| *** naohirot has joined #openstack-ironic | 00:03 | |
| *** oomichi has joined #openstack-ironic | 00:04 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Make MTU configurable https://review.openstack.org/194876 | 00:11 |
|---|---|---|
| *** ijw_ has joined #openstack-ironic | 00:18 | |
| *** smoriya has joined #openstack-ironic | 00:20 | |
| *** bradjones has quit IRC | 00:34 | |
| *** bradjones has joined #openstack-ironic | 00:35 | |
| *** bradjones has quit IRC | 00:35 | |
| *** bradjones has joined #openstack-ironic | 00:35 | |
| *** cing has joined #openstack-ironic | 00:43 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Add the rest of the cookiecutter-generated files for testing https://review.openstack.org/194893 | 00:46 |
| *** ijw_ has quit IRC | 00:48 | |
| *** ijw_ has joined #openstack-ironic | 01:05 | |
| rloo | mrda: hi, I've got a few secs now. | 01:29 |
| *** chenglch has joined #openstack-ironic | 01:37 | |
| openstackgerrit | Sinval Vieira Mendes Neto proposed openstack/ironic: Adds port creation passing the name of the node instead of the UUID of the node https://review.openstack.org/193439 | 01:38 |
| *** Marga_ has joined #openstack-ironic | 01:39 | |
| *** mitchjameson has quit IRC | 01:46 | |
| *** rloo has quit IRC | 02:02 | |
| *** cing has quit IRC | 02:16 | |
| *** zhenguo has joined #openstack-ironic | 02:18 | |
| *** zz_natorious is now known as natorious | 02:42 | |
| *** ramineni has joined #openstack-ironic | 02:44 | |
| *** natorious is now known as zz_natorious | 02:51 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Remove redundant python-pip package install https://review.openstack.org/194915 | 02:57 |
| *** yog_ has joined #openstack-ironic | 03:03 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic-specs: Add Zapping support to iLO drivers https://review.openstack.org/145404 | 03:06 |
| TheJulia | Haomeng|2: if the review I just posted above passes, we will need to go ahead and land it since the gate is job is broken on us at the moment | 03:09 |
| Haomeng|2 | TheJulia: ok, let me +a once https://review.openstack.org/#/c/194915/ pass gate, thanks Julia | 03:12 |
| TheJulia | Thank you, I'm likely going to sleep in a few minutes | 03:13 |
| Haomeng|2 | TheJulia: good night, and nice dream:) | 03:13 |
| Haomeng|2 | TheJulia: see you:) | 03:14 |
| *** bradjones has quit IRC | 03:16 | |
| *** bradjones has joined #openstack-ironic | 03:18 | |
| *** bradjones has quit IRC | 03:18 | |
| *** bradjones has joined #openstack-ironic | 03:18 | |
| *** martini has joined #openstack-ironic | 03:19 | |
| *** krtaylor has joined #openstack-ironic | 03:20 | |
| *** boris-42 has joined #openstack-ironic | 03:23 | |
| *** martini has quit IRC | 03:26 | |
| *** martini has joined #openstack-ironic | 03:33 | |
| *** martini has quit IRC | 03:33 | |
| openstackgerrit | Merged openstack/bifrost: Remove redundant python-pip package install https://review.openstack.org/194915 | 03:36 |
| *** martini has joined #openstack-ironic | 03:43 | |
| *** martini has quit IRC | 03:46 | |
| *** coolsvap|away is now known as coolsvap | 04:03 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Make ilo drivers standalone work without swift https://review.openstack.org/193478 | 04:05 |
| *** marios has quit IRC | 04:16 | |
| *** slagle_ has joined #openstack-ironic | 04:16 | |
| *** coolsvap has quit IRC | 04:17 | |
| *** slagle has quit IRC | 04:18 | |
| *** marios has joined #openstack-ironic | 04:20 | |
| *** lazy_prince has joined #openstack-ironic | 04:21 | |
| openstackgerrit | Merged openstack/bifrost: Make MTU configurable https://review.openstack.org/194876 | 04:21 |
| openstackgerrit | Merged openstack/bifrost: Remove 'Internet Access' from Requirements https://review.openstack.org/193776 | 04:21 |
| *** marios has quit IRC | 04:23 | |
| *** slagle has joined #openstack-ironic | 04:23 | |
| *** coolsvap has joined #openstack-ironic | 04:23 | |
| *** slagle_ has quit IRC | 04:26 | |
| *** Marga_ has quit IRC | 04:27 | |
| lazy_prince | o/ morning.. | 04:28 |
| *** marios has joined #openstack-ironic | 04:28 | |
| *** rameshg87 has joined #openstack-ironic | 04:34 | |
| *** Nisha has joined #openstack-ironic | 04:37 | |
| *** yuikotakada has joined #openstack-ironic | 04:39 | |
| *** Marga_ has joined #openstack-ironic | 04:40 | |
| *** achanda has joined #openstack-ironic | 04:44 | |
| *** chenglch has quit IRC | 04:51 | |
| *** chenglch has joined #openstack-ironic | 04:51 | |
| krotscheck | devananda: Comment on https://review.openstack.org/#/c/194424/ , asking for a blueprint. | 05:06 |
| krotscheck | Not certain if I want to do that or write a spec. | 05:06 |
| *** achanda has quit IRC | 05:06 | |
| *** chenglch|2 has joined #openstack-ironic | 05:10 | |
| *** chenglch has quit IRC | 05:10 | |
| openstackgerrit | Haomeng,Wang proposed openstack/bifrost: Add the rest of the cookiecutter-generated files for testing https://review.openstack.org/194893 | 05:14 |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: iLO Zapping Support https://review.openstack.org/194940 | 05:16 |
| *** saripurigopi has joined #openstack-ironic | 05:17 | |
| saripurigopi | morning Ironic | 05:17 |
| Haomeng|2 | saripurigopi: morning:) | 05:19 |
| saripurigopi | hello Haomeng|2 | 05:19 |
| Haomeng|2 | saripurigopi: :) | 05:19 |
| *** zz_natorious is now known as natorious | 05:26 | |
| *** lazy_prince has quit IRC | 05:27 | |
| *** achanda has joined #openstack-ironic | 05:30 | |
| *** natorious is now known as zz_natorious | 05:32 | |
| *** achanda has quit IRC | 05:35 | |
| *** mgarza_ has joined #openstack-ironic | 05:41 | |
| *** zz_natorious is now known as natorious | 05:50 | |
| *** saripurigopi has quit IRC | 05:51 | |
| *** Marga_ has quit IRC | 06:00 | |
| *** zhenguo has quit IRC | 06:03 | |
| *** Marga_ has joined #openstack-ironic | 06:06 | |
| *** e0ne has joined #openstack-ironic | 06:07 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add unit test for ilo_deploy _configure_vmedia_boot() https://review.openstack.org/194047 | 06:11 |
| *** radek__ has joined #openstack-ironic | 06:12 | |
| *** oomichi has quit IRC | 06:18 | |
| *** e0ne has quit IRC | 06:20 | |
| *** bradjones has quit IRC | 06:20 | |
| *** bradjones has joined #openstack-ironic | 06:22 | |
| *** bradjones has quit IRC | 06:22 | |
| *** bradjones has joined #openstack-ironic | 06:22 | |
| *** dasm|afk is now known as dasm | 06:22 | |
| *** e0ne has joined #openstack-ironic | 06:24 | |
| *** amotoki has joined #openstack-ironic | 06:29 | |
| *** ukalifon1 has joined #openstack-ironic | 06:33 | |
| *** saripurigopi has joined #openstack-ironic | 06:34 | |
| *** e0ne is now known as e0ne_ | 06:36 | |
| *** bradjones has quit IRC | 06:36 | |
| *** bradjones has joined #openstack-ironic | 06:39 | |
| *** bradjones has quit IRC | 06:39 | |
| *** bradjones has joined #openstack-ironic | 06:39 | |
| *** e0ne_ has quit IRC | 06:42 | |
| *** e0ne has joined #openstack-ironic | 06:45 | |
| *** e0ne is now known as e0ne_ | 06:46 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Make ilo drivers standalone work without swift https://review.openstack.org/193478 | 06:46 |
| *** e0ne_ has quit IRC | 06:55 | |
| *** coolsvap has quit IRC | 07:01 | |
| *** marios has quit IRC | 07:01 | |
| *** slagle_ has joined #openstack-ironic | 07:02 | |
| *** slagle has quit IRC | 07:05 | |
| *** marios has joined #openstack-ironic | 07:07 | |
| saripurigopi | is there a way we can do same operation on multiple nodes? like supported boot devices? or moving them to maintenance state? | 07:08 |
| *** romcheg has joined #openstack-ironic | 07:08 | |
| *** romcheg has quit IRC | 07:10 | |
| *** romcheg has joined #openstack-ironic | 07:10 | |
| *** coolsvap has joined #openstack-ironic | 07:10 | |
| *** romcheg has quit IRC | 07:10 | |
| *** coolsvap is now known as coolsvap|away | 07:11 | |
| *** coolsvap|away is now known as coolsvap | 07:12 | |
| *** dguerri` is now known as dguerri | 07:12 | |
| *** athomas has joined #openstack-ironic | 07:14 | |
| *** coolsvap is now known as coolsvap|away | 07:15 | |
| Haomeng|2 | saripurigopi: I think we can have such *bulk-operations*, maybe we can create some bulk-scripts to call command/api for each node | 07:17 |
| *** Haomeng|2 is now known as Haomeng | 07:17 | |
| *** coolsvap|away is now known as coolsvap | 07:18 | |
| Haomeng | saripurigopi: and it is useful for user to control multiple nodes at same time | 07:18 |
| Haomeng | I run with IPA, encountered "Failed to start Switch Root" issue during IPA OS starting, any ideas? | 07:23 |
| rameshg87 | Haomeng: which ipa are you using - built from core os ? | 07:27 |
| *** dguerri is now known as dguerri` | 07:27 | |
| Haomeng | rameshg87: it is dracut, which is created by devstack scripts | 07:28 |
| rameshg87 | Haomeng: yeah, afaik devstack creates the coreos ramdisk | 07:28 |
| Haomeng | rameshg87: maybe we need to change IPA OS type, devstack should have such env var to set | 07:29 |
| rameshg87 | Haomeng: it doesn't reach the the login prompt, is it ? | 07:29 |
| Haomeng | rameshg87: yes, break OS starting | 07:30 |
| Haomeng | rameshg87: and enter the shell mode:) | 07:30 |
| Haomeng | rameshg87: let me try to change IPA OS type | 07:31 |
| rameshg87 | Haomeng: I don't think it's possible today with devstack | 07:31 |
| Haomeng | rameshg87: yes | 07:31 |
| Haomeng | rameshg87: checking devstack scripts... | 07:31 |
| rameshg87 | Haomeng: you can try adding ssh keys into ipa ramdisk and see if you can ssh | 07:31 |
| Haomeng | rameshg87: ok, from console, I can run shell command in IPA now:) | 07:32 |
| rameshg87 | oh | 07:32 |
| Haomeng | rameshg87: can check system log, but can not find the root cause:) | 07:32 |
| rameshg87 | Haomeng: https://blog.kingj.net/2015/03/26/how-to/failed-to-start-switch-root-error-when-booting-coreos/ | 07:33 |
| rameshg87 | Haomeng: do you have enough memory to boot it ? are you trying it on a vm ? | 07:33 |
| *** mgarza_ has quit IRC | 07:33 | |
| rameshg87 | I remember we had issue with 512 mb before | 07:33 |
| *** romainh has joined #openstack-ironic | 07:35 | |
| *** Haomeng has quit IRC | 07:37 | |
| *** Haomeng has joined #openstack-ironic | 07:37 | |
| *** dguerri` is now known as dguerri | 07:42 | |
| *** yonglihe has quit IRC | 07:43 | |
| *** dguerri is now known as dguerri` | 07:44 | |
| *** lazy_prince has joined #openstack-ironic | 07:48 | |
| *** zhenguo has joined #openstack-ironic | 07:50 | |
| *** lucasagomes has joined #openstack-ironic | 08:04 | |
| *** derekh has joined #openstack-ironic | 08:07 | |
| *** dtantsur|afk is now known as dtantsur | 08:09 | |
| *** Nisha has quit IRC | 08:09 | |
| dtantsur | Morning Ironic! | 08:09 |
| lucasagomes | dtantsur, morning | 08:14 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: Get a subset of fields from Ports and Chassis https://review.openstack.org/183549 | 08:14 |
| dtantsur | lucasagomes, hey. reviewing ahc-tools stuff, I'm a bit worried that people still use node.list(associated=...) to detect active/non-active nodes | 08:15 |
| dtantsur | lucasagomes, should we start telling people to use provision_state instead? | 08:15 |
| dtantsur | (probably we don't have good means to filter by several provision states though) | 08:15 |
| lucasagomes | dtantsur, :-/ I think for AHC stuff it does make sense yeah | 08:15 |
| lucasagomes | usually associated is the nova way to figure if the instance is active | 08:16 |
| dtantsur | lucasagomes, change in question: https://review.gerrithub.io/#/c/237390/1/ahc_tools/match.py | 08:16 |
| lucasagomes | dtantsur, yes, because in the standalone version instance_uuid is optional | 08:16 |
| dtantsur | lucasagomes, oh, but we don't have means to filter by provision_state >_< | 08:17 |
| lucasagomes | no? urgh... | 08:17 |
| dtantsur | lucasagomes, https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L862-L864 | 08:17 |
| * dtantsur is reporting a bug | 08:17 | |
| lucasagomes | __ | 08:25 |
| lucasagomes | ++ | 08:25 |
| lucasagomes | plus in ahc-tool he may want to have limit=0 | 08:25 |
| lucasagomes | so it will return all nodes instead of the max number of the collection | 08:25 |
| * rameshg87 wonders what is ahc-tools | 08:25 | |
| rameshg87 | morning dtantsur lucasagomes | 08:25 |
| dtantsur | lucasagomes, right | 08:25 |
| dtantsur | rameshg87, see our mail on ML | 08:25 |
| dtantsur | and g'morning | 08:25 |
| saripurigopi | Haomeng: sorry I was away. Yes, it'll be more useful, especially on large deployments. | 08:25 |
| rameshg87 | oh yeah, thanks .. I think I remember now .. | 08:25 |
| *** openstack has quit IRC | 08:25 | |
| *** openstack has joined #openstack-ironic | 08:27 | |
| yuikotakada | dtantsur, o/ | 08:27 |
| saripurigopi | what does 'o/' mean? I'm still adapting to IRC. | 08:29 |
| dtantsur | saripurigopi, this looks like a person raising a hand in a greeting gesture :) can be also used like _o/ | 08:30 |
| saripurigopi | huh, got it now :-) dtantsur _o/ | 08:31 |
| dtantsur | :) | 08:31 |
| *** jcoufal has joined #openstack-ironic | 08:31 | |
| rameshg87 | \o/ is for a person jumping in joy | 08:33 |
| *** Haomeng has quit IRC | 08:33 | |
| rameshg87 | :) | 08:33 |
| *** Haomeng has joined #openstack-ironic | 08:33 | |
| *** e0ne is now known as e0ne_ | 08:34 | |
| saripurigopi | rameshg87 :-) \o/ | 08:34 |
| *** viktors|afk is now known as viktors | 08:35 | |
| saripurigopi | lucasagomes, rameshg87, dtantsur, Haomeng : what do you guys think about having some option for bulk-operations ?? Is it already available? | 08:35 |
| dtantsur | saripurigopi, we've been thinking about it for a while. the only problem is that it's hard to fit bulk operations into HTTP API paradigm | 08:35 |
| Haomeng | saripurigopi: dont think it is available | 08:36 |
| korean101 | hi ironic! | 08:36 |
| korean101 | i have an another question | 08:37 |
| saripurigopi | dtantsur: yeah right. | 08:37 |
| Haomeng | korean101: ok, welcome | 08:37 |
| korean101 | can i ask you? | 08:37 |
| korean101 | Haomeng: hi :) | 08:37 |
| Haomeng | korean101: sure | 08:37 |
| saripurigopi | koream101: o/ | 08:37 |
| korean101 | We don't recommend creating raw images by dd'ing block device files, we discuss how to create raw images later. | 08:37 |
| korean101 | http://docs.openstack.org/image-guide/content/ch_introduction.html | 08:37 |
| korean101 | above link | 08:38 |
| korean101 | we discuss how to create raw images later. <-- how to? | 08:38 |
| korean101 | not use dd command? | 08:38 |
| *** e0ne_ has quit IRC | 08:39 | |
| korean101 | i use dd command and qemu-img convert to qcow2 | 08:39 |
| Haomeng | korean101: our image is generated by disk image builder, it is raw image first then dib convert it to qcow or tar | 08:39 |
| Haomeng | korean101: not sure what is your concern | 08:39 |
| Haomeng | korean101: do you want to convert raw to qcow2? | 08:40 |
| Haomeng | korean101: I remeber there is option to create tar format with dib | 08:40 |
| *** e0ne has joined #openstack-ironic | 08:40 | |
| *** athomas has quit IRC | 08:40 | |
| korean101 | i do boot whole disk | 08:40 |
| korean101 | and my own windows 2008 R2 server | 08:41 |
| korean101 | i use dd command and make raw image | 08:41 |
| korean101 | and use qemu-img convert and make raw->qcow2 iamge | 08:41 |
| Haomeng | korean101: Ironic can support qcow2, but it is converted to raw before *dd* | 08:41 |
| korean101 | yes | 08:41 |
| korean101 | i also converted | 08:42 |
| Haomeng | korean101: yes, bot raw and qcow2 can be as input for ironic to consume | 08:42 |
| korean101 | ok | 08:42 |
| Haomeng | korean101: so what is your question, what is not working? | 08:42 |
| korean101 | and i put in glance qcow2 image | 08:42 |
| korean101 | and boot_whole_disk id working!!! | 08:42 |
| *** ifarkas has joined #openstack-ironic | 08:42 | |
| Haomeng | korean101: glance will not care about the image format | 08:42 |
| korean101 | but so long time | 08:43 |
| korean101 | to deploy | 08:43 |
| Haomeng | korean101: yes, dd is working on disk level, will *copy* all disk block data even they are 000 or random which are not used by filesystems | 08:44 |
| korean101 | ok block device level | 08:44 |
| Haomeng | korean101: so depends on your size of whole disk | 08:44 |
| Haomeng | korean101: yes, block device level | 08:44 |
| korean101 | http://docs.openstack.org/image-guide/content/centos-image.html | 08:45 |
| *** Nisha_away has joined #openstack-ironic | 08:45 | |
| *** e0ne has quit IRC | 08:45 | |
| korean101 | is this recommended? rather than dd command? | 08:45 |
| Haomeng | korean101: so we have more idea to speed such process, maybe we can recreate paration and filesystems, then copy source to dest on file level, to ignore the unused block data | 08:45 |
| korean101 | oh really? | 08:46 |
| korean101 | such a windows images? | 08:46 |
| Haomeng | korean101: for windows filesystems, need to support with fat32/ntfs driver I think | 08:47 |
| yuikotakada | dtantsur, until what time will you work today? I will go home now and restart to work at my home | 08:47 |
| Haomeng | korean101: your reference like *kickstart* | 08:47 |
| yuikotakada | And I'd like to talk about our integration tests a little bit later | 08:48 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic-specs: Fixes for generic RAID interface https://review.openstack.org/192936 | 08:48 |
| dtantsur | yuikotakada, I'll be relocating to the office right now. I'll be back in 1-2 hrs, then until evening (which is deep night for you) | 08:48 |
| *** Nisha has quit IRC | 08:48 | |
| Haomeng | korean101: you can check ironic conductor log to see how time consumed to *dd*, how is the speed | 08:48 |
| Haomeng | korean101: so understand your concern is about *dd* performance, right? | 08:49 |
| *** bradjones has quit IRC | 08:49 | |
| korean101 | Haomeng: yes | 08:50 |
| korean101 | Haomeng: rootwrap dd | 08:50 |
| lucasagomes | korean101, are you using pxe_ driver? | 08:50 |
| korean101 | yes | 08:51 |
| *** bradjones has joined #openstack-ironic | 08:51 | |
| *** bradjones has quit IRC | 08:51 | |
| *** bradjones has joined #openstack-ironic | 08:51 | |
| rameshg87 | dtantsur: lucasagomes: ^^ (some lines above) proposed a new patch set | 08:51 |
| Haomeng | korean101: maybe ipa will be good performance | 08:51 |
| lucasagomes | by using the agent_ may speed up things a bit since you won't need to copy it over iSCSI | 08:51 |
| yuikotakada | dtantsur, it sounds good for me | 08:51 |
| lucasagomes | rameshg87, thanks! | 08:51 |
| yuikotakada | see you later, Ironic :) | 08:51 |
| dtantsur | see you :) | 08:51 |
| Haomeng | yuikotakada: yes | 08:51 |
| lucasagomes | yuikotakada, have a great evening! | 08:51 |
| korean101 | Haomeng: oh thank you!! | 08:51 |
| rameshg87 | yuikotakada: have a good evening | 08:51 |
| korean101 | yuikotakada: see you~ | 08:51 |
| *** yuikotakada is now known as yuikot_train | 08:51 | |
| * lucasagomes brb | 08:51 | |
| Haomeng | korean101: welcome | 08:51 |
| *** romcheg has joined #openstack-ironic | 08:54 | |
| *** dtantsur is now known as dtantsur|brb | 08:55 | |
| *** Nisha has joined #openstack-ironic | 08:56 | |
| *** pelix has joined #openstack-ironic | 08:57 | |
| *** Nisha_away has quit IRC | 08:59 | |
| *** ndipanov has joined #openstack-ironic | 09:02 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Do not use "private" attribute in AuthTokenMiddleware https://review.openstack.org/195000 | 09:04 |
| lucasagomes | rameshg87, he GET states/power | 09:07 |
| lucasagomes | re GET...* | 09:07 |
| openstackgerrit | Sergey Vilgelm proposed openstack/ironic-lib: Switch to oslo.service https://review.openstack.org/195002 | 09:07 |
| lucasagomes | rameshg87, currently we have no way to GET the power state directly from the BMC via our API | 09:08 |
| *** Marga_ has quit IRC | 09:08 | |
| lucasagomes | when we do a GET on a node it returns the information from the database | 09:08 |
| lucasagomes | rameshg87, I wonder if offering a GET v1/nodes/<uuid>/states/power (that previous was only PUT) that would get the power state from the bmc would be important | 09:09 |
| *** bradjones has quit IRC | 09:09 | |
| rameshg87 | lucasagomes: hi | 09:11 |
| rameshg87 | lucasagomes: but again it would just return the cached value, right ? | 09:11 |
| *** Marga_ has joined #openstack-ironic | 09:11 | |
| lucasagomes | rameshg87, the idea was no, to talk directly to the BMC | 09:11 |
| *** bradjones has joined #openstack-ironic | 09:11 | |
| *** bradjones has quit IRC | 09:11 | |
| *** bradjones has joined #openstack-ironic | 09:11 | |
| lucasagomes | but that may hurt our idea of never talking to BMC synchronously | 09:12 |
| rameshg87 | lucasagomes: yeah, I was coming to that | 09:12 |
| rameshg87 | lucasagomes: it will be good, as we always rely on sync_power_states periodic task to update the power status for us | 09:12 |
| lucasagomes | rameshg87, we could do it async... like run a get power state and update the node | 09:12 |
| rameshg87 | lucasagomes: ah but then it doesn't fit into GET :) | 09:13 |
| lucasagomes | rameshg87, yeah, should be a PUT or POST | 09:13 |
| lucasagomes | with a refresh | 09:13 |
| *** dguerri` is now known as dguerri | 09:13 | |
| rameshg87 | yeah | 09:13 |
| rameshg87 | sounds like triggering a refresh | 09:13 |
| openstackgerrit | Tan Lin proposed openstack/ironic: Add a new boot section 'trusted_boot' for PXE https://review.openstack.org/191661 | 09:14 |
| lucasagomes | rameshg87, yeah, I will put some thoughts on this | 09:14 |
| rameshg87 | lucasagomes: as far as I have used ironic, when ever I wanted to confirm for example if bare metal is powered off | 09:15 |
| lucasagomes | rameshg87, just saw an internal email about how sync the power state of the node works | 09:15 |
| rameshg87 | lucasagomes: I try to set power state off | 09:15 |
| lucasagomes | rameshg87, right, yeah I suggested that to speed up things | 09:15 |
| lucasagomes | but doesn't seems ideal | 09:15 |
| rameshg87 | lucasagomes: well sort of okay, because most of the power drivers check current power state before triggering some operation | 09:16 |
| rameshg87 | lucasagomes: but may not be idea :) | 09:16 |
| rameshg87 | *ideal | 09:16 |
| lucasagomes | yeah | 09:16 |
| lucasagomes | I just pinged you because that may be an use case for the GET states/power | 09:17 |
| lucasagomes | that we discussed yesterday | 09:17 |
| lucasagomes | if we go with the sync approach | 09:17 |
| rameshg87 | yeah exactly :) | 09:17 |
| rameshg87 | leaving poor guy /states/provision out ;-) | 09:17 |
| rameshg87 | there might be a reason for /states/power some day | 09:18 |
| lucasagomes | heh | 09:18 |
| openstackgerrit | Sergey Vilgelm proposed openstack/ironic-python-agent: Switch to oslo.service https://review.openstack.org/195005 | 09:18 |
| openstackgerrit | Sergey Vilgelm proposed openstack/ironic-python-agent: Switch to oslo.service https://review.openstack.org/195005 | 09:19 |
| openstackgerrit | Sergey Vilgelm proposed openstack/ironic-lib: Switch to oslo.service https://review.openstack.org/195002 | 09:20 |
| *** boris-42 has quit IRC | 09:22 | |
| *** Haomeng has quit IRC | 09:23 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add iBoot driver documentation https://review.openstack.org/193546 | 09:23 |
| *** Marga_ has quit IRC | 09:27 | |
| *** rsFF has quit IRC | 09:34 | |
| sambetts | Morning all o/ | 09:35 |
| openstackgerrit | Sergey Vilgelm proposed openstack/ironic: Switch to oslo.service https://review.openstack.org/195008 | 09:36 |
| *** e0ne has joined #openstack-ironic | 09:46 | |
| *** Marga_ has joined #openstack-ironic | 09:48 | |
| *** lazy_prince has quit IRC | 09:50 | |
| *** MattMan has joined #openstack-ironic | 09:55 | |
| *** MattMan has left #openstack-ironic | 09:55 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Update cached images based on update time https://review.openstack.org/195017 | 09:57 |
| *** mark-os has joined #openstack-ironic | 09:58 | |
| mark-os | Has the Ironic API changed since the 20th of May? | 09:59 |
| mark-os | Also, is there a place where I can track the changes to the API? | 09:59 |
| *** yog_ has quit IRC | 10:02 | |
| *** Marga_ has quit IRC | 10:05 | |
| lucasagomes | mark-os, what did you see that changed? | 10:05 |
| TheJulia | mark-os: And what branch are you using? | 10:06 |
| *** Marga_ has joined #openstack-ironic | 10:06 | |
| TheJulia | btw, good morning everyone | 10:06 |
| lucasagomes | TheJulia, good morning | 10:08 |
| * TheJulia goes and finds breakfast | 10:09 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Image service should not be set in ImageCache constructor https://review.openstack.org/194752 | 10:10 |
| *** MattMan has joined #openstack-ironic | 10:13 | |
| *** zhenguo has quit IRC | 10:13 | |
| mark-os | lucasagomes , TheJulia : I'm looking at an ironic dashboard and it's receiving an error receiving node details from Ironic API | 10:14 |
| lucasagomes | mark-os, hmm do you have the error handy there? | 10:14 |
| mark-os | I was guessing that perhaps in the past it used to work and that there's been changes to the API which has caused it to stop working | 10:14 |
| mark-os | Someone raised an issue in the GitHub page for the repo: https://github.com/rackerlabs/onmetal-dashboard/issues/2 | 10:15 |
| *** Nisha has quit IRC | 10:21 | |
| lucasagomes | mark-os, hmmm "Error retrieving node details from Ironic API" this is a | 10:25 |
| lucasagomes | GET /v1/nodes/detail ? | 10:25 |
| *** romcheg has quit IRC | 10:25 | |
| openstackgerrit | Merged openstack/ironic: Refactor method to add or update capability string https://review.openstack.org/162451 | 10:25 |
| mark-os | What you mean "GET /v1/nodes/detail ?" | 10:26 |
| mark-os | I've identified the method it's failing at if that helps? | 10:26 |
| openstackgerrit | Merged openstack/ironic: Transition state machine to use automaton oslo lib. https://review.openstack.org/191985 | 10:26 |
| *** Marga_ has quit IRC | 10:28 | |
| *** coolsvap is now known as coolsvap|away | 10:28 | |
| *** Marga_ has joined #openstack-ironic | 10:28 | |
| lucasagomes | mark-os, it doesn't ring me a bell anything changing on that area | 10:28 |
| lucasagomes | mark-os, I've tried it locally http://paste.openstack.org/show/319635/ | 10:29 |
| lucasagomes | https://github.com/rackerlabs/onmetal-dashboard/blob/05389bff210394e426aa3cb6e5cab10b060e5ebf/lib/auth.js#L134 | 10:29 |
| lucasagomes | trying to mimic what you guys do there | 10:29 |
| *** yuanying has quit IRC | 10:29 | |
| *** romcheg has joined #openstack-ironic | 10:31 | |
| mark-os | lucasagomes: That dashboard isn't mine btw, I'm just trying to get it to work :) | 10:31 |
| lucasagomes | oh | 10:32 |
| mark-os | lucasagomes: What about GET /v2.0/tenants ? | 10:32 |
| lucasagomes | mark-os, ok... see if you can get a better traceback from the errors | 10:33 |
| lucasagomes | mark-os, that's keystone related | 10:33 |
| lucasagomes | sure Ironic will call it to authenticate | 10:33 |
| * lucasagomes dunno if something changed there | 10:33 | |
| *** dtantsur|brb is now known as dtantsur | 10:33 | |
| dtantsur | sambetts, o/ | 10:34 |
| sambetts | dtantsur: Hey :D | 10:34 |
| dtantsur | sambetts, are you by chance looking for a simple thing to fix? | 10:35 |
| mark-os | lucasagomes: so what you're saying is that it could be a keystone issue? | 10:35 |
| mark-os | Here's where the error is coming from: https://github.com/rackerlabs/onmetal-dashboard/blob/05389bff210394e426aa3cb6e5cab10b060e5ebf/node_modules/pkgcloud/lib/pkgcloud/openstack/context/identity.js#L158 | 10:35 |
| sambetts | dtantsur: I can take a look if you've got something | 10:35 |
| dtantsur | sambetts, https://bugs.launchpad.net/bugs/1462365 | 10:35 |
| openstack | Launchpad bug 1462365 in Ironic Inspector "Replace non-standard "authenticate" option with Ironic-alike auth_strategy" [High,Triaged] | 10:35 |
| dtantsur | the only thing we don't have on track for inspector 2.0 | 10:36 |
| *** tteggel_ has quit IRC | 10:36 | |
| lucasagomes | mark-os, I don't know exactly... but maybe, that's the only command that is failing? | 10:36 |
| mark-os | Also here too: https://github.com/rackerlabs/onmetal-dashboard/blob/05389bff210394e426aa3cb6e5cab10b060e5ebf/lib/hostcache.js#L149 | 10:36 |
| sambetts | dtantsur: Ok, I'll see what I can do, I'm working on setting up a test bed in my lab right now so theres a lot of waiting for things to install :-P | 10:37 |
| *** tteggel has joined #openstack-ironic | 10:37 | |
| dtantsur | sambetts, if you don't find time, no problem, I'll fix it once I'm less busy :) | 10:37 |
| mark-os | Yeah it seems like those methods are causing the error to appear but I can't figure out why it's being caused. I've read one of the TODO items which says ' make this more resiliant (what if multiple active tenants) ' and I have multiple active tenants | 10:38 |
| mark-os | https://github.com/rackerlabs/onmetal-dashboard/blob/05389bff210394e426aa3cb6e5cab10b060e5ebf/node_modules/pkgcloud/lib/pkgcloud/openstack/context/identity.js#L172 | 10:38 |
| mark-os | So I'm guessing that could be the issue but I'm not 100% sure | 10:39 |
| sambetts | dtantsur: I'll assign myself and see if I can get it done :) | 10:39 |
| lucasagomes | mark-os, right, yeah... try to disable the authentication in the ironic api | 10:39 |
| lucasagomes | set auth_strategy=noauth | 10:39 |
| lucasagomes | see if that works | 10:39 |
| mark-os | lucasagomes: how do I do that? | 10:40 |
| lucasagomes | just to isolate the problem | 10:40 |
| lucasagomes | mark-os, edit the ironic.conf file, under the [DEFAULT] section | 10:40 |
| lucasagomes | set "auth_strategy=noauth" | 10:40 |
| lucasagomes | then restart the ironic-api service | 10:40 |
| *** lazy_prince has joined #openstack-ironic | 10:47 | |
| mark-os | lucasagomes: How can I test to see if the noauth is working? | 10:50 |
| *** lazy_prince has quit IRC | 10:50 | |
| *** lazy_prince has joined #openstack-ironic | 10:50 | |
| lucasagomes | mark-os, just make a request to the api without passing a token | 10:56 |
| lucasagomes | mark-os, e.g | 10:56 |
| lucasagomes | $ curl -X GET http://<ip>:6385/v1/nodes/ | 10:56 |
| * rameshg87 goes home | 11:00 | |
| *** rameshg87 has quit IRC | 11:00 | |
| mark-os | {"nodes": []} | 11:01 |
| mark-os | That might be correct because I don't have an active node at this time | 11:01 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Update cached images based on update time https://review.openstack.org/195017 | 11:02 |
| *** ramineni has quit IRC | 11:05 | |
| *** chenglch|2 has quit IRC | 11:05 | |
| lucasagomes | mark-os, active you mean registered? | 11:08 |
| lucasagomes | mark-os, so register one please | 11:08 |
| mark-os | lucasagomes: I can't get access, so keystone hasn't been disabled | 11:08 |
| lucasagomes | mark-os, you can create directly via ironic api | 11:09 |
| mark-os | whenever I issue ironic node-list, it says: The request you have made requires authentication. | 11:09 |
| mark-os | lucasagomes: where can I find the api? | 11:09 |
| lucasagomes | it should allow you to create even when keystone is not present hmmm | 11:09 |
| mark-os | lucasagomes: just issuing the 'ironic' command? | 11:10 |
| lucasagomes | mark-os, curl -X POST http://192.168.122.87:6385/v1/nodes -H "Content-Type: application/json" -H "Accept: application/json" -d '{"driver": "fake"}' | 11:10 |
| lucasagomes | yeah ironic node-create -d fake | 11:10 |
| *** yuikotakada has joined #openstack-ironic | 11:13 | |
| lucasagomes | mark-os, but anyway... it seems to be something about the way ironic authenticates with keystone | 11:13 |
| lucasagomes | not a change in our API | 11:13 |
| mark-os | Yeah, thanks for your help :) | 11:17 |
| mark-os | lucasagomes I'm having issues with disabling the auth so I'll need to sort that out before testing it | 11:17 |
| lucasagomes | mark-os, ok, no problem | 11:17 |
| lucasagomes | yeah lemme know if you find somehting there | 11:17 |
| *** Marga_ has quit IRC | 11:20 | |
| *** yuikotakada has quit IRC | 11:21 | |
| *** yuikotakada has joined #openstack-ironic | 11:21 | |
| *** yuikotakada is now known as yuikot_cooking | 11:22 | |
| *** lucasagomes is now known as lucas-hungry | 11:23 | |
| *** bradjones has quit IRC | 11:29 | |
| *** bradjones has joined #openstack-ironic | 11:30 | |
| *** bradjones has quit IRC | 11:30 | |
| *** bradjones has joined #openstack-ironic | 11:30 | |
| *** yuikot_cooking is now known as yuikotakada | 11:31 | |
| yuikotakada | hi, ironic | 11:31 |
| dtantsur | yuikotakada, welcome back :) | 11:34 |
| yuikotakada | dtantsur, you too :) | 11:35 |
| yuikotakada | dtantsur, Gate patch still has not been merged...Is all we can do just waiting? | 11:36 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Allow to refer to Glance image using its name https://review.openstack.org/189341 | 11:37 |
| dtantsur | yuikotakada, unfortunately yes. I've tried pinging people on #openstack-infra, you can make one more attempt, but largely we can't do anything about it | 11:37 |
| yuikotakada | And also I've started to implement a test kicking inspector from ironic in tempest and I've taken over adam_g's patch about API microversion, but it doesn't go well | 11:38 |
| yuikotakada | dtantsur, oh well | 11:39 |
| *** trown|outttypeww is now known as trown | 11:39 | |
| *** romcheg has quit IRC | 11:48 | |
| dtantsur | yuikotakada, what's the problem? | 11:48 |
| yuikotakada | dtantsur, maybe it will take a little bit long time, it's no problem if we just want to run simple shell script on the gate, but some time we want to run more ironic/ironic+inspector tests with tempest or tempest-lib, right? | 11:52 |
| dtantsur | yuikotakada, yeah, after some time. and it can be pretty large time ;) | 11:52 |
| *** alexpilotti has joined #openstack-ironic | 11:56 | |
| *** dprince has joined #openstack-ironic | 11:57 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 11:59 |
| *** athomas has joined #openstack-ironic | 12:02 | |
| yuikotakada | dtantsur, right :) | 12:05 |
| *** romcheg has joined #openstack-ironic | 12:05 | |
| *** smoriya has quit IRC | 12:08 | |
| *** yuikotakada has quit IRC | 12:18 | |
| *** yuikotakada has joined #openstack-ironic | 12:18 | |
| *** yuikotakada is now known as yuikot_dinner | 12:19 | |
| *** yuikot_dinner has quit IRC | 12:22 | |
| *** yuikot_dinner has joined #openstack-ironic | 12:22 | |
| *** rameshg87 has joined #openstack-ironic | 12:29 | |
| openstackgerrit | Merged openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 12:33 |
| *** lucas-hungry is now known as lucasagomes | 12:34 | |
| sambetts | dtantsur: can I just remove the authenticate option? or do we need to support both for while with a deprecation warning? | 12:36 |
| trown | dtantsur: lucasagomes, thanks for the review on https://review.gerrithub.io/#/c/237390/ | 12:36 |
| dtantsur | sambetts, I'd prefer we deprecate it and remove later | 12:36 |
| dtantsur | trown, np | 12:36 |
| trown | what do you all think of iterating over the node list to check provision state | 12:36 |
| lucasagomes | trown, np | 12:36 |
| trown | clearly having a filter would be ideal, but as you said we dont | 12:37 |
| lucasagomes | trown, yeah iteracting seems hmm very costly | 12:37 |
| lucasagomes | but would work | 12:37 |
| dtantsur | lucasagomes, trown, should work for now, and we should definitely introduce filter by provision_state | 12:38 |
| lucasagomes | yes | 12:38 |
| trown | lucasagomes: I do not think it would be costly for the TripleO use case, and we can use the filter once it exists | 12:38 |
| sambetts | dtantsur: kl kl :D | 12:38 |
| trown | lucasagomes: so I can still add FIXME, but if associated is unreliable, I would rather not rely on it | 12:39 |
| lucasagomes | trown, yeah sounds good, you can iterate the nodes to find the provision state | 12:39 |
| lucasagomes | and add a TODO/FIXME there to change it once it's avaialble via the ironic's api | 12:39 |
| trown | sweet, thanks guys | 12:39 |
| rameshg87 | lucasagomes: earlier we were talking about performing synchronous operations with bmc on the api | 12:45 |
| *** romcheg has quit IRC | 12:45 | |
| rameshg87 | lucasagomes: drac vendor passthru seems to follow that - https://review.openstack.org/#/c/150488/25/ironic/drivers/modules/drac/vendor_passthru.py | 12:45 |
| rameshg87 | lucasagomes: now since it is a vendor passthru, I don't know if we really care about that | 12:45 |
| rameshg87 | any thoughts ? | 12:45 |
| lucasagomes | rameshg87, yeah we actually don't, I mean we can suggest then to not do it | 12:46 |
| rameshg87 | lucasagomes: yeah I was thinking the same, may be not for the ironic standard apis we can still follow that | 12:46 |
| rameshg87 | ack | 12:46 |
| lucasagomes | but I don't think we also offer a good way to vendors to actually offer a way to the client to know when a async operator is completed/failed | 12:47 |
| lucasagomes | even for Ironic, we currently just tell clients to poll stuff | 12:47 |
| lucasagomes | rameshg87, I think that get_boot_devices is sync no? | 12:48 |
| *** romainh has left #openstack-ironic | 12:48 | |
| *** teju has joined #openstack-ironic | 12:48 | |
| rameshg87 | lucasagomes: yeah, but that was before we decided, right ? :) | 12:49 |
| lucasagomes | yeah | 12:49 |
| lucasagomes | but even so, some stuff still complicated if not sync | 12:49 |
| *** yuikot_dinner is now known as yuikotakada | 12:49 | |
| rameshg87 | lucasagomes: yeah may be we could extend vendor passthrus more, allow results of the last to be cached in driver_internal_info and provide the results through API or something | 12:49 |
| rameshg87 | *result of the last vendor passthru of that method | 12:50 |
| *** EmilienM is now known as EmilienM|off | 12:50 | |
| rameshg87 | and may be return the url to get the result back in the API | 12:51 |
| lucasagomes | yeah, using the location header would be good | 12:51 |
| lucasagomes | good for that* | 12:52 |
| dtantsur | well, get_* stuff can't be async, others should | 12:52 |
| lucasagomes | dtantsur, yeah, in the case of power we were discussing about having a way to tell ironic to "refresh" | 12:52 |
| lucasagomes | but I wouldn't mind having GET sync for get power state on that endpoint | 12:53 |
| lucasagomes | I mean, GET /states/power | 12:53 |
| dtantsur | it would be really cool for inspector actually.. | 12:53 |
| rameshg87 | :) | 12:53 |
| lucasagomes | as a API abstracting hardware, having a way to get the power state directly from the BMC seems fine | 12:53 |
| lucasagomes | when GET /v1/nodes/ ... /v1/nodes/NNNN etc... it will come from the db | 12:54 |
| lucasagomes | the power state will come from the db* | 12:54 |
| * lucasagomes will suggest that | 12:54 | |
| dtantsur | please do | 12:55 |
| *** Marga_ has joined #openstack-ironic | 12:57 | |
| *** romcheg has joined #openstack-ironic | 13:01 | |
| *** yuikotakada has quit IRC | 13:03 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Override PXE kernel cmdline via Glance property https://review.openstack.org/178229 | 13:11 |
| *** rloo has joined #openstack-ironic | 13:12 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Correct license header on zuul change parsing code https://review.openstack.org/195073 | 13:16 |
| * TheJulia proceeds to beat self up | 13:16 | |
| *** achanda has joined #openstack-ironic | 13:20 | |
| rameshg87 | TheJulia: just got to know why now :D | 13:24 |
| TheJulia | rameshg87: the change I just put in, messed up stamping the header, I pasted the wrong one originally and didn't notice. :( | 13:26 |
| rameshg87 | :) | 13:26 |
| * TheJulia resumes beating self up, in accordance with all prophecies. | 13:29 | |
| *** Marga_ has quit IRC | 13:29 | |
| *** Marga_ has joined #openstack-ironic | 13:30 | |
| lucasagomes | rameshg87, NobodyCam JayF do you guys know whether most BMCs support IPMI 1.5 and 2.0 ? | 13:30 |
| lucasagomes | I just got a situation here where the "ipmitool power status" command works with "-I lan" (1.5) but fails with "-I lanplus" (2.0) | 13:31 |
| lucasagomes | I wonder whether Ironic should fallback to 1.5 to be more resilient to failures | 13:31 |
| lucasagomes | any thoughts? | 13:31 |
| rameshg87 | lucasagomes: no idea, I rarely work with ipmi :( | 13:32 |
| *** e0ne is now known as e0ne_ | 13:33 | |
| *** teju has quit IRC | 13:35 | |
| *** e0ne_ is now known as e0ne | 13:37 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Addition of shade support to bifrost_inventory.py https://review.openstack.org/194684 | 13:37 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Deprecate authenticate opt in favor of auth_strategy https://review.openstack.org/195082 | 13:41 |
| * rameshg87 is done for the day | 13:42 | |
| rameshg87 | good night (ugt) ironic | 13:42 |
| *** saripurigopi has quit IRC | 13:42 | |
| *** rameshg87 has quit IRC | 13:42 | |
| sambetts | Have a good evening rameshg87 | 13:42 |
| *** yuikotakada has joined #openstack-ironic | 13:49 | |
| *** yuikotakada has quit IRC | 13:49 | |
| naohirot | dtantsur: hello | 13:51 |
| dtantsur | naohirot, o/ | 13:52 |
| naohirot | dtantsur: :-) | 13:52 |
| naohirot | dtantsur: thank you for quick reply | 13:52 |
| dtantsur | yvw | 13:52 |
| naohirot | dtantsur: frankly speaking, I don't know how | 13:53 |
| dtantsur | naohirot, very tl;dr: drop the try..except.. block | 13:54 |
| naohirot | dtantsur: can I let user know the exception.InvalidParameterValue | 13:54 |
| dtantsur | why not? that's a real problem they have | 13:54 |
| naohirot | dtantsur: If my code doesn't catch the exception.InvalidParameterValue, then exception.InvalidParameterValue is propagated to the user? | 13:55 |
| dtantsur | naohirot, I hope so | 13:55 |
| naohirot | dtantsur: that's easy, isn't it? I didn't know that the conductor is so clever :-) | 13:57 |
| rloo | dtantsur, naohirot: I don't know what you're discussing, but you can always try it out if you are unsure. That's one good way to learn how/what the code is doing :) | 13:57 |
| *** zz_jgrimm is now known as jgrimm | 13:57 | |
| naohirot | rloo: hi, sure | 13:58 |
| naohirot | rloo: I'll try it out. | 13:58 |
| *** zz_natorious is now known as natorious | 14:02 | |
| *** marzif has joined #openstack-ironic | 14:02 | |
| *** natorious is now known as zz_natorious | 14:03 | |
| *** absubram has joined #openstack-ironic | 14:03 | |
| *** huats has joined #openstack-ironic | 14:06 | |
| rloo | hello everyone :) | 14:06 |
| lucasagomes | rloo, naohirot sambetts good ugt morning | 14:06 |
| rloo | dtantsur: you're so funny, I was thinking of it but was unsure and thought you'd cringe if I mentioned a microversion bump: https://review.openstack.org/#/c/193587/ | 14:07 |
| naohirot | lucasagomes: good morning | 14:07 |
| *** logan2 has quit IRC | 14:08 | |
| dtantsur | rloo, well, it's a breaking change. I have nothing against gating them :) | 14:08 |
| rloo | dtantsur: but it doesn't break anything does it? only possibly breaks future names... hmm... | 14:09 |
| rloo | dtantsur: I'm finding it harder and harder to review code | 14:09 |
| *** syh has joined #openstack-ironic | 14:10 | |
| dtantsur | rloo, that's one of the reasons why I object to complex microversioning - it makes our life harder... | 14:10 |
| dtantsur | rloo, re breaking: we're using the function in question in WSME types. I *suspect* that it may prevent users from accessing existing wrongly named nodes | 14:11 |
| dtantsur | e.g. with underscores | 14:11 |
| *** zz_natorious is now known as natorious | 14:11 | |
| huats | My understanding of the filters is that I can chosse between baremetal and nova filters, but once I have choosen the baremetal ones, even the "classical" instances will be using the baremetal filters. Is it correct ? Won't it create any issue for using another nova driver (like kvm) ? | 14:11 |
| rloo | dtantsur: oh... from mrda's comment, i assumed it was fine, but i didn't want to check myself. guess i should look. | 14:11 |
| dtantsur | I'm not 100% sure either, but it's worth checking before approving | 14:15 |
| lucasagomes | are you guys aware of http://www.opencontainers.org/ ? | 14:15 |
| *** killer_prince has joined #openstack-ironic | 14:15 | |
| lucasagomes | looks like a great initiative with coreos, docker, red hat, google (and a bunch of other companies) | 14:16 |
| *** lazy_prince has quit IRC | 14:19 | |
| sambetts | Morning lucasagomes o/ | 14:20 |
| *** killer_prince has quit IRC | 14:20 | |
| dtantsur | lucasagomes, I've heard about it, yeah | 14:21 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add example deploy all available playbook https://review.openstack.org/195100 | 14:21 |
| *** stpierre has joined #openstack-ironic | 14:22 | |
| stpierre | i've been asked to protect the change i've proposed in https://review.openstack.org/#/c/193587/ with a microversion. i'm not very familiar with ironic, though -- can someone help me figure out how to do that? | 14:23 |
| *** achanda has quit IRC | 14:27 | |
| * lucasagomes would love to drop that regex and allow free form names | 14:28 | |
| lucasagomes | stpierre, https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/__init__.py#L66 | 14:29 |
| lucasagomes | you gotta bump that version and add a safe guard function for versions > 1.X (the version you bumped) | 14:29 |
| lucasagomes | like https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L78-L82 | 14:30 |
| stpierre | hah, okay | 14:30 |
| rloo | lucasagomes: agreed wrt free-form names. | 14:30 |
| stpierre | as i understand it -- and i could be wrong -- that regex is used for multiple things, some of which are actual RFC 952 hostnames (which should be subject to that regex) and some of which are not (which needn't be). | 14:30 |
| lucasagomes | stpierre, but that said, I would encourage you to stop using those RFCs and just allow free form for the name | 14:31 |
| rloo | stpierre: I'd prefer if you just got rid of that and did the free-form names thing | 14:31 |
| *** logan2 has joined #openstack-ironic | 14:31 | |
| stpierre | honestly, i don't have a horse in this race so I don't particularly care how it gets fixed | 14:31 |
| lucasagomes | stpierre, the reason is we have no reason to use hostname or fqdn | 14:31 |
| lucasagomes | because that name won't be set as the hostname for the instance | 14:32 |
| lucasagomes | so it's confusing/misleading for the user and super restrict | 14:32 |
| lucasagomes | like "node_1" is an invalid name according to that RFC | 14:32 |
| stpierre | so would allowing freeform names be considered a backwards incompatible change? seems like it'd be more forgiving, so wouldn't need a microversion | 14:32 |
| rloo | stpierre: i just feel like you're spending time on something that we want to change (and I don't want to waste your time or reviewers' time) | 14:32 |
| *** UForgotten has quit IRC | 14:33 | |
| rloo | stpierre: it needs a microversion still (although I think we're trying to change that so microversions are only needed for backwards-incompat changes) | 14:33 |
| stpierre | the suggestion in the bug you linked me to is string.letters + string.digits + string.punctuation; is that what you want to go with, or would you prefer all printable unicode? (fwiw, that's what nova allows in most of its resource names) | 14:35 |
| openstackgerrit | Merged openstack/ironic: API: Get a subset of fields from Ports and Chassis https://review.openstack.org/183549 | 14:35 |
| *** Nisha has joined #openstack-ironic | 14:36 | |
| rloo | stpierre: that's a good question. I hadn't put any thought into it. sec. | 14:36 |
| stpierre | if you're curious: https://github.com/openstack/nova/commit/f40b6a1d5c91628f78fdbc0417a3a394a69ab716 | 14:38 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add additional note dependent upon os_ironic_facts https://review.openstack.org/195111 | 14:39 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/194320 | 14:39 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-lib: Updated from global requirements https://review.openstack.org/194321 | 14:39 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/193986 | 14:39 |
| rloo | stpierre: thx for the link. are you in a hurry to fix whatever problem you're trying to fix? | 14:39 |
| *** Marga_ has quit IRC | 14:40 | |
| rloo | stpierre: i took a look at the bug, and it seems like an update to the spec may be needed first to get 'agreement' with the change to freeform. | 14:40 |
| rloo | stpierre: which means it isn't clear how long it will take, but quite possibly longer than 'fixing a bug' | 14:40 |
| stpierre | sort of. it's blocking a rally spec; this is one of the most restrictive resource naming schemes in any project, so we need to figure out how it will get changed before we can decide on a scheme for naming ephemeral resources created by rally | 14:40 |
| stpierre | for instance, one of the comments on the bug opposed allowing punctuation, which is something we'd have to take into account | 14:41 |
| rloo | stpierre: so if we made it freeform, it would be fine for rally. | 14:42 |
| stpierre | depending on the definition of "freeform," yes :) | 14:42 |
| rloo | stpierre: ha ha. yeah... | 14:42 |
| *** Marga_ has joined #openstack-ironic | 14:42 | |
| rloo | hey, TheJulia, if you have a minute to comment on your comment in https://bugs.launchpad.net/ironic/+bug/1434376 | 14:43 |
| openstack | Launchpad bug 1434376 in Ironic "Node name rule relaxation" [Low,Confirmed] - Assigned to Lucas Alvares Gomes (lucasagomes) | 14:43 |
| rloo | TheJulia: did you mean that you are opposed to any punctuation in the name? | 14:43 |
| * lucasagomes removes assigneee | 14:43 | |
| *** achanda has joined #openstack-ironic | 14:45 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: WIP: Attempt to upgrade iPXE if no mac found https://review.openstack.org/193385 | 14:45 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Deprecate authenticate opt in favor of auth_strategy https://review.openstack.org/195082 | 14:46 |
| TheJulia | rloo: context switching while on conference call, one moment | 14:46 |
| *** Nisha has quit IRC | 14:47 | |
| *** Nisha has joined #openstack-ironic | 14:47 | |
| *** e0ne is now known as e0ne_ | 14:47 | |
| *** absubram has quit IRC | 14:48 | |
| *** e0ne_ is now known as e0ne | 14:48 | |
| *** Marga_ has quit IRC | 14:48 | |
| dtantsur | stpierre, I'm also strongly +1 for just dropping any checks on names, and I'm also sad to see _ disabled | 14:49 |
| dtantsur | stpierre, making less restrictive check does not break backward compatibility. there's no consensus among cores whether to bump microversion in this case.. | 14:50 |
| TheJulia | rloo: not any punctuation, anything outside what would fit in DNS standards. I SEEM to remember somebody was using name for something important and that making it free form would break them, but I just can't recall it right now. | 14:51 |
| rloo | TheJulia: why would ironic care though? I'm not sure I understand. | 14:52 |
| dtantsur | ++ people can misuse everything... | 14:52 |
| TheJulia | Ironic shouldn't care, but a consumer could easilly care | 14:52 |
| rloo | TheJulia: it is the name of the node. What they do with the name is their business? | 14:52 |
| rloo | TheJulia: so the consumer enters the name using whatever rules they want, as long as it fits into our 'freeform' | 14:53 |
| dtantsur | crazy corner cases will always happen. if someone tries to parse Node.driver_info (my favorite example) expecting fixed structure, they will fail one day | 14:53 |
| * dtantsur experiences similar case one | 14:53 | |
| rloo | TheJulia: are you saying that the consumer may expect Ironic to check that the name that they specified, conforms to eg DNS standards? | 14:54 |
| *** Marga_ has joined #openstack-ironic | 14:54 | |
| TheJulia | rloo: agree, however in my mind, a logical name shouldn't be a description. If we determine it's more of a description, then I guess its okay, but then someone could query the API with a name such as "Julia's server of \o/ win" | 14:55 |
| rloo | TheJulia: stpierre mentioned that nova has changed to allow printable unicode chars etc: https://github.com/openstack/nova/commit/f40b6a1d5c91628f78fdbc0417a3a394a69ab716 | 14:56 |
| rloo | TheJulia: Honestly, I don't care if they put a 'description' in the 'name' field. | 14:56 |
| * TheJulia looks at the change | 14:56 | |
| TheJulia | Okay, then it makes sense to remove constraints | 14:57 |
| rloo | TheJulia: thx :) | 14:57 |
| rloo | stpierre: ^^ | 14:57 |
| stpierre | that was what we decided in nova, too. if someone wants to name a server with the unicode pile of poo character, that's their problem :) | 14:57 |
| stpierre | cool, thanks | 14:57 |
| stpierre | i'll get an updated patch set proposed | 14:58 |
| * rloo just learned of the pile of poo character recently. Sigh. | 14:58 | |
| TheJulia | rloo: I reserve the right to sulk privately when somebody complains ;) | 14:58 |
| rloo | TheJulia: sure, you have that right, but really, the person complaining should be sulking | 14:58 |
| rloo | stpierre: you'll probably need to submit a patch to the spec | 14:58 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add unit test for ilo_deploy _configure_vmedia_boot() https://review.openstack.org/194047 | 14:59 |
| rloo | stpierre: https://github.com/openstack/ironic-specs/blob/master/specs/kilo/logical-names.rst | 14:59 |
| * naohirot good night all, please review it again ^^^ rloo, dtantsur | 14:59 | |
| TheJulia | rloo: true :) | 15:00 |
| *** naohirot has quit IRC | 15:00 | |
| *** Marga_ has quit IRC | 15:00 | |
| *** Marga_ has joined #openstack-ironic | 15:01 | |
| *** Marga_ has quit IRC | 15:01 | |
| * TheJulia needs to re-think bifrost's inventory node name concept slightly | 15:01 | |
| *** Marga_ has joined #openstack-ironic | 15:01 | |
| stpierre | do we need to impose a maximum length, or should that be left to the database implementation? nova kind of did both; there's an explicit maximum length of 255 characters, but mysql handles unicode strangely so there's an additional limit of something like 163 characters on mysql-backed installs | 15:02 |
| rloo | stpierre: I believe we've got some max set in the db | 15:03 |
| *** rwsu has joined #openstack-ironic | 15:04 | |
| stpierre | so does the new is_safe_hostname() (which i'll probably rename) need to enforce a max length? | 15:04 |
| NobodyCam | Good (UGT) morning Ironicers | 15:04 |
| stpierre | additionally: do we want to continue to reject uuid-like logical names? | 15:04 |
| *** r-daneel has joined #openstack-ironic | 15:05 | |
| rloo | stpierre: you're making me think of stuff i don't want to think about! :) | 15:05 |
| devananda | g'morning, all | 15:06 |
| rloo | stpierre: we increased the name length once, from 63 to 255 chars. so i think we should stick with 255 chars | 15:07 |
| *** logan2 has quit IRC | 15:07 | |
| rloo | morning devananda | 15:07 |
| stpierre | ok | 15:07 |
| dtantsur | devananda, morning | 15:07 |
| TheJulia | good morning | 15:07 |
| rloo | stpierre: am trying to remember why we're rejecting uuid-like logical names. i think it was for this backwards-compat whatever. to be safe, i'd leave it in (continue rejecting uuid-like logical names) | 15:08 |
| trown | good morning devananda | 15:08 |
| stpierre | it's in the spec | 15:08 |
| devananda | stpierre: utf8 is variable byte length, from 1 to 3 per character. MySQL's VARCHAR() field type is limited by byte length, not # of chars | 15:08 |
| stpierre | "It is recognised that a valid <logical name> could also be a valid <node uuid>, which could lead to confusion. As a consequence, logical names will be rejected as invalid if they are valid UUIDs." | 15:08 |
| NobodyCam | morning devananda rloo TheJulia stpierre dtantsur trown lucasagomes | 15:08 |
| lucasagomes | NobodyCam, devananda good ugt morning | 15:09 |
| dtantsur | NobodyCam, o/ | 15:09 |
| NobodyCam | :) | 15:09 |
| rloo | stpierre: i'd leave that in the spec and code then. can always be changed later or not. | 15:09 |
| stpierre | ok | 15:09 |
| rloo | stpierre: and yeah, devananda knows more about the db stuff than me :) | 15:09 |
| stpierre | yeah, we hashed through all of this when making the change in nova | 15:09 |
| dhellmann | could I get someone to look at https://review.openstack.org/#/c/193338/2 quickly? it's a small change that's blocking some oslo namespace removal work | 15:09 |
| trown | morning NobodyCam | 15:09 |
| rloo | stpierre: oh, so it should be easy peasy for you. deja vu etc | 15:10 |
| rloo | dhellmann: done | 15:10 |
| dhellmann | rloo: thanks! | 15:10 |
| devananda | dhellmann: done | 15:10 |
| devananda | heh | 15:10 |
| rloo | devananda: :D | 15:11 |
| dhellmann | devananda: how do you feel about an ironic_lib release after that merges? | 15:11 |
| rloo | dhellmann: ++. I'd like to get that into requirements | 15:11 |
| dhellmann | oh, there are no releases yet -- is nothing using that? | 15:12 |
| dtantsur | also requirements update https://review.openstack.org/#/c/194321/ | 15:12 |
| mark-os | lucasagomes: I got the ironic dashboard working | 15:12 |
| * rloo sets things up so she's aware of ironic-lib patches | 15:12 | |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic-specs: Allow all printable unicode and horizontal spaces in logical names https://review.openstack.org/195138 | 15:12 |
| mark-os | lucasagomes: turns out it was an authentication error | 15:12 |
| lucasagomes | mark-os, w00t, what was it? | 15:12 |
| lucasagomes | ah ok :-) good | 15:12 |
| dhellmann | rloo, devananda : it's easy enough to cut a release when you're ready to start experimenting with adoption, just let me know | 15:12 |
| mark-os | lucasagomes: I looked inside the auth.js file which was importing auth variables | 15:13 |
| openstackgerrit | Merged openstack/ironic-lib: Drop use of 'oslo' namespace package https://review.openstack.org/193338 | 15:13 |
| mark-os | lucasagomes: does 'option.username' in JS mean that the username is being imported from something called options? | 15:14 |
| * lucasagomes no idea | 15:14 | |
| mark-os | lucasagomes: I've been trying to look through this code and I've barely done any JS. Though I some how managed to get it working so I'm happy enough :) | 15:14 |
| lucasagomes | yeah I dunno JS as well neither the dashboard | 15:15 |
| *** lazy_prince has joined #openstack-ironic | 15:15 | |
| mark-os | lucasagomes: I think this is separate from Horizon because I haven't seen a component in Horizon which is built mostly in Javascript. Horizon seems to be more Django / Python based | 15:16 |
| rloo | dhellmann: it is that chicken-egg thing. it'd be nice to have a release (even w/o much in it) so we can add it to global requirements and add it to ironic requirements so that when we move stuff to ironic-lib, we can iterate quickly to get the code out of ironic before things get out of sync. | 15:16 |
| mark-os | * test | 15:16 |
| * mark-os tests this out | 15:16 | |
| lucasagomes | right yeah sounds like... I dind't know horizon supported JS | 15:16 |
| * mark-os is learning | 15:16 | |
| lucasagomes | maybe it doesn't and that's a complete different codebase | 15:17 |
| dhellmann | rloo: yep. A 0.1.0 release with something small and easy is a good first step | 15:17 |
| * lucasagomes brb in a call | 15:17 | |
| *** rwsu has quit IRC | 15:17 | |
| dhellmann | is bifrost an ironic thing? | 15:17 |
| rloo | dhellmann: yeah. TheJulia ^^ is the person! | 15:17 |
| *** rwsu has joined #openstack-ironic | 15:17 | |
| dhellmann | rloo, TheJulia : ok, I'm trying to run the unit tests and testr is reporting an error. Maybe I'm doing that wrong? I'm running "tox -e py27" | 15:17 |
| *** killer_prince has joined #openstack-ironic | 15:17 | |
| dhellmann | rloo, TheJulia : http://paste.openstack.org/show/320157/ | 15:18 |
| *** logan2 has joined #openstack-ironic | 15:18 | |
| openstackgerrit | Merged openstack/ironic-lib: Updated from global requirements https://review.openstack.org/194321 | 15:19 |
| TheJulia | dhellmann: bifrost doesn't have python tests landed yet, so it is not defined in tox.ini. It is a review pending right now. | 15:19 |
| dhellmann | TheJulia: ah, ok | 15:20 |
| *** lazy_prince has quit IRC | 15:20 | |
| rloo | g'morning NobodyCam :) | 15:20 |
| TheJulia | dhellmann: since it is very largely ansible, we have a test script to functionally exercise deployment | 15:20 |
| *** ijw_ has quit IRC | 15:20 | |
| openstackgerrit | Doug Hellmann proposed openstack/bifrost: Drop use of 'oslo' namespace package https://review.openstack.org/195143 | 15:20 |
| dhellmann | TheJulia: ^^ | 15:21 |
| dhellmann | if the tests aren't expected to run, that's fine, I'll ignore the failure locally for now | 15:21 |
| *** ijw_ has joined #openstack-ironic | 15:21 | |
| TheJulia | dhellmann: the change for the test for that is still in review. I'll test locally and check my oustanding reviews. Thanks! | 15:22 |
| dhellmann | TheJulia: sounds good -- feel free to take that over and rebase or whatever you need to make it land cleanly | 15:22 |
| * TheJulia is sure it will work | 15:22 | |
| TheJulia | dhellmann: will do, thanks! | 15:22 |
| dtantsur | folks, devananda, can I get your opinion on https://review.openstack.org/#/c/194619/ please? It gives us some gray hair now... | 15:24 |
| rloo | dtantsur: I thought you meant the patch was giving gray hairs, but you mean the patch will reduce the gray hairs? :) | 15:25 |
| devananda | dtantsur: it's inspecting the call stack ...that's never a good sign | 15:25 |
| dtantsur | devananda, that's why I ask. but I see no more sane options to make it debuggable | 15:26 |
| *** ijw_ has quit IRC | 15:26 | |
| dtantsur | devananda, well, we can extend task_manager.acquire with "purpose" argument | 15:26 |
| dtantsur | and then maybe even store this "purpose" in database | 15:26 |
| dtantsur | but that's gonna be a huge patch | 15:26 |
| devananda | dtantsur: so what is the problem this is solving? | 15:27 |
| devananda | dtantsur: I can imagine some things ... but want to know what the issue at hand actually is | 15:27 |
| *** achanda_ has joined #openstack-ironic | 15:27 | |
| dtantsur | devananda, we get numerous failures due to "node locked error". while I can increase retry in ironicclient, I want to know what is causing nodes to be locked for > 30 secs | 15:28 |
| dtantsur | when nothing is seemingly going on on them (i.e. no deploy) | 15:28 |
| dtantsur | what I want to is to be able to tell "oh, this was power sync for a node, and it took 55 sec" for example | 15:29 |
| devananda | gotcha | 15:29 |
| devananda | I dont think this is the right approach to get that answer, but I do think that's a good thing to log :) | 15:29 |
| *** achanda has quit IRC | 15:29 | |
| dtantsur | devananda, any better ideas then? :) I don't like it either tbh | 15:30 |
| devananda | dtantsur: i'll comment on review so it's more clear, one minute | 15:30 |
| *** ijw_ has joined #openstack-ironic | 15:32 | |
| dtantsur | thanks! | 15:32 |
| *** achanda_ has quit IRC | 15:32 | |
| *** ifarkas has quit IRC | 15:32 | |
| *** Nisha has quit IRC | 15:33 | |
| openstackgerrit | Matt Keenan proposed openstack/ironic: get_supported_boot_devices() returns static device list https://review.openstack.org/188466 | 15:34 |
| dtantsur | ok folks, I'll be going now, see you tomorrow! | 15:35 |
| rloo | bye dtantsur | 15:35 |
| *** dtantsur is now known as dtantsur|afk | 15:35 | |
| NobodyCam | night dtantsur|afk | 15:36 |
| *** logan2 has quit IRC | 15:39 | |
| *** e0ne is now known as e0ne_ | 15:40 | |
| *** viktors is now known as viktors|afk | 15:41 | |
| rloo | hey, anyone have any thoughts wrt updating existing/approved specs? eg https://review.openstack.org/#/c/195138/ | 15:42 |
| JayF | lucasagomes: I'd be +1 to a configuration to change 2.0/1.5. I'd be very -1 to changing the default. | 15:43 |
| rloo | I suggested that (cuz the bug suggested that the spec be updated). But... do we care that this spec is in kilo but we're changing it in Liberty | 15:43 |
| lucasagomes | JayF, ++ oh yeah totally 2.0 will be the default | 15:43 |
| lucasagomes | JayF, I'm working on a patch to add ipmi_protocol_version in the node's driver_info | 15:44 |
| lucasagomes | so one can switch between it | 15:44 |
| lucasagomes | (default to 2.0) | 15:44 |
| *** e0ne_ is now known as e0ne | 15:45 | |
| NobodyCam | mornign JayF :) | 15:45 |
| TheJulia | good morning JayF | 15:46 |
| devananda | rloo: replied | 15:48 |
| rloo | devananda: ok thx. Good to know. | 15:49 |
| huats | My understanding of the filters is that I can chosse between baremetal and nova filters, but once I have choosen the baremetal ones, even the "classical" instances will be using the baremetal filters. Is it correct ? Won't it create any issue for using another nova driver (like kvm) ? | 15:54 |
| *** athomas has quit IRC | 15:54 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add IPMI 1.5 support for the ipmitool power driver https://review.openstack.org/195157 | 15:55 |
| openstackgerrit | Merged openstack/bifrost: Make git URLs consistent with each other https://review.openstack.org/194881 | 16:01 |
| *** mgoddard1 has joined #openstack-ironic | 16:02 | |
| devananda | huats: you can use host aggregates to separate the nova-compute hosts by hypervisor, for example | 16:03 |
| *** mgoddard has quit IRC | 16:04 | |
| lucasagomes | o wow api getting a subset of attributes for ports and chassis was merged :-) | 16:05 |
| lucasagomes | w00t | 16:05 |
| rloo | lucasagomes: this is your lucky day :) | 16:06 |
| lucasagomes | yeah feels like :-) | 16:06 |
| lucasagomes | rloo, thanks for the reviews | 16:06 |
| rloo | I updated our wiki to reflect 'changes to existing specs': https://wiki.openstack.org/wiki/Ironic/Specs_Process | 16:06 |
| * lucasagomes works on the client patches | 16:06 | |
| rloo | lucasagomes: yw. and thx for doing the client side too :) | 16:06 |
| devananda | rloo: thx | 16:06 |
| lucasagomes | rloo, yeah will start that heh there's no code yet | 16:07 |
| rloo | devananda: yw. thx for the answer :) | 16:07 |
| huats | devananda: thanks ! | 16:07 |
| lucasagomes | gotta change nova after that as well to make polling lighter | 16:07 |
| lucasagomes | " changing a previously approved and completed spec should only be done for cosmetic cleanup or fixing errors" ++ | 16:08 |
| openstackgerrit | Merged openstack/bifrost: Adding agent_ucs driver to the driver list https://review.openstack.org/191219 | 16:08 |
| openstackgerrit | Merged openstack/bifrost: Drop use of 'oslo' namespace package https://review.openstack.org/195143 | 16:12 |
| *** Nisha has joined #openstack-ironic | 16:13 | |
| *** davideagnello has quit IRC | 16:16 | |
| *** Nisha has quit IRC | 16:19 | |
| lucasagomes | folks if you guys have some free time, please take a look at https://review.openstack.org/#/c/194132/ (it's small) | 16:22 |
| NobodyCam | lucasagomes: helding ??? | 16:24 |
| lucasagomes | holding :-( | 16:25 |
| * lucasagomes fix commit message | 16:25 | |
| NobodyCam | lol | 16:25 |
| *** jcoufal has quit IRC | 16:26 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Clean nodes stuck in DEPLOYING state when ir-cond restarts https://review.openstack.org/194132 | 16:26 |
| NobodyCam | lgtm lucasagomes :) | 16:29 |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic-specs: Allow all printable unicode and horizontal spaces in logical names https://review.openstack.org/195138 | 16:29 |
| lucasagomes | NobodyCam, w00t :-) thanks | 16:29 |
| *** Nisha has joined #openstack-ironic | 16:30 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add a dynamic prepare for test role https://review.openstack.org/192001 | 16:30 |
| *** lazy_prince has joined #openstack-ironic | 16:33 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Addition of shade support to bifrost_inventory.py https://review.openstack.org/194684 | 16:34 |
| *** killer_prince has quit IRC | 16:35 | |
| *** ijw_ has quit IRC | 16:36 | |
| *** lazy_prince has quit IRC | 16:37 | |
| *** ijw_ has joined #openstack-ironic | 16:37 | |
| *** mgoddard1 has quit IRC | 16:37 | |
| *** mgoddard has joined #openstack-ironic | 16:38 | |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic-specs: Allow all printable unicode and horizontal spaces in logical names https://review.openstack.org/195138 | 16:40 |
| *** trown is now known as trown|lunch | 16:40 | |
| *** ijw_ has quit IRC | 16:41 | |
| *** Shrews has quit IRC | 16:42 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add iBoot driver documentation https://review.openstack.org/193546 | 16:43 |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/194320 | 16:43 |
| *** ijw_ has joined #openstack-ironic | 16:44 | |
| *** Shrews has joined #openstack-ironic | 16:44 | |
| *** jgrimm is now known as zz_jgrimm | 16:47 | |
| *** Sukhdev has joined #openstack-ironic | 16:51 | |
| *** Nisha has quit IRC | 16:51 | |
| *** Nisha has joined #openstack-ironic | 16:51 | |
| *** achanda has joined #openstack-ironic | 16:52 | |
| *** dontalton has joined #openstack-ironic | 16:52 | |
| lucasagomes | folks I'm going to call it a day | 16:53 |
| *** lsmola has quit IRC | 16:53 | |
| lucasagomes | have a great evening everyone | 16:55 |
| *** lucasagomes is now known as lucas-dinner | 16:55 | |
| *** derekh has quit IRC | 16:55 | |
| NobodyCam | have a good night lucas-dinner | 16:57 |
| NobodyCam | :) | 16:57 |
| *** ijw__ has joined #openstack-ironic | 16:57 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add example deploy all available playbook https://review.openstack.org/195100 | 16:59 |
| *** ijw_ has quit IRC | 17:00 | |
| *** achanda has quit IRC | 17:03 | |
| *** davideagnello has joined #openstack-ironic | 17:06 | |
| sambetts | night lucas-dinner | 17:06 |
| sambetts | I'm also going to call it an evening, good night ironic o/ | 17:07 |
| TheJulia | goodnight sambetts | 17:07 |
| sambetts | night TheJulia | 17:07 |
| NobodyCam | night sambetts | 17:09 |
| NobodyCam | :) | 17:09 |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic: Permit logical names to be freeform strings https://review.openstack.org/193587 | 17:11 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Remove un-necessary checks from the dynamic test https://review.openstack.org/193687 | 17:12 |
| devananda | stpierre: have a link to the nova change re: names? | 17:14 |
| stpierre | https://github.com/openstack/nova/commit/f40b6a1d5c91628f78fdbc0417a3a394a69ab716 | 17:15 |
| stpierre | it's also in the spec | 17:15 |
| *** e0ne has quit IRC | 17:16 | |
| devananda | stpierre: http://specs.openstack.org/openstack/nova-specs/specs/kilo/implemented/relax-resource-name-restrictions.html?highlight=relax#proposed-change | 17:17 |
| devananda | "Fields that have legitimate reasons to be restricted can still impose further restrictions. For instance, hostnames can be restricted per RFC 1178, which is a strict subset of the restrictions described above" | 17:17 |
| stpierre | in nova, the same regex was used to validate *lots* of different kinds of resource names | 17:17 |
| devananda | stpierre: the nova spec references "resource names" such as flavors, images, etc | 17:18 |
| stpierre | yeah | 17:18 |
| devananda | host names are explicitly called out as being appropriate for more restrictions | 17:18 |
| *** pelix has quit IRC | 17:20 | |
| devananda | stpierre: can you explain why node names should not be treated like a hostname? | 17:24 |
| krotscheck | devananda: So, sean was asking about a blueprint for the webclient. I'm thinking... probably not a blueprint, because blueprints are horrible, but perhaps a spec? | 17:30 |
| stpierre | devananda: no, i can't. my initial review was to do exactly that -- fix is_hostname_safe() to properly check hostnames. rloo suggested, per bug 1434376, opening it up to freeform strings | 17:31 |
| openstack | bug 1434376 in Ironic "Node name rule relaxation" [Low,In progress] https://launchpad.net/bugs/1434376 - Assigned to Chris St. Pierre (stpierre) | 17:31 |
| devananda | krotscheck: honestly, a spec is there for folks to discuss/approve a design of some change to a project | 17:31 |
| devananda | krotscheck: you've started a new project. it doesn't need a spec | 17:32 |
| devananda | krotscheck: but things you choose to do (or others propose) within that may warrant specs | 17:32 |
| krotscheck | Allllrighteythen | 17:32 |
| devananda | it'd be like asking someone to create a spec for the creation of ironic-inspector | 17:32 |
| stpierre | i don't even use ironic, so i don't have a horse in this race, i'm just trying to get the current (broken, excessively restriction) regex fixed so that rally can implement a ephemeral resource naming scheme that works across all openstack projects | 17:32 |
| krotscheck | I'm going to create a spec for scotch. Because scotch drives all good code. | 17:33 |
| *** mitchjameson has joined #openstack-ironic | 17:33 | |
| devananda | stpierre: ahh ok. I thought you were driving the change | 17:33 |
| rloo | devananda, stpierre: yeah, it was me. cuz I knew there was that bug open for relaxing the name. I personally don't care. | 17:33 |
| stpierre | not really. we (rally) just found the broken regex and wanted to fix it for our own devious ends :) | 17:33 |
| devananda | krotscheck: OTOH, it is quite possible that some folks don't understand what you're trying to do, and what htey really want is just to understand the architecture/design/goals | 17:33 |
| rloo | devananda, stpierre: I just didn't want stpierre to go to all that work with the patch if we wanted freeform. | 17:34 |
| devananda | krotscheck: whether you communicate that via a spec or via docs is... meh. which ever works for you | 17:35 |
| rloo | devananda: I think when logical names code was being reviewed, people were asking why the restrictions to the logical name. | 17:35 |
| krotscheck | devananda: See, I don't really know where a large-scale vision discussion for javascript in openstack should go. Right now I'm still just trying to get the pieces in place that would permit us to create the projects. | 17:36 |
| devananda | krotscheck: you're welcome to do a spec if you want to. i woul imagine it will be easy to approve since I've already approved your approach AND it requires almost no changes in Ironic (just add an optional middleware) | 17:36 |
| devananda | krotscheck: exactly! that large scale discussion is definitely not in an ironic spec | 17:36 |
| krotscheck | I'm guessing x-project specs. | 17:36 |
| devananda | yup | 17:36 |
| krotscheck | Hrm. | 17:36 |
| krotscheck | Actually, I should probably do a javascript infra spec. | 17:36 |
| krotscheck | And the necessary governance update. | 17:36 |
| devananda | krotscheck: and I do not want someone asking you to create an ironic spec to slow you down when what you're doing doesn't NEED a design discussion within Ironic itself | 17:37 |
| *** dguerri is now known as dguerri` | 17:37 | |
| devananda | your project ought to have some doc somewhere, though. so that others can try it out // contribute to it | 17:37 |
| *** openstackgerrit has quit IRC | 17:38 | |
| *** openstackgerrit has joined #openstack-ironic | 17:38 | |
| krotscheck | devananda: Ugh. I need to spend a week just writing documentation for all this stuff. | 17:39 |
| devananda | rloo: stpierre: if neither of you care, why are either of you working on this? I'm confused | 17:39 |
| krotscheck | maybe a cycle. | 17:39 |
| devananda | krotscheck: hah | 17:39 |
| krotscheck | Because srsly. | 17:39 |
| devananda | krotscheck: or just make something cool that works well and ignore the "but wait! we need moar process!" folks | 17:39 |
| krotscheck | I think I'm too german for that. | 17:39 |
| devananda | :) | 17:40 |
| krotscheck | Also, I committed to writing CORS docs in the TC Spec. | 17:40 |
| rloo | devananda: cuz stpierre proposed a fix to the existing validation of the logical (hostname) code. which might entail a microversion bump. so if we want freeform, why not do that instead. | 17:40 |
| rloo | devananda: and stpierre wants *something* fixed cuz of rally. | 17:40 |
| rloo | devananda: there's lots of patches where I don't care, but I may review them anyway :) | 17:41 |
| devananda | what's actually broken? | 17:41 |
| devananda | stpierre: ^ ? | 17:42 |
| stpierre | devananda: https://review.openstack.org/#/c/193587/1//COMMIT_MSG | 17:42 |
| stpierre | oh, and the max length is wrong, that didn't get into the commit message | 17:42 |
| stpierre | ah, PS2 has it: https://review.openstack.org/#/c/193587/2//COMMIT_MSG | 17:44 |
| devananda | oh - i didn't see anything besides rev3 | 17:44 |
| krotscheck | Incidentally, devananda, the guys in the grommet project have a "Medium example app" which, surprise surprise, allows someone to manage dummy racks, servers, and server configurations. | 17:45 |
| krotscheck | Link: http://grommet.io/medium-app/login (any password and login will work) | 17:45 |
| *** mgarza_ has joined #openstack-ironic | 17:45 | |
| devananda | stpierre: yea, so rev2 looks quite agreeable to me (other folks comments notwithstanding) | 17:48 |
| *** mgoddard has left #openstack-ironic | 17:50 | |
| devananda | stpierre: another thing to keep in mind -- logical names are included in resource URI's, so they must be encodable in an HTTP URL | 17:51 |
| devananda | eg, https://tools.ietf.org/html/rfc3986#section-2.2 | 17:54 |
| *** amotoki has quit IRC | 17:55 | |
| *** achanda has joined #openstack-ironic | 17:56 | |
| *** logan2 has joined #openstack-ironic | 18:00 | |
| *** harlowja has quit IRC | 18:00 | |
| *** e0ne has joined #openstack-ironic | 18:01 | |
| *** ukalifon1 has quit IRC | 18:02 | |
| *** romcheg has quit IRC | 18:03 | |
| *** harlowja has joined #openstack-ironic | 18:04 | |
| *** trown|lunch is now known as trown | 18:07 | |
| stpierre | so now i've got conflicting recommendations from two cores. can you guys come to an agreement and then i'll be happy to implement what you decide on -- i've basically got it written already, whatever it is :) | 18:07 |
| *** e0ne has quit IRC | 18:08 | |
| rloo | stpierre: devananda is bigger than me. go with him :) Esp since he isn't in agreement with the freeform stuff. | 18:09 |
| stpierre | ok | 18:09 |
| stpierre | sounds good | 18:09 |
| rloo | stpierre: sorry that i led you astray. I was hoping to minimize work. | 18:10 |
| devananda | stpierre: apologies for the back and forth on that. don't mean to put you in the middle -- dtantsur|afk and I should talk about this :) | 18:10 |
| stpierre | rloo: well isn't that.... ironic | 18:11 |
| stpierre | http://i.ytimg.com/vi/6YMPAH67f4o/hqdefault.jpg | 18:11 |
| rloo | stpierre: :D | 18:11 |
| stpierre | honestly, Nova would have taken eight months to figure this out so getting it all taken care of in a morning really isn't so bad | 18:12 |
| stpierre | so, in PS3 i took the approach of putting the microversion abstractions in the API libraries that call is_hostname_safe() -- v1/types.py and v1/utils.py. it'd be easier to just put that abstraction directly in ironic.common.utils, but i wasn't sure how kosher it would be to import pecan and check the API microversion in there, since it's not explicitly an API library | 18:13 |
| stpierre | thoughts? | 18:13 |
| rloo | stpierre: in a meeting and i haven't actually looked at your patch yet, only read the commit msg :). Later this aft... | 18:14 |
| devananda | stpierre: https://review.openstack.org/#/c/193587/3/ironic/api/controllers/v1/utils.py,cm looks good to me. I would not put that in ironic.common.utils because, as you say, it's a general utils area, not api specific | 18:18 |
| *** boris-42 has joined #openstack-ironic | 18:18 | |
| *** mmorais has quit IRC | 18:21 | |
| *** achanda has quit IRC | 18:27 | |
| *** achanda has joined #openstack-ironic | 18:28 | |
| *** dguerri` is now known as dguerri | 18:30 | |
| *** dguerri is now known as dguerri` | 18:32 | |
| *** achanda has quit IRC | 18:33 | |
| *** ukalifon1 has joined #openstack-ironic | 18:36 | |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic: Improve is_hostname_safe check https://review.openstack.org/193587 | 18:37 |
| *** achanda has joined #openstack-ironic | 18:40 | |
| *** mmorais has joined #openstack-ironic | 18:40 | |
| *** logan2 has quit IRC | 18:46 | |
| *** absubram has joined #openstack-ironic | 18:49 | |
| *** Sukhdev has quit IRC | 18:51 | |
| *** amotoki has joined #openstack-ironic | 18:56 | |
| *** amotoki has quit IRC | 19:01 | |
| *** harlowja has quit IRC | 19:15 | |
| *** harlowja has joined #openstack-ironic | 19:15 | |
| *** mgarza_ has quit IRC | 19:17 | |
| *** Marga_ has quit IRC | 19:21 | |
| *** achanda has quit IRC | 19:23 | |
| *** mgarza_ has joined #openstack-ironic | 19:24 | |
| *** alexpilotti has quit IRC | 19:26 | |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic: Improve is_hostname_safe check https://review.openstack.org/193587 | 19:28 |
| *** Marga_ has joined #openstack-ironic | 19:31 | |
| *** Sukhdev has joined #openstack-ironic | 19:33 | |
| *** e0ne has joined #openstack-ironic | 19:33 | |
| *** Marga_ has quit IRC | 19:34 | |
| *** Marga_ has joined #openstack-ironic | 19:34 | |
| *** Nisha has quit IRC | 19:37 | |
| *** ukalifon1 has quit IRC | 19:37 | |
| *** dguerri` is now known as dguerri | 19:41 | |
| *** dguerri is now known as dguerri` | 19:44 | |
| *** Marga_ has quit IRC | 19:46 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Remove un-necessary checks from the dynamic test https://review.openstack.org/193687 | 19:47 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add dynamic node unprovision role https://review.openstack.org/193688 | 19:47 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Correct ipv4 networking config issue in dynamic path https://review.openstack.org/193689 | 19:47 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add delete role for dynamic inventory testing https://review.openstack.org/193690 | 19:48 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Individual playbook for creating a test VM https://review.openstack.org/193691 | 19:48 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Update the dynamic test playbook https://review.openstack.org/193692 | 19:49 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add dynamic inventory description to README https://review.openstack.org/193709 | 19:49 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Addition of dynamic enrollment and deployment playbooks https://review.openstack.org/193737 | 19:49 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Change Bifrost over to using dynamic inventory https://review.openstack.org/193885 | 19:49 |
| *** romcheg has joined #openstack-ironic | 19:50 | |
| *** dguerri` is now known as dguerri | 19:52 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add dynamic node unprovision role https://review.openstack.org/193688 | 19:55 |
| *** Marga_ has joined #openstack-ironic | 19:55 | |
| *** amotoki has joined #openstack-ironic | 19:57 | |
| *** dguerri is now known as dguerri` | 19:59 | |
| *** Marga_ has quit IRC | 20:00 | |
| *** amotoki has quit IRC | 20:01 | |
| *** Marga_ has joined #openstack-ironic | 20:02 | |
| *** e0ne has quit IRC | 20:06 | |
| *** natorious is now known as zz_natorious | 20:08 | |
| *** ijw_ has joined #openstack-ironic | 20:11 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Correct ipv4 networking config issue in dynamic path https://review.openstack.org/193689 | 20:12 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add delete role for dynamic inventory testing https://review.openstack.org/193690 | 20:12 |
| *** ijw__ has quit IRC | 20:14 | |
| *** harlowja has quit IRC | 20:18 | |
| *** harlowja has joined #openstack-ironic | 20:18 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Individual playbook for creating a test VM https://review.openstack.org/193691 | 20:30 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Update the dynamic test playbook https://review.openstack.org/193692 | 20:30 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add dynamic inventory description to README https://review.openstack.org/193709 | 20:37 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Addition of dynamic enrollment and deployment playbooks https://review.openstack.org/193737 | 20:38 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Change Bifrost over to using dynamic inventory https://review.openstack.org/193885 | 20:38 |
| rloo | devananda: when you get a chance, please unblock this https://review.openstack.org/#/c/193587/ | 20:44 |
| rloo | devananda: unless I misunderstood, but no spec related to it any more | 20:44 |
| rloo | devananda: by the way, i see quite a few patches submitted w/o any bug associated with them. do you have guidelines as to when a bug should be opened? I keep thinking that maybe there is no need cuz no one else seems to be asking. | 20:46 |
| *** dprince has quit IRC | 20:51 | |
| *** ndipanov has quit IRC | 20:59 | |
| *** vishwana_ has quit IRC | 21:17 | |
| *** vishwanathj has joined #openstack-ironic | 21:18 | |
| *** trown is now known as trown|outttypeww | 21:22 | |
| *** morganfainberg is now known as ayspryn | 21:24 | |
| *** ayspryn is now known as morganfainberg | 21:24 | |
| *** hblixt has joined #openstack-ironic | 21:25 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Follow-up fix for H306 in bifrost_inventory.py https://review.openstack.org/195324 | 21:26 |
| *** alexpilotti has joined #openstack-ironic | 21:30 | |
| openstackgerrit | Chris St. Pierre proposed openstack/ironic: Improve is_hostname_safe check https://review.openstack.org/193587 | 21:30 |
| *** zz_natorious is now known as natorious | 21:31 | |
| *** natorious is now known as zz_natorious | 21:31 | |
| *** Sukhdev has quit IRC | 21:33 | |
| devananda | rloo: use your judgement -- if the patch is clearly fixing something, it probably should have a bug report to track that in release notes | 21:34 |
| devananda | rloo: if it's part of ongoing work we're all doing, either there's a spec or there isn't. | 21:35 |
| rloo | devananda: Thx, that's what I thought. Wonder why others don't think the same way... | 21:39 |
| devananda | rloo: any particular cases in mind? | 21:39 |
| *** alexpilotti has quit IRC | 21:40 | |
| rloo | devananda: well, for instance, that hostname thing. No one else asked for a bug. And it has happened before. So thought I'd get a reality check. | 21:40 |
| devananda | huh. thanks | 21:40 |
| *** stpierre has quit IRC | 21:41 | |
| *** alexpilotti has joined #openstack-ironic | 21:42 | |
| *** dguerri` is now known as dguerri | 21:44 | |
| *** zz_natorious is now known as natorious | 21:46 | |
| mrda | Morning | 21:46 |
| mrda | rloo, dtantsur: Yes, nodes with logical names containing underscores (+ a few other conditions) would need to be renamed via a 'ironic node-update <uuid> update name=<new name>' first | 21:48 |
| mrda | I mentioned that in the review comments | 21:48 |
| mrda | (and rloo, thanks for the review) | 21:49 |
| rloo | morning mrda! (and yw) | 21:49 |
| TheJulia | good morning mrda | 21:49 |
| mrda | Hi rloo and TheJulia ;) | 21:50 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Update cached images based on update time https://review.openstack.org/195017 | 21:50 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Update cached images based on update time https://review.openstack.org/195017 | 21:52 |
| *** lucas-dinner has quit IRC | 21:53 | |
| *** dguerri is now known as dguerri` | 21:58 | |
| *** amotoki has joined #openstack-ironic | 21:58 | |
| devananda | heading out for the afternoon spate of errands ... bbl | 22:00 |
| *** amotoki has quit IRC | 22:04 | |
| NobodyCam | enjoy devananda | 22:05 |
| *** hblixt has quit IRC | 22:15 | |
| *** alexpilotti has quit IRC | 22:19 | |
| *** ijw_ has quit IRC | 22:22 | |
| *** Sukhdev has joined #openstack-ironic | 22:23 | |
| *** ijw_ has joined #openstack-ironic | 22:23 | |
| *** radek__ has quit IRC | 22:23 | |
| *** ThomasPB has quit IRC | 22:25 | |
| *** mgoddard has joined #openstack-ironic | 22:27 | |
| *** mgoddard has left #openstack-ironic | 22:27 | |
| *** ijw_ has quit IRC | 22:29 | |
| JayF | devananda: I talked to mariojv about that barbican spec; I think there's value there but it wasn't well communicated in the first drafts. Would love to have the conversation about it with you. | 22:34 |
| mariojv | +1 ^ me too | 22:35 |
| JayF | devananda: this is one of those things we're likely doing with or without upstream, and I'd love to have it upstreamed | 22:35 |
| *** ijw_ has joined #openstack-ironic | 22:38 | |
| *** romcheg has quit IRC | 22:53 | |
| *** romcheg has joined #openstack-ironic | 22:53 | |
| *** mgarza_ has quit IRC | 22:57 | |
| *** romcheg has quit IRC | 22:57 | |
| *** amotoki has joined #openstack-ironic | 22:59 | |
| *** amotoki has quit IRC | 23:04 | |
| *** hblixt has joined #openstack-ironic | 23:05 | |
| *** rloo has quit IRC | 23:05 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Add the rest of the cookiecutter-generated files for testing https://review.openstack.org/194893 | 23:09 |
| *** blixt_ has joined #openstack-ironic | 23:11 | |
| *** hblixt has quit IRC | 23:15 | |
| *** ijw__ has joined #openstack-ironic | 23:22 | |
| *** ijw_ has quit IRC | 23:25 | |
| openstackgerrit | Laura Moore proposed openstack/ironic-specs: Update of the Ironic ML2 Integration spec https://review.openstack.org/188528 | 23:26 |
| *** bradjones has quit IRC | 23:26 | |
| *** bradjones has joined #openstack-ironic | 23:26 | |
| *** bradjones has quit IRC | 23:26 | |
| *** bradjones has joined #openstack-ironic | 23:26 | |
| *** Marga_ has quit IRC | 23:30 | |
| *** Sukhdev has quit IRC | 23:40 | |
| *** mtanino has joined #openstack-ironic | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
