevent updates if power transition is in progress  https://review.openstack.org/20456509:52
*** naohirot has quit IRC09:54
openstackgerritYuiko Takada proposed openstack/ironic-inspector: Migrate to oslo_log  https://review.openstack.org/20591209:56
openstackgerritShivanand Tendulker proposed openstack/ironic: grub2 bootloader support for uefi boot mode  https://review.openstack.org/16619209:58
openstackgerritShivanand Tendulker proposed openstack/ironic: grub2 bootloader support for uefi boot mode  https://review.openstack.org/16619210:00
openstackgerritShivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver  https://review.openstack.org/15480810:08
Haomengramineni1: ping10:11
ramineni1Haomeng, hi10:14
ramineni1Haomeng, I just wondering if this bug is still valid , https://bugs.launchpad.net/ironic/+bug/143435310:14
openstackLaunchpad bug 1434353 in python-ironicclient "commands occasionally show exception traceback" [Medium,Confirmed] - Assigned to Haomeng,Wang (whaom)10:14
ramineni1Haomeng, debug is set to False , we cant reproduce it right?10:15
Haomengramineni1: let me check the bug10:16
Haomengramineni1: I turn debug to false, it will work as well10:18
Haomengramineni1: so this should be invalid bug10:18
Haomengramineni1: if debug=true, the traceback will be shown out10:18
*** Nisha_away has quit IRC10:18
Haomengramineni1: if it set to false, will not print the traceback info10:19
Haomengramineni1: let me mark this bug is invalid10:19
ramineni1Haomeng, yes , I´m also wondering if it should be marked as invalid10:19
ramineni1Haomeng, ok :)10:19
Haomengramineni1: sorry for confusing10:19
Haomengramineni1: thanks for your reminder10:19
Haomengramineni1: :)10:20
ramineni1Haomeng, np..just want to confirm :)10:20
Haomengramineni1: :)10:21
lucasagomesrameshg87, +a10:27
rameshg87\o/10:28
rameshg87lucasagomes: thanks :)10:28
lucasagomestested with pxe and ipxe10:28
lucasagomesworks great!10:28
rameshg87great ..10:30
rameshg87dtantsur: wdyt about the comment on https://review.openstack.org/#/c/166521/ ?10:31
rameshg87can you please take a look at it when you get some time ..10:31
*** Marga_ has quit IRC10:34
*** yog__ has quit IRC10:35
dtantsurrameshg87, sorry, I'm about to take a small sick break, as I'm not feeling really well10:36
dtantsurrameshg87, I remember I didn't have a strong opinion on this patch10:36
dtantsurbut I'll rereview it once I'm better10:37
rameshg87dtantsur: sure..thanks a  lot10:37
*** bradjones has quit IRC10:41
*** athomas has joined #openstack-ironic10:42
*** thrash|g0ne is now known as thrash10:44
lucasagomesmrda, jlvillal when you guys have a time please take a look https://review.openstack.org/#/c/204995/10:49
lucasagomesthank you10:49
*** romcheg has quit IRC10:50
*** romainh has quit IRC11:01
*** ishant__ has quit IRC11:05
* rameshg87 goes home11:14
openstackgerritRamakrishnan G proposed openstack/ironic-specs: Update neutron integration spec  https://review.openstack.org/20648711:14
rameshg87exit11:14
*** rameshg87 has quit IRC11:14
openstackgerritMerged openstack/ironic: Allow inspector driver to work in standalone mode  https://review.openstack.org/20243511:21
openstackgerritLucas Alvares Gomes proposed openstack/ironic-specs: Lazy delete instances  https://review.openstack.org/20416211:25
lucasagomesdevananda, added the nova interactions and how they will be handled (Nova driver impact section) to the spec ^11:26
*** ramineni1 has quit IRC11:26
*** romcheg has joined #openstack-ironic11:27
*** Marga_ has joined #openstack-ironic11:27
*** pal_ has joined #openstack-ironic11:30
*** pal has quit IRC11:30
openstackgerritSinval Vieira Mendes Neto proposed openstack/ironic: Add more info level log to deploy_utils.work_on_disk() method  https://review.openstack.org/20538711:30
*** Marga_ has quit IRC11:31
*** Marga_ has joined #openstack-ironic11:32
*** stendulker has quit IRC11:33
*** coolsvap is now known as coolsvap|away11:37
*** karimb has quit IRC11:38
*** lucasagomes is now known as lucas-hungry11:38
*** romainh has joined #openstack-ironic11:41
*** karimb has joined #openstack-ironic11:42
TheJuliagood morning11:48
*** karimb has quit IRC11:48
*** Marga_ has quit IRC11:55
sinvalmorning TheJulia and Ironic11:56
*** Marga_ has joined #openstack-ironic11:58
*** karimb has joined #openstack-ironic12:09
*** lucas-hungry is now known as lucasagomes12:35
lucasagomessinval, TheJulia g'morning12:35
*** Nisha_away has joined #openstack-ironic12:36
*** karimb has quit IRC12:39
*** zhenguo has quit IRC12:42
openstackgerritNaohiro Tamura proposed openstack/ironic: Add iRMC vendor passthru for soft power off and NMI  https://review.openstack.org/18708212:45
*** coolsvap|away is now known as coolsvap12:45
lucasagomesauto merging failed for the pxe boot interface12:45
lucasagomesrame<tab>12:45
*** trown|outttypeww is now known as trown12:47
sambettsMorning TheJulia o/12:47
*** karimb has joined #openstack-ironic12:50
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Refactor pxe - New PXEBoot and ISCSIDeploy interfaces  https://review.openstack.org/16651312:50
lucasagomesrebased and fixed the auto merge ^12:51
*** karmatronic has joined #openstack-ironic12:52
openstackgerritGopi Krishna S proposed openstack/ironic: UCS: node-get-boot-device is failing for Cisco servers  https://review.openstack.org/20652812:53
*** karimb has quit IRC12:54
*** rloo has joined #openstack-ironic12:55
*** thiagop has joined #openstack-ironic12:55
thiagopGood morning, Ironic12:58
*** cdearborn has joined #openstack-ironic13:02
sambettso/ thiagop13:03
*** pal_ has quit IRC13:05
*** amotoki has joined #openstack-ironic13:08
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Clean up instance_uuid as part of the node's tear down  https://review.openstack.org/20503313:13
*** Haomeng|2 has joined #openstack-ironic13:15
*** Haomeng has quit IRC13:18
NobodyCamgood morning Ironicers... says the man waiting on coffee13:21
saripurigopiNobodyCam: o/13:21
TheJuliaGood morning NobodyCam13:21
NobodyCamsaripurigopi: TheJulia: good morning :)13:22
lucasagomesNobodyCam, good morning13:25
NobodyCamgood (UGT) morning lucasagomes :)13:25
*** kkoski has joined #openstack-ironic13:27
sambettso/ NobodyCam13:27
jrollmorning all :)13:28
dtantsurjroll, NobodyCam, TheJulia, morning13:28
* dtantsur got fresh air, wine and espresso, and seemingly recovered13:28
lucasagomesjroll, morning!13:29
jrolldtantsur: lol13:29
jrollmorning lucasagomes dtantsur TheJulia NobodyCam sambetts and everyone else \o13:29
jrolllucasagomes: looking at https://review.openstack.org/#/c/204995/3/nova/virt/ironic/driver.py13:29
lucasagomesjroll, yup13:30
lucasagomesthanks13:30
jrollI'm wondering why we do node = validate_...13:30
jrollthe node (uuid, I assume) is passed to _unprovision13:30
jrollso why not just check the state, and drop the get by instance call13:30
jrollyeah, we do the same call in destroy to get the node object13:31
lucasagomesjroll, I think it's to avoid some race13:31
lucasagomeslike the node changes the instance_uuid and now is provisioning another instance13:32
lucasagomesI don't think that may happen tho... 1 sec13:32
lucasagomesI think deva commented about it in another patch13:32
jrolllucasagomes: your new code would hit the same race though13:32
thiagopmorning NobodyCam sambetts jroll13:32
jrolloh no it wouldn't13:32
jrollurgh13:32
lucasagomesjroll, https://review.openstack.org/#/c/183107/13:32
lucasagomesno it wouldn't13:32
lucasagomesI would just exit if that instance was already deleted13:32
jrollright13:33
jrollok13:33
jrollthanks13:33
jroll+113:33
NobodyCammornign jroll sambetts dtantsur :)13:33
lucasagomesjroll, I want to remove that loop as part of the lazy instance deletion13:33
jrolllucasagomes: indeed, though I think we'll still need it in case of api blip or whatever13:34
jrollthough that's probably a client wrapper thing13:34
*** Nisha has joined #openstack-ironic13:35
lucasagomesjroll, yeah tear down request still goes to the conductor13:36
lucasagomesand in the conductor we start the thread _spawn_worker to do the rest making in async13:36
jrolldtantsur: let's do the whole "ironic has-capability" thing and have the return values like this: https://imgur.com/ZBpai0413:36
lucasagomesfor marking the node to be deleted we jsut will mark the node in the db and return, so it's basically a sync operation13:36
*** Nisha_away has quit IRC13:36
jrolllucasagomes: right, yep13:36
lucasagomesso if the command succeed it's reliable that the node is marked for deletion13:37
dtantsurjroll, lol, exactly13:37
*** saripurigopi has quit IRC13:42
rloohi lucasagomes, dtantsur, sambetts. morning NobodyCam, jroll, thiagop, TheJulia, and everyone else :)13:43
jrollmorning rloo :)13:43
lucasagomesrloo, hi there! good morning13:44
dtantsurrloo, morning13:44
thiagophey, morning rloo13:45
NobodyCamgood mornign rloo thiagop :)13:46
*** Nisha has quit IRC13:51
*** cdearborn has quit IRC13:52
*** karmatronic has quit IRC13:54
lucasagomesjroll, lol damn I thought it was a joke! But no, it's actually for real13:55
lucasagomeshttp://www.w3schools.com/tags/av_met_canplaytype.asp13:55
lucasagomesshite13:55
jrollheh, yep13:56
dtantsuromg14:01
openstackgerritYuriy Zveryanskyy proposed openstack/ironic: Cache Keystone client instance  https://review.openstack.org/18668514:01
jrollthis is why I don't do frontend work anymore :)14:04
lucasagomeswould be awesome to submit a patch adding a new return value "hopefully" - "if you have enough faith the browser might play that audio/video."14:09
*** boris-42 has joined #openstack-ironic14:10
*** pal has joined #openstack-ironic14:13
*** pal has quit IRC14:19
*** pal has joined #openstack-ironic14:19
*** thrash has quit IRC14:21
*** mestery has joined #openstack-ironic14:24
mesteryJoshNang: Did you put something on the neutron agenda from the nova mid-cycle last week?14:24
mesteryFor reference: https://review.openstack.org/#/c/152703/14:24
mesteryJoshNang: nm14:24
*** thrash has joined #openstack-ironic14:27
openstackgerritJarrod Johnson proposed stackforge/pyghmi: Implement read of IPv4 network configuration  https://review.openstack.org/20655714:28
*** thrash has quit IRC14:31
*** thrash has joined #openstack-ironic14:35
*** Nisha has joined #openstack-ironic14:38
openstackgerritLaura Moore proposed openstack/ironic: Addition of portgroups to support LAG interfaces  https://review.openstack.org/20623214:42
*** cdearborn has joined #openstack-ironic14:43
*** meghal has joined #openstack-ironic14:46
*** zz_natorious is now known as natorious14:52
openstackgerritLaura Moore proposed openstack/ironic: Addition of portgroups to support LAG interfaces  https://review.openstack.org/20623214:52
openstackgerritMerged openstack/ironic: Refactor pxe - New PXEBoot and ISCSIDeploy interfaces  https://review.openstack.org/16651314:55
dtantsur\o/   ^^14:55
NobodyCamnice!14:55
jroll\o/14:56
dtantsur*now* we're ready to release :)14:56
*** achanda has joined #openstack-ironic14:56
thiagop\o/14:56
openstackgerritLaura Moore proposed openstack/ironic: Addition of portgroups to support LAG interfaces  https://review.openstack.org/20623814:57
jrolldevananda: let's release pls14:59
jrollseriously, it's L-2 everywhere else14:59
jrollwe wanted to move faster14:59
jrolland we still haven't released since kilo14:59
jrollsuch sad :(15:00
rloojroll: what is needed wrt deciding stuff, before a release happens? there was something about semver15:00
openstackgerritMerged openstack/bifrost: Documentation reformatting  https://review.openstack.org/20621715:00
rloojroll: and we don't want to release cuz of enroll stuff?15:00
jrollrloo: yeah, we need to fix the version thing15:01
jrollfor the second bit, define "we"15:01
rloojroll: so let me ask another question. how do "we" (ironic group) decide when to do a release :-)15:02
jrollrloo: we say "we've built some cool stuff, let's get it to users" :)15:02
rloojroll: how many people have to say that? It seems like there is at least 2 cores that say let's release...15:02
*** pal has quit IRC15:03
jrollrloo: well, I'm not sure. but right now, only devananda can actually do the logistics of the release15:03
devanandaexactly one person has to say that - the release manager for the project, which currently is me15:03
rloojroll: ha ha. ok, so devananda has the power. that's what I wanted to know.15:03
lucasagomes:-) boot interface!15:04
sambettso/ rloo15:04
devanandajroll: have folks done any extra testing / documenting of the current state of the code, checking on open bugs to make sure it, well, doesn't contain known bugs that we would usually consider a release blocker?15:07
jrolldevananda: not sure15:08
jrollso let's back up15:08
jrollwhat do we need to do to ship some code15:09
devanandajroll: I'd love to know what to comunicate to users in the release notes15:09
devanandaheh15:09
devanandaseveral things :)15:09
jrollright15:09
jrollis there a documented checklist?15:09
devanandaso - let's actually talk through that in an hour (after this meeting)15:09
*** achanda has quit IRC15:09
dtantsurright, these things should be done during this "week of soft freeze" that we agreed on. so before releasing anything we should at least officially start it :)15:09
devanandafrankly, I do not want to write relesae notes that say "oh by the way, we just made a breaking API change in our first release"15:09
devanandadtantsur: right!15:10
jrollsure15:10
dtantsurdevananda, our previous release also had a breaking change, it's kind of a tradition already :D15:10
* dtantsur is kidding of course15:10
jrollit isn't a breaking change15:10
* jroll is not kidding15:11
lucasagomesyeah we have to work it around with the client still, are we releasing the client right after the release of ironic right?15:11
*** natorious is now known as zz_natorious15:12
*** ukalifon has quit IRC15:12
jrolllet's keep client chatter to the ML15:12
jrollbecause that's an *entirely* separate conversation15:12
rloowhether we release now or (no choice) do a Liberty release, it seems like the enroll/client/microversion is still a question mark?15:12
dtantsurI don't think we question ENROLL, it's a bit too late15:13
openstackgerritRicardo Carrillo Cruz proposed openstack/bifrost: Fix roles and playbooks not putting deploy image on /httpboot  https://review.openstack.org/20658515:13
rlooI honestly don't understand the microversioning stuff. I thought we had decided how it was going to work.15:13
jrollright, just how the client handles versions is in the air15:13
*** rcarrillocruz has joined #openstack-ironic15:13
rloobut jroll's email seems to indicate that we're opening it all up again?15:13
jrollyes, because how we're doing it today is fundamentally broken15:13
rcarrillocruzTheJulia: hiya, just pushed https://review.openstack.org/206585 for the issue of deploy images not being placed under /httpboot15:14
jrollit attempts to completely hide the fact that versioning is a thing from users15:14
*** Nisha has quit IRC15:14
rloojroll: we have a spec on that: http://specs.openstack.org/openstack/ironic-specs/specs/kilo/api-microversions.html15:14
rloojroll: should we be addressing that spec then?15:15
lucasagomesrloo, re microversion, ignoring the bits how it currently gets negotiated by the client <-> server. I like sean's view of it15:15
openstackgerritMerged openstack/ironic: Don't prevent updates if power transition is in progress  https://review.openstack.org/20456515:15
lucasagomes(not sure if you have been following the emails about it)15:15
dtantsurlucasagomes, I also like pretending we don't have versioning, but it's dishonest towards non-CLI users15:15
rloolucasagomes: i have sort of been following the emails but i get a bit lost with what we're trying to do.15:15
lucasagomesdtantsur, no I don't like that15:15
lucasagomesIMO I would prefer the version to be explicit15:16
jrollrloo: sure. I brought it up in irc, was told to take it to the meeting. I brought it up in the meeting, was told to take it to the mailing list. I brought it up in the mailing list, and here we are15:16
jrollrloo: so we can take it to the spec if you'd like, but...15:16
dtantsurLOL15:17
dtantsurjroll++15:17
rloojroll: ok, so basically, we want to start from scratch with 1. we will have microversioning in ironic. 2. default microversion in server is the lowest supported version 3. the question is how our client / server interacts?15:17
dtantsuractually we have an API-WG spec as well. It got a load of -1s and is silent now :)15:17
dtantsurrloo, 4. what does versioning means for us?15:18
jrollrloo: correct, and in 3 the python client and the CLI are two separate itmes15:18
jrolland to be clear, the CLI is for running commands, it is NOT a bash SDK.15:18
dtantsurrloo, 5. depending on 4, what's the scope of versioning?15:18
*** karimb has joined #openstack-ironic15:18
*** mtanino has joined #openstack-ironic15:18
rlooso I think dtantsur is correct. it is clear that it is unclear to folks what the scope of versioning is. (and/or the scope seems to fluctuate, dunno.)15:20
*** vishwanathj has quit IRC15:20
jrollso I think it's pretty clear. if we change the API semantics, we change the version. people disagree with that and it creates confusion.15:20
*** mestery has quit IRC15:20
rloodtantsur: you OK with jroll's description?15:21
jroll"if we change the API semantics, we change the version." is what our rule is today, at least15:21
lucasagomesjroll, ++15:21
lucasagomesI like what sean's said here http://lists.openstack.org/pipermail/openstack-dev/2015-June/065849.html15:21
lucasagomesabout that15:22
jroll+115:22
rloothere's more details in the nova spec i think, about what might bump the microversion.15:22
lucasagomesmaking each version unique and equal across clouds15:22
devanandajroll: "if we change ... then we change ..." < +115:23
dtantsurrloo, no, it's not complete :)15:23
dtantsurjroll, 1. define "API semantics", 2. your definitions says nothing about feature hiding, which is the source of our problems15:23
jrolldtantsur: adding a feature is changing the API semantics15:24
dtantsur(and which only 1 person ever tried to explain to me of the whole community)15:24
devanandaalso what sdague says about "the fallacy of time only flowing forward" is important to consider, and this is why guarding non-breaking features is important.15:24
rloothen there's devananda patch (I admit, I didn't read the versioning part): https://review.openstack.org/#/c/196320/15:24
devanandanow, before dtantsur says we disagree on that ...15:24
devanandalet me point at my patch for "compatible"15:24
devanandabecause indicating the point in time that the last semantic change happened != the last time that any change happened15:24
jrolldtantsur: API semantics == "how the API works and is interacted with"15:24
dtantsurjroll, driver_info and other JSON fields? node states?15:25
devanandawe didn't define that in either our spec, or nova's spec, but it has come up in several other discussions since then15:25
dtantsurdevananda, you still didn't start with *WHY* in your spec15:25
*** meghal has quit IRC15:25
jrolldtantsur: so for example, the semantics of driver_info is "it is a json field that may be updated"15:25
jrolldtantsur: ++15:25
dtantsurand without *WHY* we can create whatever *HOW*15:25
jrollI don't understand the usefulness of the compaible header15:26
jrollcompatible.15:26
devanandadtantsur: hm. I thought I had. lemme see15:26
lucasagomesdevananda, the header is merely informative? Or it will be used on the negotiation of the version ?15:26
jrollI really wish this conversation was happening in the mailing list thread I started15:26
jrolljust saying.15:26
dtantsurdevananda, I'm speaking about broad *WHY*. What we achieve by versioning, what we achieve by feature hiding, what we achieve by compatible header, examples of use cases15:26
rloojroll: I'm hoping that we can summarize this discussion in the ML if we end up agreeing on anything.15:27
dtantsurjroll, I said the same words in 2-3 threads already, not sure I want to be ignored for the 4th time...15:27
lucasagomes++15:27
lucasagomesrloo, ^++15:27
jrollrloo: we've touched about 10 topics here in IRC this morning, best I can tell, but I hope so too15:27
jrolldtantsur: I understand, it seems it's the same people on the list again and again15:27
devanandadtantsur: oh, gotcha. yea, that wasn't in the initial spec either15:27
devanandajroll: ++ to ML. I'll go reply there after this meeting15:28
dtantsurdevananda, maybe if I read all the use cases, I would just stop arguing :)15:28
devanandadtantsur: lol15:28
TheJuliarcarrillocruz: thank you!15:28
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Clean up instance_uuid as part of the node's tear down  https://review.openstack.org/20503315:29
dtantsurdevananda, on an unrelated note: may I ask you to rereview https://review.openstack.org/#/c/202562/ please? A bit tricky change IMO15:31
devanandadtantsur: ack15:31
dtantsurjroll, I've probably asked, but may I ask again: what log level to you folks run your services at?15:33
*** mestery has joined #openstack-ironic15:34
jrolldtantsur: debug all the things15:34
dtantsurjroll, then could you +1 this thingy for me please: https://review.openstack.org/#/c/206437/15:34
jrollthere's no way to actually track down problems without it15:34
jrollheh15:34
lucasagomesjroll, yeah we really should improve the logs15:35
lucasagomes:-/15:35
lucasagomesspecially INFO stuff15:35
jrolllucasagomes: it doesn't matter, when something breaks you're going to want debug15:35
jrollso unless we s/debug/info/ everywhere I'll always run at debug15:35
jrollunless we just never have any failures15:36
jroll(lol)15:36
lucasagomeswe are close :-) #notnot15:36
lucasagomes#butnot15:36
dtantsurjroll, yep, but grepping over INFO should give one a complete picture of state transitions, so it still makes sense to improve15:36
jrolldtantsur: sure. I still conjecture INFO is useless, but hey :)15:37
dtantsurFWIW we run all undercloud services as DEBUG too15:37
jrollnod15:38
jrollrackspace runs all of our public cloud at debug15:38
jrollaggregated nova-compute logs are so massive it's hilarious15:38
lucasagomesdamn yeah I can imagine15:38
dtantsurnow to something completely different:15:41
dtantsurThe Retry-After response-header field can be used with a 503 (Service Unavailable) response to indicate how long the service is expected to be unavailable to the requesting client.15:41
dtantsurthe spec does not expect to return this header with 409 response..15:42
*** ijw has joined #openstack-ironic15:42
*** alexpilotti has quit IRC15:47
openstackgerritRuby Loo proposed openstack/python-ironicclient: Node names support RFC 3986 unreserved characters  https://review.openstack.org/20659615:47
*** thrash has quit IRC15:47
rloodtantsur: maybe we shouldn't be returning 409 for those cases where you want a retry-after?15:48
openstackgerritRicardo Carrillo Cruz proposed openstack/bifrost: Fix roles and playbooks not putting deploy image on /httpboot  https://review.openstack.org/20658515:49
dtantsurrloo, changing error code is a breaking change, though versioning might help15:49
dtantsurrloo, well, various node locked errors are not user's fault, so they probably should not be 4xx15:49
*** meghal has joined #openstack-ironic15:49
rloodtantsur: seems like we discussed that at an irc meeting, and yeah, they shouldn't be 4xx15:50
devanandadtantsur: the problem with 503 is, AIUI, it means the whole service is unavailable15:50
devanandadtantsur: unless I'm missing a context and that *is* what you're intending15:51
devanandaalso - done with meeting. reviewing that patch now15:51
dtantsuryeah, IIRC we didn't invent a good code for that15:51
*** thrash has joined #openstack-ironic15:52
lucasagomesdevananda, also, if you have some time I've updated https://review.openstack.org/#/c/204162/ to explain the nova interactions and how it's going to work (driver nova impact session)15:52
*** thrash has quit IRC15:52
*** thrash has joined #openstack-ironic15:54
*** logan2 has quit IRC15:55
*** cdearborn has quit IRC15:57
*** cdearborn has joined #openstack-ironic15:58
jrolldtantsur: devananda lucasagomes rloo so, I don't see any clear outcomes of that discussion besides "let's talk on the ML", is that accurate?15:59
rloojroll: yeah. devananda said he'd post on the ML.15:59
dtantsuryes :(15:59
lucasagomesseems correct yes15:59
jrollok, I will anxiously await some activity15:59
*** [1]cdearborn has joined #openstack-ironic16:01
rloojroll: I'm not quite sure if this is a pop-corn moment, but maybe you should have a drink or two ;)16:01
jrollrloo: seems like two might not be enough16:01
lucasagomesjroll, btw on ur ML, any reason why you prefer 1. and 3.  best? I thought you liked the idea of having the version mandatory16:01
*** praneshp has joined #openstack-ironic16:02
jrolllucasagomes: mostly because I thought about it more and the deprecation thing will be painful16:03
jrollbut I'm also fine with 416:03
*** kkoski has quit IRC16:03
rloojroll, lucasagomes: I believe I argued for 1 but I was outvoted. many moons ago.16:03
jroll(mandatory version)16:03
jrollrloo: it wouldn't be any fun if we never changed our minds, now, would it :)16:03
lucasagomesyeah, the problem I see with 1. is that it will be another breakable change16:03
jrollso would 2 and 3 and 416:04
rloojroll: what's your definition of 'fun' ? :)16:04
lucasagomesyes, but 4. in the long run educates people about version16:04
jrollright16:04
lucasagomesthe others still hide it and gambles16:04
jrollI mean, they all do in one way or another16:04
jrollwhether it's "I need to specify version to get this feature" or "my random cli command broke because I didn't use a version"16:05
*** kkoski has joined #openstack-ironic16:05
lucasagomessure, maybe I just prefer the explicit better than implicit16:05
jrollyeah, +116:05
rlooi think it would help if we had some requirements/criteria. eg, 1. must not break users (hi priority); 2. must educate people about versioning (medium priority), 3. must make dmitry happy (critical), ...16:05
*** coolsvap is now known as coolsvap|dinner16:06
jrolllucasagomes: I think the real problem here is that people think the CLI is a bash SDK16:06
*** penick has joined #openstack-ironic16:06
lucasagomesjroll, right, just like Clynt have said in the email16:07
dtantsurrloo, I like you prioritization, thanks :)16:07
lucasagomesand the fix for that "hey pin a version"16:07
lucasagomestherefore I think that making version mandatory so people won't break they scripts when updating the API is important16:07
jrollsure16:09
openstackgerritMerged openstack/ironic: Remove outdated TODO.rst file  https://review.openstack.org/20618216:09
lucasagomess/important/ideal16:09
*** davideagnello has joined #openstack-ironic16:14
*** romcheg has quit IRC16:15
*** cdearborn has quit IRC16:15
*** meghal has quit IRC16:17
*** guy_rodrigue has joined #openstack-ironic16:19
*** e0ne has quit IRC16:20
*** praneshp_ has joined #openstack-ironic16:21
*** praneshp has quit IRC16:21
*** praneshp_ is now known as praneshp16:21
*** jistr has quit IRC16:22
*** romcheg has joined #openstack-ironic16:23
*** saripurigopi has joined #openstack-ironic16:23
*** logan2 has joined #openstack-ironic16:24
*** dtantsur is now known as dtantsur|afk16:27
*** karimb has quit IRC16:28
*** romcheg has quit IRC16:32
*** meghal has joined #openstack-ironic16:32
NobodyCamhave a good night dtantsur|afk16:33
*** Nisha has joined #openstack-ironic16:35
*** pal has joined #openstack-ironic16:36
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Add delete_instance field to the nodes table  https://review.openstack.org/20661616:37
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Add support for deferring instance deletion to later  https://review.openstack.org/20661716:37
lucasagomescode for the lazy delete instance ^16:38
lucasagomesspec is not merged so no bother reviewing, unless you want a reference for the spec16:39
*** lintan has quit IRC16:39
openstackgerritJosh Gachnang proposed openstack/ironic-python-agent: Improve IPA logging and exception handling  https://review.openstack.org/19027916:39
*** lintan has joined #openstack-ironic16:41
*** pal_ has joined #openstack-ironic16:42
*** pal has quit IRC16:43
openstackgerritMerged openstack/ironic: Make inspector driver test correctly  https://review.openstack.org/20497316:46
devanandadtantsur|afk: I'm looking at how this uses upgrade_lock() and wondering if that method should use a try around _lock() to call release_resources() if _lock() fails16:46
*** puranamr has joined #openstack-ironic16:54
devanandadtantsur|afk: ah - perhaps not. the exception is caught outside the with: block, so that should implicitly release() things. :)16:55
lucasagomesdevananda, I think that if we do that the task object will be invalidate if upgrade_lock() fails16:56
lucasagomesI think the idea is that the shared lock is valid in case the exclusive one failed16:56
devanandalucasagomes: it doesn't call release internally16:56
devanandathe way that acquire or __init__ will16:57
lucasagomesdevananda, yes, I believe that's to avoid ^16:57
devanandaright16:57
lucasagomesunless you wanna to start with a exclusive lock, then it should call realease if it fails16:57
devanandano, you're correct. I had to re-read the outer loop in ConductorManager to see where the release is called from, that's all16:58
lucasagomesyeah, tho I think some methods that patch calls we should enforce the exclusive lock with @task_manager.require_exclusive_lock16:59
lucasagomesbecause they change the states of the node16:59
* lucasagomes comments on the patch16:59
*** coolsvap|dinner is now known as coolsvap|mt17:01
*** coolsvap|mt is now known as coolsvap|mtg17:01
sambettsnight all o/17:03
NobodyCamnight sambetts17:04
lucasagomessambetts, good night17:04
*** guy_rodrigue has quit IRC17:04
*** athomas has quit IRC17:05
*** praneshp has quit IRC17:10
*** thrash has quit IRC17:12
NobodyCamTheJulia: if you have a free cycle or two: https://review.openstack.org/#/c/17860717:12
NobodyCam^^^ tempest test ironic without glance17:12
openstackgerritRuby Loo proposed openstack/ironic: Save and re-raise exception  https://review.openstack.org/20662917:13
NobodyCambrb quick post bbt walkies17:15
*** thrash has joined #openstack-ironic17:15
*** pal_ has quit IRC17:15
*** thrash has quit IRC17:15
*** davideagnello has quit IRC17:16
*** davideagnello has joined #openstack-ironic17:16
*** thrash has joined #openstack-ironic17:16
*** lintan has quit IRC17:19
*** lintan has joined #openstack-ironic17:19
*** boris-42 has quit IRC17:20
*** mestery has quit IRC17:20
*** alex_xu has quit IRC17:21
*** boris-42 has joined #openstack-ironic17:22
*** ifarkas has quit IRC17:22
openstackgerritGopi Krishna S proposed openstack/ironic-specs: Out-of-band hardware introspection support for Cisco B/C/M-series servers  https://review.openstack.org/20473317:23
*** alex_xu has joined #openstack-ironic17:23
*** thrash has quit IRC17:24
*** meghal has quit IRC17:27
*** romainh has quit IRC17:31
*** romainh has joined #openstack-ironic17:33
*** thrash has joined #openstack-ironic17:33
*** jjohnson2 has joined #openstack-ironic17:39
*** romainh has left #openstack-ironic17:40
*** achanda has joined #openstack-ironic17:42
*** jjohnson2 has quit IRC17:45
*** lintan_ has joined #openstack-ironic17:46
lucasagomesfolks I will call it a day17:46
lucasagomeshave a great night!17:47
lucasagomessee y'all17:47
*** lucasagomes is now known as lucas-dinner17:47
*** lintan has quit IRC17:48
*** amotoki has quit IRC17:48
NobodyCamnight lucas-dinner17:49
thiagopnight lucas-dinner17:49
*** meghal has joined #openstack-ironic17:50
*** mestery has joined #openstack-ironic17:54
*** jjohnson2 has joined #openstack-ironic17:55
*** mestery has quit IRC17:55
openstackgerritJosh Gachnang proposed openstack/ironic-specs: Implement Zapping States  https://review.openstack.org/18512218:00
*** achanda has quit IRC18:02
thiagopGuys, I'm trying to generate a "ubuntu ironic-agent" deploy image with DIB and it's failing with a "No such file or directory" on /usr/lib/systemd/system/ironic-python-agent.service18:02
*** achanda has joined #openstack-ironic18:02
thiagopanybody has a clue of what is happening?18:02
NobodyCamthiagop: when are you getting that18:04
*** alex_xu has quit IRC18:04
*** penick has quit IRC18:05
thiagopNobodyCam: appears to be in the cleaning after the IPA installation18:07
jrollthe ubuntu support just landed, it could be broken18:07
thiagopdefault is fedora jroll ?18:08
NobodyCamit was18:08
*** romainh has joined #openstack-ironic18:08
jrollthiagop: yes, until recently only option was fedora18:08
jrollseems like ubuntu element shouldn't be using systemd18:09
thiagopmodifying the question: the "mainstream" is fedora? :)18:09
*** alex_xu has joined #openstack-ironic18:09
thiagoplet me try...18:09
jrollyeah :)18:09
jrollwell I'd argue "mainstream" isn't DIB at all, but... people like dib for some reason18:09
thiagopI would vote for it due to easiness18:11
*** davideag_ has joined #openstack-ironic18:11
jrollthiagop: the alternative is "cd ironic-python-agent/imagebuild/coreos/ && make"18:12
*** davideagnello has quit IRC18:12
*** meghal has quit IRC18:12
thiagopI didn't know that...18:13
thiagopknew*18:13
*** davideagnello has joined #openstack-ironic18:13
NobodyCamTheJulia: happen to still be around .. or are you at lunch18:14
jrollthiagop: even has docs! https://github.com/openstack/ironic-python-agent/tree/master/imagebuild/coreos18:14
thiagopjroll: I'm beginning to walk through this land now. Thanks.18:15
jroll:)18:15
*** davideag_ has quit IRC18:16
thiagopcorrect me if I'm wrong, but I can run *any* user image with the deploy ramdisk generated with fedora?18:17
*** davideagnello has quit IRC18:17
*** davideagnello has joined #openstack-ironic18:17
NobodyCamthiagop: run == deploy?18:17
TheJuliaNobodyCam: around but on phone being grumpy home seller18:17
*** meghal has joined #openstack-ironic18:17
thiagopyep18:17
jrollthiagop: yep18:17
thiagopnice18:17
NobodyCamTheJulia: ack18:17
TheJuliaNobodyCam: whats up18:18
TheJuliahmmmm18:18
TheJuliai bet the folder doesnt exist....18:19
*** penick has joined #openstack-ironic18:19
NobodyCamI'm attempting to create a enrollment file within the vagrant yaml file and was woundering what format is best.18:20
NobodyCamcsv I think would be easiest but There was a issue converting csv (converted but exited with non zero value)18:20
*** vishwanathj has joined #openstack-ironic18:21
NobodyCamis htat still true of am I out in left field?18:22
*** davideagnello has quit IRC18:22
jrollyou being off in left field is tangential to this :)18:23
NobodyCamhehehhe18:23
*** davideagnello has joined #openstack-ironic18:23
TheJuliaNobodyCam: yaml or json would be most flexible in describing the other host.  i thought ehat you ran into was just stderr, stdout was fine, but you didnt indicate an non-zero exit18:23
TheJulias/field/hemisphere/?18:23
NobodyCamI'll dbl check... ICBW ofc 4realz18:24
*** vishwana_ has joined #openstack-ironic18:25
*** vishwanathj has quit IRC18:25
rloothis is awesome (sort of). We approved this patch before Jenkins had a chance to vote. The gate passed and code got merged, but the original jenkins job failed! https://review.openstack.org/#/c/166513/18:25
rloothere are advantages to approve w/o waiting for jenkins then ;)18:25
jrollrloo: the ultimate recheck18:25
*** vishwana_ has quit IRC18:27
*** Marga_ has quit IRC18:33
*** saripurigopi has quit IRC18:37
*** Marga_ has joined #openstack-ironic18:37
*** pelix has quit IRC18:43
*** puranamr has quit IRC18:43
*** Haomeng has joined #openstack-ironic18:43
*** Haomeng|2 has quit IRC18:46
*** puranamr has joined #openstack-ironic18:47
harlowjarloo u want to plugin the automaton converter into https://github.com/openstack/ironic/blob/master/tools/states_to_dot.py :-P18:59
harlowjau know u want to ;)18:59
harlowjahttp://docs.openstack.org/developer/automaton/api.html#module-automaton.converters.pydot (that thing)18:59
rlooharlowja: oh, I thought we had already changed that.18:59
harlowjanah, haven't yet afaik18:59
rlooharlowja: thx for the reminder.19:00
harlowjaif u want a easy code change :-P19:00
harlowjaif not, let me know and i'll try to get around to it someday, lol19:00
rlooharlowja: it'd be easier if you did the code change and put my name on it :)19:00
* harlowja the state diagram might need to be refreshed again anyway19:00
harlowjarloo haha19:00
rlooharlowja: I'll do it this afternoon. unless I get distracted ;)19:00
harlowja:)19:01
harlowjano distractions!19:01
rlooharlowja: oh yeah, i think the diagram needs to be updated too.19:01
harlowjarloo likely19:01
harlowjathe monthly diagram refresh, lol19:01
*** e0ne has joined #openstack-ironic19:03
*** ijw has quit IRC19:05
*** zz_natorious is now known as natorious19:05
*** ijw has joined #openstack-ironic19:08
*** natorious is now known as zz_natorious19:10
*** ijw has quit IRC19:12
openstackgerritJarrod Johnson proposed stackforge/pyghmi: Implement read of IPv4 network configuration  https://review.openstack.org/20655719:15
*** ijw has joined #openstack-ironic19:17
*** ijw has quit IRC19:21
*** mestery has joined #openstack-ironic19:30
openstackgerritJulia Kreger proposed openstack/bifrost: Allow use of extra_dib_elements  https://review.openstack.org/20666419:30
openstackgerritJulia Kreger proposed openstack/bifrost: Remove un-necessary role  https://review.openstack.org/20666719:35
openstackgerritMerged openstack/ironic: Follow-on to b6ed09e297 to fix docstrings/comments  https://review.openstack.org/20643319:36
NobodyCam22219:39
TheJulia222?19:41
jrolltimes 319:44
NobodyCamlol19:44
NobodyCamwas a wrong window19:44
NobodyCam:-p19:44
openstackgerritJulia Kreger proposed openstack/bifrost: Error if deploy_image is not found.  https://review.openstack.org/20668219:49
NobodyCamTheJulia: lol doh19:49
*** ijw has joined #openstack-ironic19:51
*** ijw has quit IRC19:56
*** alexpilotti has joined #openstack-ironic20:06
*** mestery has quit IRC20:08
*** alexpilotti has quit IRC20:12
openstackgerritNisha Agarwal proposed openstack/ironic: Remove the hard dependency of swift from ilo drivers  https://review.openstack.org/19865620:14
*** ijw has joined #openstack-ironic20:14
*** igordcard has joined #openstack-ironic20:16
*** e0ne has quit IRC20:19
*** e0ne has joined #openstack-ironic20:21
*** e0ne has quit IRC20:21
*** mtanino has quit IRC20:23
*** puranamr has quit IRC20:36
*** penick has quit IRC20:38
*** mtanino has joined #openstack-ironic20:41
*** logan2 has quit IRC20:42
*** alexpilotti has joined #openstack-ironic20:42
*** mestery has joined #openstack-ironic20:43
*** puranamr has joined #openstack-ironic20:44
*** logan2 has joined #openstack-ironic20:45
*** alexpilotti has quit IRC20:52
*** thrash is now known as thrash|g0ne21:03
*** kkoski has quit IRC21:03
openstackgerritRuby Loo proposed openstack/ironic: enroll/verify/cleanwait in state machine diagram  https://review.openstack.org/20669821:08
*** jjohnson2 has quit IRC21:10
mrdaMorning Ironic21:14
NobodyCamgood mornign mrda21:15
mrdahey NobodyCam21:15
NobodyCam:)21:16
*** penick has joined #openstack-ironic21:17
*** lucas-dinner has quit IRC21:19
NobodyCamrunning out to a doctors apt...21:25
NobodyCambbiaw21:25
*** mestery has quit IRC21:25
*** meghal has quit IRC21:27
*** meghal has joined #openstack-ironic21:27
*** meghal has quit IRC21:28
*** meghal has joined #openstack-ironic21:29
*** vishwanathj has joined #openstack-ironic21:31
*** meghal has quit IRC21:33
*** lucas-dinner has joined #openstack-ironic21:33
*** meghal has joined #openstack-ironic21:33
openstackgerritRuby Loo proposed openstack/ironic: Use automaton's converters/pydot  https://review.openstack.org/20672621:34
*** vishwanathj has quit IRC21:36
*** Nisha has quit IRC21:42
*** vishwanathj has joined #openstack-ironic21:52
*** vishwanathj has quit IRC22:00
*** Pradip has joined #openstack-ironic22:00
*** coolsvap|mtg is now known as coolsvap|away22:00
*** vishwanathj has joined #openstack-ironic22:04
mrdaHas anyone used IPMI/PXE drivers on Dell kit?  Or has anyone who has been using Ironic on Dell been using the DRAC driver only?22:06
*** vishwanathj has quit IRC22:09
jrollmrda: lucas-dinner or dtantsur|afk might have insight22:09
jrollI would assume it just works if ipmi is enabled22:09
mrdathanks jroll - I did some basic poking of a Dell 720 with IPMI once upon a time, but I'd like to know whether its usable and reliable in practice for Ironic.  I'll bug those guys in 12 hours or so.22:11
mrdaTimezones are hard.22:11
jrollmhmmm22:11
Pradiphow can I have ironic devstack installed. I can boot up BM perfectly. But the problem is now I can't boot up normal VM's. I know that I need to run another compute node for normal vms, but i am not sure how to integrate them. Please help22:15
*** kkoski has joined #openstack-ironic22:15
* mrda thanks jroll for his help22:17
*** kkoski has quit IRC22:21
openstackgerritStephanie Miller proposed openstack/bifrost: Correct reference to deploy_image variable in install playbook  https://review.openstack.org/20674522:23
jrollPradip: as I understand it, you should use two compute nodes, one for VM one for metal, and host_aggregates to define what does what22:27
jrollbeyond that I'm not sure how it works22:27
*** zz_natorious is now known as natorious22:30
mrdajust scheduling filters?22:30
Pradipjroll: okay. I am confused bcoz in only VM case, there is one controller and multiple compute node. and the compute node doesn't run all the components of openstack. But here if there is a contol node, onevm compute and one metal compute node. what should be the localrc of the controller and the BM compute22:31
jrollPradip: I don't think devstack supports this setup by just editing localrc, but I am not sure22:31
*** [1]cdearborn has quit IRC22:31
Pradipjroll: Okay I see.22:32
mrdaohh, not focused on devstack. Yeah I'm not sure that you can have different nova compute drivers in play "out of the box" with devstack.22:32
*** meghal has quit IRC22:33
Pradipmrda:  okay got it. So if I want to create such setup, its not possible with devstack, right? What's the best way to do it?22:34
*** meghal has joined #openstack-ironic22:35
mrdaPradip: Well, you could use devstack to standup either VM or BM compute, and manually start the other?22:35
mrdaPradip: I haven't done it personally, but it should be possible22:36
Pradipmrda: okay got it. thanks22:36
openstackgerritJim Rollenhagen proposed openstack/ironic: Add Nova scheduler_tracks_instance_changes config to docs  https://review.openstack.org/20675222:38
jrolldocs review for sale ^^22:38
jrollif you don't want your ironic environment to completely fall over at scale, I highly recommend reading this :P22:39
*** igordcard has quit IRC22:39
Pradipjroll: thanks.22:41
jroll:)22:42
jrollthat's for everyone, but you're welcome :)22:42
* NobodyCam is back22:56
*** Guest47142 is now known as dan22:56
*** penick has quit IRC23:01
*** puranamr has quit IRC23:02
*** Haomeng|2 has joined #openstack-ironic23:09
*** Haomeng has quit IRC23:11
*** romainh has left #openstack-ironic23:15
*** r-daneel has joined #openstack-ironic23:24
*** NobodyCam has quit IRC23:24
*** r-daneel has quit IRC23:24
*** NobodyCam has joined #openstack-ironic23:24
*** r-daneel has joined #openstack-ironic23:25
*** ijw has quit IRC23:29
*** sinval_ has joined #openstack-ironic23:32
*** ijw has joined #openstack-ironic23:34
*** davideagnello has quit IRC23:42
*** davideagnello has joined #openstack-ironic23:45
*** lucas-dinner has quit IRC23:51
jrollthis broke our hardware manager API :( :( :( https://review.openstack.org/#/c/191751/23:53
NobodyCambadly?23:56
NobodyCamand can we add a test to catch it?23:56
jrollwe probably could23:58
jrollsimilar to the one nova has23:58
JoshNangnot too badly, but we need to add an extra param in the downstream hardware managers. but still :(23:58
jrollI mean, this is how bad, not a huge deal https://github.com/rackerlabs/onmetal-ironic-hardware-manager/pull/4123:58
jrollbut we try to guarantee that API to some extent23:59
* jroll off for the night, cheers23:59

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