| *** lucas-dinner has quit IRC | 00:01 | |
| *** marcoemorais has quit IRC | 00:02 | |
| devananda | NobodyCam: https://bugs.launchpad.net/ironic/+bug/1373671 | 00:05 |
|---|---|---|
| devananda | have I said lately that I hate that we have that driver? | 00:05 |
| devananda | I know you wrote it, and it's been really helpful in all of our ability to develop and test this project .... | 00:06 |
| NobodyCam | lol :( | 00:06 |
| devananda | but it also makes me sad when I'm trynig to test translations and that happens | 00:06 |
| jroll | devananda: there's a patch up to add a 'default_locale' arg to execute() | 00:06 |
| jroll | which would make fixing this super easy | 00:06 |
| devananda | jroll: sure, we'd need to put translated versions of that grep inline, which is not really good either | 00:07 |
| jroll | oh. &^*(&*&^ | 00:07 |
| devananda | jroll: read the bug :) | 00:07 |
| jroll | devananda: that makes me wonder about ipmitool | 00:07 |
| jroll | and the version check | 00:07 |
| devananda | yep | 00:07 |
| jroll | at startup | 00:07 |
| * devananda tries it | 00:07 | |
| jroll | so many expletives. | 00:07 |
| jroll | that said, I don't expect any deployers to run the ssh driver :P | 00:08 |
| devananda | jroll: no, it's fine | 00:08 |
| jroll | ipmitool is? | 00:09 |
| devananda | jroll: the ipmitool option support code checks for support by actually running it and catching an error | 00:09 |
| jroll | neat | 00:09 |
| jroll | aha | 00:09 |
| jroll | cool | 00:09 |
| *** marcoemorais has joined #openstack-ironic | 00:10 | |
| *** openstackgerrit has quit IRC | 00:16 | |
| *** openstackgerrit has joined #openstack-ironic | 00:18 | |
| *** todd_dsm has joined #openstack-ironic | 00:18 | |
| *** Haomeng|2 has joined #openstack-ironic | 00:26 | |
| *** Haomeng has quit IRC | 00:27 | |
| *** rloo has quit IRC | 00:33 | |
| *** todd_dsm has quit IRC | 00:39 | |
| *** ChuckC has quit IRC | 00:49 | |
| *** harlowja_ has quit IRC | 00:51 | |
| *** marcoemorais has quit IRC | 01:02 | |
| devananda | adam_g: https://bugs.launchpad.net/nova/+bug/1310135 -- see long post at the end | 01:08 |
| lifeless | must read? | 01:09 |
| devananda | lifeless: this one affects you guys too, I bet | 01:09 |
| devananda | workign on a fix now, though I should be taking a break | 01:09 |
| lifeless | ohh yes I can imagine so | 01:10 |
| *** marcoemorais has joined #openstack-ironic | 01:15 | |
| *** harlowja has joined #openstack-ironic | 01:18 | |
| *** marcoemorais has quit IRC | 01:19 | |
| *** todd_dsm has joined #openstack-ironic | 01:25 | |
| *** todd_dsm has quit IRC | 01:39 | |
| *** todd_dsm has joined #openstack-ironic | 01:40 | |
| *** nosnos has joined #openstack-ironic | 01:45 | |
| *** todd_dsm has quit IRC | 01:45 | |
| *** todd_dsm has joined #openstack-ironic | 01:50 | |
| adam_g | devananda, yeah, i think something like that would work. tho i worry about races, and im not sure a reboot ever sets target_power_state='reboot'. | 01:56 |
| *** todd_dsm has quit IRC | 01:56 | |
| adam_g | when we're in paris we should work with nova and come up with a good way to make the neutron/nova callback stuff generally consumable elsewhere | 01:57 |
| adam_g | i know we have use cases for it in other parts of ironic, and i heard mumblings on the list about cinder wanting to use something like it as well | 01:58 |
| lifeless | webhooks forever | 02:00 |
| lifeless | or websockets would fit some cases | 02:00 |
| *** annegentle has quit IRC | 02:25 | |
| *** annegentle has joined #openstack-ironic | 02:27 | |
| *** Poornima has joined #openstack-ironic | 02:33 | |
| *** todd_dsm has joined #openstack-ironic | 02:47 | |
| *** marcoemorais has joined #openstack-ironic | 02:48 | |
| *** marcoemorais1 has joined #openstack-ironic | 02:50 | |
| *** marcoemorais has quit IRC | 02:53 | |
| *** todd_dsm has quit IRC | 02:59 | |
| devananda | adam_g: tested it - ironic sets target power state = "power on" | 03:10 |
| *** Haomeng|2 has quit IRC | 03:14 | |
| *** ramineni has joined #openstack-ironic | 03:15 | |
| *** pcrews has quit IRC | 03:22 | |
| openstackgerrit | A change was merged to openstack/ironic: Do not return 'id' in REST API error messages https://review.openstack.org/123150 | 03:33 |
| *** todd_dsm has joined #openstack-ironic | 03:35 | |
| *** todd_dsm has quit IRC | 03:41 | |
| *** todd_dsm has joined #openstack-ironic | 03:47 | |
| *** coolsvap is now known as coolsvap|gone | 03:58 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/123662 | 04:06 |
| *** lynxman has quit IRC | 04:10 | |
| *** lynxman has joined #openstack-ironic | 04:10 | |
| *** ChuckC_ has joined #openstack-ironic | 04:11 | |
| *** todd_dsm has quit IRC | 04:20 | |
| *** todd_dsm has joined #openstack-ironic | 04:22 | |
| *** todd_dsm has quit IRC | 04:24 | |
| *** todd_dsm has joined #openstack-ironic | 04:33 | |
| *** nikunj2512 has joined #openstack-ironic | 04:34 | |
| *** killer_prince is now known as lazy_prince | 04:37 | |
| *** vinbs has joined #openstack-ironic | 04:40 | |
| *** Poornima has quit IRC | 04:50 | |
| *** todd_dsm has quit IRC | 04:54 | |
| *** lsmola has quit IRC | 04:57 | |
| *** rakesh_hs has joined #openstack-ironic | 05:01 | |
| *** Poornima has joined #openstack-ironic | 05:07 | |
| *** lsmola has joined #openstack-ironic | 05:11 | |
| *** todd_dsm has joined #openstack-ironic | 05:14 | |
| *** coolsvap|gone is now known as coolsvap | 05:26 | |
| *** todd_dsm has quit IRC | 05:31 | |
| *** Ng has quit IRC | 05:47 | |
| *** Ng has joined #openstack-ironic | 05:48 | |
| *** victor_lowther__ has quit IRC | 05:48 | |
| *** victor_lowther__ has joined #openstack-ironic | 05:49 | |
| *** harlowja is now known as harlowja_away | 05:53 | |
| *** pensu has joined #openstack-ironic | 06:07 | |
| *** Poornima has quit IRC | 06:22 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/123342 | 06:25 |
| *** k4n0 has joined #openstack-ironic | 06:37 | |
| *** dtantsur|afk is now known as dtantsur | 06:56 | |
| dtantsur | Morning Ironic | 06:56 |
| *** todd_dsm has joined #openstack-ironic | 06:58 | |
| *** vigneshvar has joined #openstack-ironic | 06:59 | |
| vigneshvar | Hi guys need help | 06:59 |
| vigneshvar | ERROR ironic.drivers.modules.agent [-] node a020de10-3233-4d43-8259-2b97665f609e command status errored: {u'message': u'Error downloading image.', u'code': 500, u'type': u'ImageDownloadError', u'details': u'Could not download image with id b481cf87-72fe-4760-9316-d36a9f7243dd.'} | 06:59 |
| vigneshvar | Glance has no error | 06:59 |
| vigneshvar | where could i have gone wrong. Any workaround | 07:01 |
| lazy_prince | vigneshvar: is your glance api able to receive request from ironic..? | 07:01 |
| vigneshvar | lazy_prince: Yes , i can see the request there | 07:02 |
| lazy_prince | vigneshvar: now can you see any issues in your glance service..? how big is the image..? | 07:02 |
| vigneshvar | lazy_prince: In glance services i dont see any error. I am just using the basic cirros image, to test with | 07:03 |
| lazy_prince | vigneshvar: aha.. and you are using it with baremetal or VM as baremetal..? | 07:04 |
| vigneshvar | VM as baremetal. As per the document. I want to give it a try before i deploy on my real server | 07:05 |
| lazy_prince | vigneshvar: i am not sure if those images can be used with baremetal i have not tried it.. also, do you have the deploy images uploaded to glance..? | 07:05 |
| vigneshvar | lazy_prince: I have deploy images. I followed devstack | 07:06 |
| lazy_prince | vigneshvar: okay.. then i am not sure .. | 07:06 |
| lazy_prince | does that image have kernel and ramdisk associaated with it..? | 07:07 |
| *** pensu1 has joined #openstack-ironic | 07:07 | |
| vigneshvar | lazy_prince: cirros image does not have, kernel ramdisk associated | 07:08 |
| *** bluex-pl has joined #openstack-ironic | 07:08 | |
| vigneshvar | lazy_prince: i think so | 07:08 |
| vigneshvar | lazy_prince: But i have ir-deploy kernel and ramdisk | 07:08 |
| *** pensu has quit IRC | 07:09 | |
| vigneshvar | lazy_prince: glance.wsgi.server [b1d00498-ab14-4b7e-abea-3bd4614fed64 3c305fadb2e44caf9e3ede8ac26ab729 e4692122354d42e985e3b7572d804b7b - - -] 127.0.0.1 - - [24/Sep/2014 17:41:16] "GET /images/b481cf87-72fe-4760-9316-d36a9f7243dd HTTP/1.1" 200 989 0.019835 | 07:10 |
| vigneshvar | lazy_prince: b481cf87-72fe-4760-9316-d36a9f7243dd is the id of image mentioned above | 07:11 |
| *** pensu1 has quit IRC | 07:11 | |
| lazy_prince | so for ironic to depploy images, it needs a kernel and ramdisk associated with the user image. when using pxe driver.. | 07:12 |
| *** ifarkas has joined #openstack-ironic | 07:12 | |
| *** jcoufal has joined #openstack-ironic | 07:12 | |
| lazy_prince | vigneshvar: if they are not found, it will error out and no glance error will be seen.. not sure about agent driver.. | 07:13 |
| vigneshvar | lazy_prince: oh ok. I was just trying to follow the docs as it is. In the glance image-list , i can find the image existing. | 07:14 |
| vigneshvar | lazy_prince: Also to add to it. I tried to vnc the instance manually | 07:14 |
| vigneshvar | lazy_prince: And i have seen login prompt | 07:14 |
| lazy_prince | the image properties point to the respective kernel and ramdisk images in glance. when image is downloaded, the properties are examined and those images are also downloaded. otherwise error is thrown.. | 07:15 |
| vigneshvar | lazy_prince: and after few min, it is turned off with error on ironic conductor that "ImageDownloadError" | 07:15 |
| vigneshvar | lazy_prince: ok | 07:16 |
| vigneshvar | lazy_prince: How do i check if the image is associated with proper kernel and ramdisk | 07:16 |
| lazy_prince | vigneshvar: glance image-show <image-id> | 07:17 |
| vigneshvar | lazy_prince: It does not show any association but in my case flavor "baremetal" has kernel and ramdisk specififed | 07:19 |
| vigneshvar | lazy_prince: {"cpu_arch": "x86_64", "baremetal:deploy_kernel_id": "80eff265-bb9d-41ea-9560-f0bf0b1287d5", "baremetal:deploy_ramdisk_id": "5fad4c22-9583-45a9-9d21-09a24c304025"} | 07:19 |
| *** romcheg has joined #openstack-ironic | 07:19 | |
| vigneshvar | lazy_prince: these are ir-deploy kernel and ram | 07:19 |
| vigneshvar | and in glance logs i can see that these kernel and ramdisk are downloaded | 07:20 |
| *** Poornima has joined #openstack-ironic | 07:20 | |
| *** romcheg has quit IRC | 07:20 | |
| *** romcheg has joined #openstack-ironic | 07:23 | |
| *** pensu has joined #openstack-ironic | 07:27 | |
| vigneshvar | lazy_prince: I guess cirros image has kernel and ramdisk in it, i tried to boot the vm manually with qemu cmds and it works without passing kernel or ramdisk | 07:34 |
| lazy_prince | vigneshvar: tir-deploy is used to boot the bm into deployment mode. and then when user image is provisioned, it uses user image associated kernel and ramdisk to pxe boot the instance. these user image associated kernel and ramdisk are diffrenet from ir-deploy*. ir-deploy is customized to start the IPA or iSCSI for Image deployment.. | 07:34 |
| vigneshvar | lazy_prince: Yes i got it | 07:34 |
| vigneshvar | lazy_prince: I got confused a bit earlier and now am clear | 07:34 |
| vigneshvar | the cirros image is not getting downloaded | 07:34 |
| *** saurabhs has left #openstack-ironic | 07:35 | |
| vigneshvar | lazy_prince: cirros image in itself has kernel and ramdisk and it boots perfectly when i boot it manually | 07:35 |
| vigneshvar | lazy_prince: i found it is a mere REST call to glance to download image. I even tried to download the image manually from glance and it works fine | 07:36 |
| *** igordcard has joined #openstack-ironic | 07:37 | |
| *** jcoufal has quit IRC | 07:46 | |
| *** jcoufal has joined #openstack-ironic | 07:46 | |
| *** rushiagr_away is now known as rushiagr | 08:00 | |
| *** bluex-pl has quit IRC | 08:03 | |
| lazy_prince | vigneshvar: correct.. but nodes are generally not configured to boot from disk. they are configured to boot from PXE. | 08:05 |
| lazy_prince | and so the ironic conductor needs to know what kernel and ramdisk to provide when node tries to boot. | 08:06 |
| *** yuanying has quit IRC | 08:08 | |
| openstackgerrit | A change was merged to openstack/ironic: Add documentation for ironic-dbsync command https://review.openstack.org/123518 | 08:08 |
| *** jistr has joined #openstack-ironic | 08:09 | |
| *** Poornima has quit IRC | 08:10 | |
| * lazy_prince back from lunch | 08:11 | |
| *** athomas has joined #openstack-ironic | 08:11 | |
| *** derekh has joined #openstack-ironic | 08:13 | |
| *** lucasagomes has joined #openstack-ironic | 08:15 | |
| openstackgerrit | A change was merged to openstack/ironic: Don't reraise Exceptions from agent driver https://review.openstack.org/122224 | 08:21 |
| *** andreykurilin_ has joined #openstack-ironic | 08:24 | |
| *** Haomeng has joined #openstack-ironic | 08:28 | |
| *** bluex-pl has joined #openstack-ironic | 08:28 | |
| vigneshvar | lazy_prince: So it means i need to associate kernel and ramdisk to cirros image before launching | 08:29 |
| lazy_prince | well.. cirros image is meant for VM deployment.. for baremetal, you can generate an image using diskimage-builder and then give it a try.. | 08:32 |
| *** MattMan has quit IRC | 08:33 | |
| Haomeng | lazy_prince: looks like cirros image has the root paration image type for bm | 08:33 |
| *** MattMan has joined #openstack-ironic | 08:33 | |
| Haomeng | lazy_prince: the disk image is used for vm I think | 08:33 |
| lazy_prince | not sure.. is it a full disk image or just file-system image..? | 08:34 |
| Haomeng | lazy_prince: but not sure if our ipa support both diskimage and root paration image | 08:34 |
| lazy_prince | Haomeng: ack.. | 08:34 |
| Haomeng | https://launchpad.net/cirros/+download , we can find these "uec style tarball (kernel, ramdisk, rootfs.img) for i386" | 08:35 |
| lazy_prince | Haomeng: yup.. thats true.. but is that the image uploaded to glance by devstack..? | 08:37 |
| *** marcoemorais1 has quit IRC | 08:38 | |
| lazy_prince | Haomeng: I guess vigneshvar is using the image which is uploaded to glance by default. | 08:38 |
| Haomeng | lazy_prince: yes, downloaded by devstack | 08:39 |
| *** andreykurilin_ has quit IRC | 08:41 | |
| *** openstackgerrit has quit IRC | 08:48 | |
| *** igordcard has quit IRC | 08:51 | |
| *** rushiagr is now known as rushiagr_away | 08:52 | |
| *** ndipanov_gone is now known as ndipanov | 08:59 | |
| *** r-mibu has left #openstack-ironic | 09:03 | |
| *** Poornima has joined #openstack-ironic | 09:07 | |
| *** lazy_prince has quit IRC | 09:08 | |
| *** Haomeng|2 has joined #openstack-ironic | 09:16 | |
| *** Haomeng has quit IRC | 09:16 | |
| *** pensu has quit IRC | 09:22 | |
| *** pensu has joined #openstack-ironic | 09:23 | |
| *** dtantsur is now known as dtantsur|brb | 09:25 | |
| *** killer_prince has joined #openstack-ironic | 09:26 | |
| *** killer_prince is now known as lazy_prince | 09:27 | |
| *** igordcard has joined #openstack-ironic | 09:43 | |
| *** lazy_prince has quit IRC | 09:48 | |
| *** lazy_prince has joined #openstack-ironic | 09:48 | |
| vigneshvar | lazy_prince: Haomeng|2 It is the image loaded by devstack | 09:49 |
| vigneshvar | lazy_prince: Haomeng|2 : sorry for the late reply | 09:49 |
| lazy_prince | vigneshvar: then it should have the corresponding kernel and ramdisk in glance.. | 09:50 |
| vigneshvar | lazy_prince: Haomeng|2 : i have both images. one cirros.disk and cirros-use which is a ker,ram and root pattern | 09:50 |
| vigneshvar | *cirros-uec | 09:50 |
| lucasagomes | folks should we start abandoning patches that no longer belongs to Ironic? For e.g the nova ironic driver patches | 09:53 |
| lazy_prince | vigneshvar: provision the image and see if it works.. if you get issues, send us link to your ironic logs. | 09:55 |
| vigneshvar | lazy_prince: Haomeng|2 - with kernel ramdisk style image too it fails | 09:56 |
| vigneshvar | lazy_prince: sure | 09:56 |
| *** rushiagr_away is now known as rushiagr | 09:56 | |
| vigneshvar | lazy_prince: Also i noticed that image is being tried to fetch through swift (port 8080) | 09:56 |
| vigneshvar | lazy_prince: http://10.0.2.15:8080/v1/AUTH_/glance/1ee568c1-54d1-4a39-9046-11e788bd46dd?temp_url_sig=aec4a73ac3d80db80ad5a929de2ff801138fc489&temp_url_expires=1411593827' | 09:57 |
| vigneshvar | lazy_prince: is this the expected behaviour | 09:57 |
| lazy_prince | vigneshvar: well.. i have not used swift | 09:57 |
| *** foexle has joined #openstack-ironic | 10:00 | |
| vigneshvar | lazy_prince: Ironic has changed the method of getting images via swift ? may be i am not upto date | 10:01 |
| lazy_prince | vigneshvar: I have my devstack running my devstack and i am using pxe_ipmitools driver and i do not have swift running.. | 10:07 |
| *** loki184 has joined #openstack-ironic | 10:07 | |
| vigneshvar | lazy_prince: https://github.com/openstack/ironic-specs/blob/master/specs/juno/swift-temp-url.rst | 10:08 |
| vigneshvar | lazy_prince: it is the swift-temp-url | 10:09 |
| vigneshvar | lazy_prince: I guess i am hitting some strange issues with temporary utl stuff | 10:09 |
| vigneshvar | lazy_prince: e2eaaaf1b50a2088b9f9205cb31a0499a6a7d1f5 | 10:10 |
| vigneshvar | lazy_prince: the above commit uses swift temporary url. | 10:10 |
| lazy_prince | vigneshvar: aha so the blueprint says it is only for IPA.. and probably thats why you are facing this issue.. | 10:11 |
| vigneshvar | lazy_prince: am sorry what is IPA | 10:12 |
| lazy_prince | vigneshvar: or may be i misinterpreted it.. | 10:12 |
| lazy_prince | vigneshvar: Ironic python agent | 10:12 |
| vigneshvar | lazy_prince: oh ok. I have seen this blue print commit id in my /opt/stack/ironic | 10:13 |
| lazy_prince | can you check the node details and see what driver node is using..? | 10:15 |
| vigneshvar | lazy_prince: yap | 10:15 |
| vigneshvar | lazy_prince: compute_driver = nova.virt.ironic.IronicDriver | 10:16 |
| vigneshvar | lazy_prince: also noticed in ironic.conf , the following | 10:17 |
| vigneshvar | [glance] | 10:17 |
| vigneshvar | swift_temp_url_duration = 3600 | 10:17 |
| vigneshvar | swift_container = glance | 10:17 |
| vigneshvar | swift_account = AUTH_ | 10:17 |
| vigneshvar | swift_api_version = v1 | 10:17 |
| vigneshvar | swift_endpoint_url = http://10.0.2.15:8080 | 10:17 |
| vigneshvar | swift_temp_url_key = xxxxx | 10:17 |
| lazy_prince | vigneshvar: ironic node-show <node-id> | 10:19 |
| lazy_prince | thats where you will see the driver being used by node.. | 10:19 |
| lazy_prince | vigneshvar: by node, i mean baremetal node | 10:21 |
| *** viktors|afk is now known as viktors | 10:26 | |
| vigneshvar | lazy_prince: driver | agent_ssh | 10:28 |
| vigneshvar | lazy_prince: previously i thought the other agent | 10:28 |
| lazy_prince | vigneshvar: okay, so this agent_ssh points to IPA which means you are using IPA for provisioning.. | 10:28 |
| vigneshvar | lazy_prince: yes you are correct | 10:29 |
| lazy_prince | vigneshvar: i wont be of much help as i have not used IPA yet.. BUt i plan to use it soon.. | 10:29 |
| vigneshvar | lazy_prince: No error in swift services as well | 10:30 |
| vigneshvar | lazy_prince: no problem. Thanks for the help | 10:30 |
| vigneshvar | lazy_prince: I am new to ironic. I am soon planning to contribute to ironic | 10:30 |
| vigneshvar | lazy_prince: After playing around with it for some time | 10:31 |
| lazy_prince | vigneshvar: cool.. | 10:32 |
| *** bluex-pl has quit IRC | 10:33 | |
| *** dtantsur|brb is now known as dtantsur | 10:48 | |
| dtantsur | lucasagomes, yes of course, I even started already | 10:49 |
| *** pensu1 has joined #openstack-ironic | 10:49 | |
| lucasagomes | dtantsur, heh right, yeah I abandoned some there | 10:49 |
| lucasagomes | might have more | 10:49 |
| *** pensu has quit IRC | 10:50 | |
| ramineni | Hi , I'm facing some issues while deploying latest devstack . Its throwing ovs-vsctl: tag=: argument does not end in "=" followed by a value. error and exiting. anybody , any idea on how to resolve this error | 10:54 |
| dtantsur | ramineni, no idea. maybe ask on #openstack-dev ? | 10:55 |
| ramineni | dtantsur: ok, thanks | 10:56 |
| *** pradipta_away has quit IRC | 10:57 | |
| *** ramineni has quit IRC | 10:59 | |
| *** pensu1 has quit IRC | 11:13 | |
| *** lucasagomes is now known as lucas-hungry | 11:20 | |
| *** bluex-pl has joined #openstack-ironic | 11:22 | |
| *** bluex-pl has quit IRC | 11:23 | |
| *** bluex-pl has joined #openstack-ironic | 11:24 | |
| *** bluex-pl has joined #openstack-ironic | 11:24 | |
| *** rakesh_hs has quit IRC | 11:26 | |
| *** vinbs has quit IRC | 11:33 | |
| *** Poornima has quit IRC | 11:35 | |
| *** loki184 has quit IRC | 11:35 | |
| *** coolsvap is now known as coolsvap|afk | 11:37 | |
| *** nikunj2512 has quit IRC | 11:44 | |
| dtantsur | jroll, PXE tests refactor: https://review.openstack.org/#/c/122137/ | 11:54 |
| dtantsur | jroll, tests didn't work back then, because we created files of size 0 and cache didn't clean them even though maximum cache size was 0 :) | 11:54 |
| *** zz_naotok is now known as naotok | 12:07 | |
| *** rushiagr is now known as rushiagr_away | 12:26 | |
| *** dprince has joined #openstack-ironic | 12:26 | |
| *** rushiagr_away is now known as rushiagr | 12:28 | |
| *** vigneshvar has quit IRC | 12:29 | |
| *** lucas-hungry is now known as lucasagomes | 12:37 | |
| *** vigneshvar has joined #openstack-ironic | 12:41 | |
| *** jjohnson2 has joined #openstack-ironic | 12:41 | |
| *** naotok is now known as zz_naotok | 12:48 | |
| *** pensu has joined #openstack-ironic | 12:51 | |
| *** bluex-pl has quit IRC | 13:08 | |
| *** bluex-pl has joined #openstack-ironic | 13:08 | |
| *** marcoemorais has joined #openstack-ironic | 13:09 | |
| *** dhellmann has quit IRC | 13:10 | |
| *** dhellmann has joined #openstack-ironic | 13:11 | |
| *** marcoemorais1 has joined #openstack-ironic | 13:11 | |
| *** marcoemorais has quit IRC | 13:14 | |
| lucasagomes | dahell! how can I change the locale >.< tried so many things still nothing works... | 13:18 |
| *** rloo has joined #openstack-ironic | 13:25 | |
| dtantsur | lucasagomes, what are you trying to achieve? | 13:25 |
| lucasagomes | https://bugs.launchpad.net/ironic/+bug/1373671 | 13:25 |
| lucasagomes | got it now | 13:25 |
| lucasagomes | had to install a language pack | 13:26 |
| *** bluex-pl has quit IRC | 13:35 | |
| NobodyCam | Good Morning Ironic, Says the man mak'n coffee | 13:52 |
| lucasagomes | NobodyCam, yo! morning | 13:53 |
| NobodyCam | morning lucasagomes | 13:53 |
| *** bluex-pl has joined #openstack-ironic | 13:53 | |
| NobodyCam | how are things going? | 13:53 |
| lucasagomes | NobodyCam, all is well :) what about urself? | 13:54 |
| dtantsur | NobodyCam, morning | 13:55 |
| *** jjohnson2 has quit IRC | 13:56 | |
| *** pcrews has joined #openstack-ironic | 13:56 | |
| NobodyCam | Doing good so far.. Just about to get the first cup of coffee | 14:00 |
| NobodyCam | morning dtantsur :) | 14:00 |
| *** nosnos has quit IRC | 14:00 | |
| *** nosnos has joined #openstack-ironic | 14:00 | |
| *** nosnos has quit IRC | 14:04 | |
| lucasagomes | NobodyCam, :D nice enjoy! | 14:10 |
| NobodyCam | lucasagomes: you on https://bugs.launchpad.net/ironic/+bug/1373671 | 14:11 |
| lucasagomes | NobodyCam, yup, just submitted the fix | 14:12 |
| lucasagomes | tho I -2ed my own patch because we need to fix OoO first | 14:12 |
| lucasagomes | NobodyCam, https://review.openstack.org/#/c/124061 | 14:12 |
| *** pensu has quit IRC | 14:12 | |
| dtantsur | lucasagomes, are we targetting it to rc1? | 14:12 |
| lucasagomes | dtantsur, so... I don't think we have time because we need to fix tripleo ci first | 14:13 |
| lucasagomes | and that requires a patch there | 14:13 |
| lucasagomes | + rebuild the CI enviroment | 14:13 |
| lucasagomes | :( | 14:13 |
| dtantsur | ah I see | 14:13 |
| lucasagomes | so may take days | 14:13 |
| NobodyCam | :( | 14:13 |
| lucasagomes | unless we delay the rc1 release to next week or so | 14:13 |
| lucasagomes | if paramiko supported passing env variables it would be diff :( I wouldn't have to change the ssh commnds | 14:14 |
| lucasagomes | commands* | 14:14 |
| NobodyCam | lol I was thinking of another wat to address the issue | 14:15 |
| NobodyCam | lucasagomes: your way is much easier | 14:15 |
| lucasagomes | NobodyCam, how would do it? | 14:15 |
| lucasagomes | how would u* | 14:15 |
| lucasagomes | if that way doesnt involve change the commands, we may want to use it | 14:16 |
| NobodyCam | I was thinking something link "An Ansible" power driver, | 14:16 |
| NobodyCam | oh no it would change things | 14:16 |
| *** openstackgerrit has joined #openstack-ironic | 14:16 | |
| lucasagomes | oh | 14:16 |
| lucasagomes | alright, so simple way is it :) | 14:16 |
| NobodyCam | :) | 14:16 |
| NobodyCam | brb | 14:17 |
| *** dlaube has joined #openstack-ironic | 14:18 | |
| *** ChuckC has joined #openstack-ironic | 14:19 | |
| *** ChuckC_ has quit IRC | 14:20 | |
| *** linggao has joined #openstack-ironic | 14:21 | |
| *** ChuckC has quit IRC | 14:25 | |
| *** pensu has joined #openstack-ironic | 14:26 | |
| *** viktors has quit IRC | 14:34 | |
| NobodyCam | hum I wounder if there is a way to trnaslate bash on the fly? | 14:34 |
| lucasagomes | ? | 14:36 |
| *** openstackgerrit has quit IRC | 14:40 | |
| *** openstackgerrit has joined #openstack-ironic | 14:42 | |
| *** marcoemorais1 has quit IRC | 14:46 | |
| lazy_prince | 1is there a way to delete a locked node in ironic..? | 14:47 |
| lucasagomes | lazy_prince, hmm I hope not | 14:48 |
| * lucasagomes checks | 14:48 | |
| lucasagomes | lazy_prince, I would say we would need to break the lock first (in case it's stuck) and then delete it | 14:48 |
| lucasagomes | problem is, breaking a lock is not something you can do via our api :( | 14:49 |
| lazy_prince | aha.. then not user friendly.. | 14:49 |
| lucasagomes | yeah :( | 14:49 |
| lazy_prince | is there a plan to fix it going forward..? | 14:50 |
| lucasagomes | we thought about some ways to do that, but we still didn't come up with a good solution for it | 14:50 |
| lucasagomes | lazy_prince, I will add it right now to the list of things to talk about in Paris | 14:50 |
| NobodyCam | lazy_prince: hack the database | 14:50 |
| lucasagomes | cause we talked about it in the best, we thought about two ways of doing that | 14:50 |
| lucasagomes | via api, or having a special utility to do that | 14:50 |
| lucasagomes | afair... but we didn't actually implemented it yet because we didn't have a consensus | 14:51 |
| lucasagomes | lazy_prince, but that's important,thanks for bringing it up | 14:51 |
| *** pensu has quit IRC | 14:51 | |
| lucasagomes | yeah right now as NobodyCam said, you have to hack the db :( | 14:51 |
| lucasagomes | remove the lock and then delete it | 14:51 |
| lazy_prince | hacing db can be recommended to a developer but not to an end user.. | 14:52 |
| *** andreykurilin_ has joined #openstack-ironic | 14:52 | |
| lucasagomes | lazy_prince, yeah indeed | 14:52 |
| lucasagomes | I know it's a lame excuse from my part, but users of ironic are sys admins... so it's likely to them to have access to the db | 14:53 |
| lucasagomes | but it's just a lame excuse and I should be shot by saying it | 14:53 |
| lucasagomes | for saying* | 14:54 |
| lazy_prince | well.. having access is one thing and being good at db command is another.. what if user screws up with db tables.. | 14:54 |
| lucasagomes | +1 | 14:54 |
| NobodyCam | lazy_prince: mysql -u user -p -e 'update nodes set reservation=NULL where id==<uuid>' ironic | 14:54 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Remove unneeded context initialization in tests https://review.openstack.org/124073 | 14:54 |
| lucasagomes | NobodyCam, where "uuid=" | 14:54 |
| NobodyCam | doh | 14:54 |
| rloo | I wonder if we should document stuff like this, until we have an API or something. Maybe open a bug about it? | 14:55 |
| NobodyCam | rloo: something like a developer quick hack page | 14:56 |
| lucasagomes | rloo, so... idk too. The real bug would be "why the node go locked and stuck?" | 14:56 |
| lucasagomes | where's the timeout or something | 14:56 |
| lucasagomes | that's a workaround/hack | 14:56 |
| rloo | lucasagomes: yes, there's a real bug there. BUT we will never be able to guarantee that a node will be in the 'correct' state or whatever. | 14:57 |
| lucasagomes | offering a tool that people can use is fine but we should tell them to just use it if it's the last attempt ever! | 14:57 |
| lucasagomes | rloo, well... idk never but we should aim that | 14:57 |
| rloo | lucasagomes: yes, agree. anything that is 'modify the db' should be a last attempt. | 14:58 |
| lucasagomes | timeouts, periodic tasks to check for it | 14:58 |
| lucasagomes | idk there's a couple of tool we can/should use | 14:58 |
| lazy_prince | agree.. | 14:58 |
| rloo | lucasagomes: for sure we should aim, but we have all these drivers etc. and bugs can get introduced... | 14:58 |
| rloo | sorry, i've seen it with nova itself. where we've had to muck with the db to get a vm out of some state. | 14:59 |
| lucasagomes | rloo, yeah, tho this kinda of errors ^ should be handled by the conductor manager | 14:59 |
| lucasagomes | it should be resilient to failures on bug drivers | 14:59 |
| rloo | lucasagomes: 'should be'. | 14:59 |
| lucasagomes | "timeout because the driver didn't completed the task X" | 14:59 |
| lucasagomes | removing lock, cleaning things up | 14:59 |
| lucasagomes | rloo, yeah | 14:59 |
| * lucasagomes is trying to think positive here | 15:00 | |
| rloo | lucasagomes: that's the problem. the operator doesn't care if things should work. something doesn't work... wil they be frustrated cuz it should work but it doesn't? | 15:00 |
| *** dtantsur is now known as dtantsur|brb | 15:00 | |
| *** jjohnson2 has joined #openstack-ironic | 15:00 | |
| lucasagomes | rloo, they will | 15:00 |
| lucasagomes | that's why I still agree we need the tool | 15:00 |
| lucasagomes | rloo, we are in the same page here :) | 15:00 |
| rloo | lucasagomes: ok good ;) | 15:01 |
| rloo | lucasagomes: in the meantime, w/o such a tool, we might want to have a hack/faq/wiki somewhere with helpful suggestions... | 15:01 |
| lucasagomes | rloo, +1 | 15:01 |
| lucasagomes | yeah maybe a wiki for hackings would be great | 15:02 |
| rloo | not sure what/if other projects do. | 15:02 |
| lucasagomes | until we decide where this tool should live | 15:02 |
| lucasagomes | either on ironic api directly or a utility like ironic-dbsync etc | 15:02 |
| lucasagomes | ironic-release-node heh | 15:02 |
| lucasagomes | rloo, hmm me netheir | 15:02 |
| * lucasagomes looks | 15:02 | |
| NobodyCam | hummm... could we do somehting simple like, clear any lock when a node is put in to maintance mode? | 15:04 |
| *** pensu has joined #openstack-ironic | 15:05 | |
| Shrews | NobodyCam: wouldn't that be dangerous for valid locks? | 15:07 |
| lucasagomes | NobodyCam, hmm idk, gotta look into it... but sounds like two different things | 15:07 |
| NobodyCam | if the node is going into maintenace then nothing will operate on it | 15:07 |
| lucasagomes | yeah it's something to think about | 15:10 |
| lucasagomes | but IMO I would like that cleaning locks should be something like last attempt | 15:10 |
| lucasagomes | therefor not really part of the whole ironic workflow | 15:11 |
| jroll | or we could like, use zookeeper, where you can set a ttl :) | 15:11 |
| lucasagomes | if we have to break a lock means we have a bug | 15:11 |
| lucasagomes | I don't want to hide it | 15:11 |
| jroll | (but agree, we need an API or something to fix these) | 15:11 |
| jroll | (or --force in the client) | 15:11 |
| rloo | +1 lucasagomes. We just need some one-off thing to fix something I think. | 15:12 |
| jroll | lucasagomes: if a conductor crashes, locks may get stuck | 15:12 |
| jroll | also, good morning rloo lucasagomes Shrews NobodyCam lazy_prince :) | 15:12 |
| NobodyCam | good morning :) on the call | 15:12 |
| NobodyCam | on a call | 15:12 |
| *** ChuckC has joined #openstack-ironic | 15:12 | |
| yuriyz | morning/evening Ironic | 15:12 |
| * rloo is anti-social. forgets to say hi to everyone. | 15:13 | |
| lazy_prince | jroll: morning :) | 15:13 |
| jroll | hiya yuriyz :) | 15:13 |
| romcheg | Good evening/morning folks! | 15:13 |
| jroll | \o romcheg | 15:13 |
| lucasagomes | jroll, morning | 15:13 |
| Shrews | hey jroll | 15:14 |
| lucasagomes | jroll, well yeah, but and if the conductor that takes the node over has a periodic task to check the locks | 15:14 |
| dlaube | g'morning | 15:14 |
| jroll | lucasagomes, rloo, hack wiki++, we do a combination of that and some scripts internally :) | 15:14 |
| lucasagomes | compare with the time of the last ping | 15:14 |
| lucasagomes | and based on that break the lock with a timeout | 15:14 |
| jroll | lucasagomes: sure, when we have that periodic task, but we don't right now | 15:14 |
| *** r-daneel has joined #openstack-ironic | 15:14 | |
| lucasagomes | jroll, yeah I know | 15:14 |
| lucasagomes | I added to the spreadsheet this topic | 15:15 |
| lucasagomes | so we can talk about it | 15:15 |
| jroll | yeah | 15:15 |
| jroll | kylestev wrote a spec and some code for pluggable locking | 15:16 |
| jroll | so we don't have to use the db | 15:16 |
| jroll | he wrote a zookeeper plugin | 15:16 |
| lucasagomes | nice | 15:16 |
| lucasagomes | it's for K right? i've seem some patches up | 15:17 |
| lucasagomes | haven't looked into it tho | 15:17 |
| jroll | yeah | 15:18 |
| jroll | I'll add links to the google doc | 15:18 |
| jroll | there was something else I was going to add but forgot :/ | 15:19 |
| lucasagomes | :( | 15:19 |
| * jroll looks | 15:19 | |
| jroll | oh! | 15:21 |
| jroll | policy.json | 15:21 |
| * jroll adds | 15:21 | |
| lucasagomes | policy json? | 15:21 |
| *** athomas has quit IRC | 15:24 | |
| rloo | hey, has anyone looked at the hash ring issue(s)? Based on what I see (or don't see), I don't know what we'll be able to accomplish by RC1. | 15:26 |
| jroll | lucasagomes: yeah, fine-grained control of api access | 15:26 |
| jroll | lucasagomes: read-only users, unprivileged operators that may only e.g. update nodes | 15:26 |
| *** athomas has joined #openstack-ironic | 15:27 | |
| lucasagomes | rloo, the take_over() or the balancing problem? | 15:27 |
| lucasagomes | jroll, gotcha, that's good! | 15:27 |
| rloo | lucasagomes: that's what I meant by issues :-( ANY of it. | 15:27 |
| lucasagomes | rloo, right the take_over() I was investigating | 15:27 |
| lucasagomes | tl;dr I think a better way to solve it for the PXE driver is working on an iPXE driver that will be stateless | 15:28 |
| lucasagomes | we discussed a bit it yesterday, me devananda JayF and jroll | 15:28 |
| lucasagomes | but not for J | 15:28 |
| rloo | lucasagomes: yes, I saw a bit of that discussion wrt ipxe. So what, if anything, do we think we can do for J? | 15:29 |
| lucasagomes | for J ... I really dunno if we should tackle that, can't find any good way to solve it | 15:29 |
| jroll | stateless drivers ftw | 15:29 |
| lucasagomes | just horrible ways that involves looping through many nodes x.x | 15:30 |
| lucasagomes | jroll, +1!! | 15:30 |
| jroll | lucasagomes: I assume you're not in charge of this, but maybe you know who to yell at about ssl errors :) https://securityblog.redhat.com/2014/09/24/bash-specially-crafted-environment-variables-code-injection-attack | 15:30 |
| rloo | There is this patch to fix the hash ring stability but there's a caveat about adding the fix to an existing ironic env: https://review.openstack.org/#/c/118932/ | 15:30 |
| lucasagomes | jroll, heh yeah I'm not in charge :P hmm lemme try to find someone | 15:31 |
| NobodyCam | morning yuriyz and romcheg | 15:31 |
| lucasagomes | jroll, for now, that may help https://plus.google.com/103866178509451949283/posts/j9LPZfBpd7F | 15:31 |
| jroll | rloo: "I'll probably leave this patch to be handled by greg when his rebalancing work is implemented... as thats needed for landing this." | 15:32 |
| jroll | rloo: not sure why there's no dependency, but yeah | 15:32 |
| rloo | jroll: exactly. | 15:32 |
| rloo | and there isn't any patch for the rebalancing yet. | 15:32 |
| *** rwsu has quit IRC | 15:33 | |
| rloo | it looks like everything targeted for rc1 is done EXCEPT the hash-ring-etc stuff. | 15:33 |
| jroll | lucasagomes: yeah, I'm all patched up, but someone in another channel complained about the ssl error :) | 15:33 |
| jroll | thank you though! | 15:33 |
| jroll | rloo: yeah, I'll leave it to deva but I don't think anything is going to land :/ | 15:33 |
| *** ramineni has joined #openstack-ironic | 15:34 | |
| lucasagomes | np | 15:34 |
| rloo | jroll: it WOULD be good to have the hash-ring-stability patch for new ironic envs though, but probably not worth it at the expense of existing environments (like yours) | 15:36 |
| jroll | rloo: this doesn't affect the agent driver :) | 15:37 |
| rloo | jroll: oh yeah, I forgot about that. | 15:37 |
| *** andreykurilin_ has quit IRC | 15:40 | |
| *** k4n0 has quit IRC | 15:43 | |
| *** pensu has quit IRC | 15:45 | |
| *** rwsu has joined #openstack-ironic | 15:45 | |
| * NobodyCam makes a fresh pot od coffee...brb | 15:46 | |
| NobodyCam | ahh coffee :-p | 15:52 |
| *** slagle_ has joined #openstack-ironic | 15:52 | |
| *** slagle_ has quit IRC | 15:52 | |
| *** slagle has quit IRC | 15:53 | |
| *** slagle has joined #openstack-ironic | 15:54 | |
| *** jistr has quit IRC | 15:54 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Update ironic.conf.sample https://review.openstack.org/124090 | 15:54 |
| lucasagomes | jroll, rloo NobodyCam dtantsur|brb ^ again :(((((((((((( | 15:54 |
| lucasagomes | we really need to talk whether we should get rid of the sample config or not | 15:55 |
| lucasagomes | the failure: http://logs.openstack.org/43/122743/4/gate/gate-ironic-pep8/95da55d/ | 15:55 |
| *** marcoemorais has joined #openstack-ironic | 15:55 | |
| *** andreykurilin_ has joined #openstack-ironic | 15:56 | |
| NobodyCam | happy happy joy joy | 15:57 |
| jroll | lucasagomes: :| | 15:57 |
| NobodyCam | lucasagomes: +2'd ... do we want a bug for this? | 15:57 |
| lucasagomes | NobodyCam, I think it would be good for us to centralize the discussion at least | 15:57 |
| jroll | +2'd | 15:57 |
| lucasagomes | I'm +1 for getting rid of the sample I think | 15:58 |
| lucasagomes | and for those who wants it, they can generate it locally | 15:58 |
| Shrews | lucasagomes++ | 15:58 |
| NobodyCam | what if we auto generated it in the docs? would be of any value | 15:59 |
| rloo | so I might be the only one that likes having it. | 15:59 |
| NobodyCam | I like having it | 15:59 |
| rloo | I suspect we might want to bring it up in the ML | 15:59 |
| rloo | yay NobodyCam! | 15:59 |
| jroll | rloo: did you just volunteer? :) | 15:59 |
| lucasagomes | rloo, +1 can you send it to the ML? | 15:59 |
| lucasagomes | :P | 15:59 |
| lucasagomes | jroll, was faster | 15:59 |
| jroll | :P | 16:00 |
| rloo | You can read English, right? 'we' is not 'me' :D | 16:00 |
| *** hemna has joined #openstack-ironic | 16:00 | |
| jroll | somebody has to do it | 16:00 |
| jroll | and I hear you're a really good writer :) | 16:00 |
| rloo | to be honest, I try not to contribute to the ML. There's something funky about my email (yahoo) and I haven't gotten around to changing to a non-work email. | 16:00 |
| lucasagomes | paper rock scissors? | 16:00 |
| rloo | jroll: I'm trying to improve my delegation skills ;) | 16:01 |
| jroll | heh | 16:01 |
| jroll | rloo: you don't have to be subscribed to the list to send email to it, just use your personal email :) | 16:02 |
| jroll | and don't reply from the work email :P | 16:02 |
| rloo | jroll: I like to keep my work and personal emails separate. haven't yet figure out how to deal with it. | 16:02 |
| jroll | lucasagomes: there's a cartoon for this... | 16:02 |
| * jroll looks | 16:02 | |
| lucasagomes | heh | 16:02 |
| *** ndipanov is now known as ndipanov_gone | 16:02 | |
| *** vigneshvar has quit IRC | 16:02 | |
| rloo | oops, did I just let it slip that this is "just" work for me? :) | 16:03 |
| NobodyCam | lol | 16:03 |
| jroll | rloo: do you see us working on ironic outside of work hours? :P | 16:03 |
| rloo | anyway, the folks that WANT to get rid of it, should propose it. Not the folks that are fine with status quo. (That's the real reason, ha.) | 16:03 |
| lucasagomes | alright I can send the email | 16:04 |
| lucasagomes | 1 sec | 16:04 |
| rloo | thx lucasagomes, you're a trooper! | 16:04 |
| *** andreykurilin_ has quit IRC | 16:04 | |
| rloo | hmm, not sure why I'm thanking you. I don't want to get rid of it... | 16:04 |
| lucasagomes | rloo, heh | 16:05 |
| jroll | ha | 16:05 |
| *** ChuckC has quit IRC | 16:06 | |
| *** dtantsur|brb is now known as dtantsur | 16:08 | |
| *** bluex-pl has quit IRC | 16:10 | |
| *** rushiagr is now known as rushiagr_away | 16:11 | |
| *** lsmola has quit IRC | 16:12 | |
| *** igordcard has quit IRC | 16:13 | |
| lucasagomes | rloo, jroll Shrews email sent | 16:23 |
| lucasagomes | to openstack-dev | 16:23 |
| lucasagomes | I didn't elaborate it much... but yeah it's there | 16:23 |
| jroll | thanks for that :) | 16:23 |
| * jroll bbl | 16:24 | |
| *** ramineni has quit IRC | 16:24 | |
| *** pensu has joined #openstack-ironic | 16:28 | |
| lucasagomes | jroll, JayF btw https://review.openstack.org/#/c/120149/ I think it affect u guys | 16:28 |
| JayF | looking | 16:28 |
| JayF | lucasagomes: It won't; we don't have a single environment up to four digits yet | 16:28 |
| JayF | lucasagomes: but thanks for the pointer and I'll make sure we have this downstreamed before we get there | 16:29 |
| lucasagomes | JayF, I c, no problem | 16:30 |
| *** openstackgerrit has quit IRC | 16:33 | |
| *** ramineni has joined #openstack-ironic | 16:35 | |
| rloo | lucasagomes: I'm not quite sure, but I think you may have to rebase 120149. (Hopefully I'm wrong.) | 16:35 |
| lucasagomes | :( | 16:35 |
| lucasagomes | lemme try | 16:36 |
| lucasagomes | rloo, seems grand :) | 16:37 |
| rloo | lucasagomes: ok good. I only noticed that the file had changed. | 16:37 |
| lucasagomes | rloo, yeah thanks git for being smart and tracking content and not files | 16:37 |
| lucasagomes | instead of files* | 16:37 |
| rloo | lucasagomes: yeah! | 16:38 |
| *** athomas has quit IRC | 16:38 | |
| *** hemna has quit IRC | 16:38 | |
| *** ramineni has quit IRC | 16:40 | |
| *** rloo has quit IRC | 16:45 | |
| lucasagomes | alright I will call it a day | 16:47 |
| lucasagomes | came to the office and have to get the train back home yet | 16:47 |
| JayF | have a good night lucas | 16:47 |
| NobodyCam | night lucasagomes | 16:47 |
| lucasagomes | have a good night everybody! | 16:47 |
| *** derekh has quit IRC | 16:48 | |
| *** lucasagomes has quit IRC | 16:48 | |
| *** rushiagr_away is now known as rushiagr | 16:48 | |
| *** romcheg1 has joined #openstack-ironic | 16:52 | |
| *** openstackgerrit has joined #openstack-ironic | 16:54 | |
| ryanpetrello | hey Ironic folks | 16:54 |
| ryanpetrello | https://review.openstack.org/#/c/123520/ just merged | 16:54 |
| ryanpetrello | which should make https://bugs.launchpad.net/ironic/+bug/1271317 much easier on you | 16:55 |
| *** hemna has joined #openstack-ironic | 16:55 | |
| JayF | I think we won't be able to update that requirement in time for Juno | 16:56 |
| JayF | but it should be fixed in early Kilo | 16:56 |
| JayF | thanks :) | 16:56 |
| *** pensu has quit IRC | 17:10 | |
| *** harlowja_away is now known as harlowja | 17:12 | |
| *** yarker has quit IRC | 17:22 | |
| *** pensu has joined #openstack-ironic | 17:25 | |
| *** ChuckC has joined #openstack-ironic | 17:30 | |
| *** pensu has quit IRC | 17:30 | |
| *** openstackgerrit has quit IRC | 17:33 | |
| *** yarker has joined #openstack-ironic | 17:33 | |
| *** openstackgerrit has joined #openstack-ironic | 17:40 | |
| *** pcrews has quit IRC | 17:41 | |
| *** pcrews has joined #openstack-ironic | 17:42 | |
| *** ifarkas has quit IRC | 17:43 | |
| *** pensu has joined #openstack-ironic | 17:45 | |
| NobodyCam | should I put up a tripleO-ci patch to handle https://review.openstack.org/#/c/124061 | 17:55 |
| *** penick has joined #openstack-ironic | 17:57 | |
| *** rloo has joined #openstack-ironic | 17:58 | |
| jroll | ryanpetrello: what jay said, I'm told we can't update requirements this late | 18:19 |
| *** marcoemorais has quit IRC | 18:20 | |
| *** marcoemorais has joined #openstack-ironic | 18:20 | |
| jroll | JayF: we can also bump the api limit rather than pulling in that nova change | 18:20 |
| *** marcoemorais has quit IRC | 18:20 | |
| *** rushiagr is now known as rushiagr_away | 18:20 | |
| JayF | ah, true | 18:20 |
| *** marcoemorais has joined #openstack-ironic | 18:21 | |
| *** dtantsur is now known as dtantsur|afk | 18:22 | |
| *** dprince has quit IRC | 18:24 | |
| *** lazy_prince is now known as killer_prince | 18:25 | |
| rloo | NobodyCam: wrt 124061, shouldn't it get two +2s first, before updating tripleO-CI? | 18:30 |
| jroll | rloo: -1, let's have the tripleo patch ready to go | 18:31 |
| jroll | the tripleo patch can be updated if the ironic patch changes | 18:31 |
| jroll | also, lucas put up a tripleo patch already :P | 18:32 |
| rloo | jroll: well, as long as the tripleo patch isn't merged before the ironic patch is OKd but not yet approved. | 18:32 |
| jroll | https://review.openstack.org/#/c/124072 | 18:32 |
| *** pcrews has quit IRC | 18:32 | |
| jroll | yeah | 18:32 |
| rloo | why can't 124072 depend on the tripleO-CI patch? | 18:33 |
| jroll | because they're different repos, you can't do cross-repo dependencies | 18:33 |
| rloo | why not? | 18:33 |
| jroll | because dependencies are just git branches | 18:34 |
| rloo | seems like a deficiency. | 18:34 |
| jroll | it's a pain point | 18:37 |
| rloo | just seems like we have a bunch of smart people -- someone should be able to figure out how to do it :) | 18:38 |
| jroll | heh | 18:39 |
| jroll | we have bigger infra problems to work on :P | 18:39 |
| rloo | oh yeah, I forgot... :-( | 18:39 |
| *** jcoufal has quit IRC | 18:40 | |
| lifeless | sdague has a patch up to alow cross-repo deps | 18:45 |
| rloo | jroll: ^^, lifeless, good to know! | 18:48 |
| jroll | omg yesssssssss | 18:48 |
| *** mikedillion has joined #openstack-ironic | 18:48 | |
| NobodyCam | lifeless: AWESOME!!!!! | 18:59 |
| NobodyCam | jroll: great I didn't see that one | 18:59 |
| * NobodyCam also thinks that the tripleo ci regex is getting tuffer to read | 19:00 | |
| *** AJaeger has joined #openstack-ironic | 19:00 | |
| *** marcoemorais has quit IRC | 19:01 | |
| *** marcoemorais has joined #openstack-ironic | 19:01 | |
| AJaeger | hi devananda , looking at your comment on https://review.openstack.org/#/c/123342/, let me check transifex for ironic... | 19:01 |
| *** marcoemorais has quit IRC | 19:02 | |
| AJaeger | devananda: it works as designed - are you around to explain how it's designed? ;) | 19:02 |
| *** marcoemorais has joined #openstack-ironic | 19:03 | |
| devananda | AJaeger: I could be wrong about it being broken again ... but i'm seeing translation patches posted to Ironic without any translations in them again | 19:03 |
| devananda | AJaeger: and IIRC, that's what we fixed last cycle | 19:03 |
| NobodyCam | added my 2 cents to the tripleo ci patch | 19:03 |
| *** marcoemorais has quit IRC | 19:04 | |
| AJaeger | devananda: thanks for double checking | 19:04 |
| AJaeger | We only import translated files that are at least 75 % translated. The major files are not translated - see https://www.transifex.com/projects/p/ironic/ | 19:04 |
| *** marcoemorais has joined #openstack-ironic | 19:04 | |
| *** marcoemorais has quit IRC | 19:04 | |
| devananda | AJaeger: right. so. right now, none of the Ironic languages are >= 75% translated in transifex | 19:04 |
| devananda | AJaeger: so why are there a bunch of changes? | 19:04 |
| NobodyCam | hey hey devananda hows the meetings going | 19:04 |
| *** marcoemorais has joined #openstack-ironic | 19:05 | |
| AJaeger | devananda: those are the changes in line numbers mainly | 19:05 |
| AJaeger | I've reworked the scripts in the last weeks so that they propose far less changes | 19:05 |
| devananda | AJaeger: sure. but those files don't have any substantive translations in them to begin with | 19:05 |
| AJaeger | give me a few minutes to double check the patch... | 19:05 |
| devananda | AJaeger: should we have .po files that contain only 3 lines with any translation? | 19:07 |
| devananda | AJaeger: http://paste.openstack.org/show/115385/ | 19:08 |
| AJaeger | devananda: Indeed there are some files that have far less than 75 % of translations - and we update them when changes happen | 19:08 |
| AJaeger | devananda: msgfmt --statistics zh_TW/LC_MESSAGES/ironic-log-error.po | 19:09 |
| AJaeger | 3 translated messages, 40 untranslated messages. | 19:09 |
| devananda | AJaeger: that's much cleaner than my grep :) | 19:09 |
| AJaeger | devananda: let's do the following: I'll propose a patch that removes all those files, and then you have far less imports. | 19:09 |
| AJaeger | Once a file is translated to 75 % again, the proposal bot will propose it ironic again... | 19:10 |
| * AJaeger works on a patch... | 19:10 | |
| devananda | AJaeger: thanks! | 19:10 |
| *** pcrews has joined #openstack-ironic | 19:12 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack/ironic: Remove untranslated PO files https://review.openstack.org/124141 | 19:19 |
| AJaeger | devananda: not a single file had enough translations ;( | 19:19 |
| devananda | AJaeger: yep | 19:19 |
| AJaeger | devananda: once 124141 is merged, abandon the bot patch - https://review.openstack.org/#/c/123342 - and you won't get a new one proposed until a team has finished translating... | 19:20 |
| devananda | cool, thanks much | 19:20 |
| AJaeger | with my recent changes to the scripts, not even the pot (source file) will get updated... | 19:20 |
| *** jbjohnso__ has quit IRC | 19:20 | |
| pensu | Hi all, I have added a new physical node in my devstack Ironic setup, but I am not able to boot an instance on it... | 19:21 |
| NobodyCam | pensu: are you getting an error | 19:22 |
| pensu | nova-scheduler is not recognizing the node, any idea how can I boot an instance on a new node? | 19:22 |
| pensu | NobodyCam: no actually, there are three virtual nodes made by devstack, only they are being considered by scheduler.. | 19:22 |
| NobodyCam | pensu: are you trying to rm vm and BM from the same devstack? | 19:23 |
| NobodyCam | s/rm/run/ | 19:23 |
| pensu | NobodyCam: I can see the node in ironic node list, I can shut the power off and switch it on with ironic API.... | 19:23 |
| NobodyCam | so ironic is working | 19:24 |
| pensu | NobodyCam: yes, I am.... | 19:24 |
| NobodyCam | sounds like a nova config issue | 19:24 |
| NobodyCam | I don't think you can do that un soething i am unaware of has changed | 19:25 |
| pensu | NobodyCam: yes it is, it's nova-scheduler which is not able to understand that there is a new node... | 19:25 |
| NobodyCam | https://github.com/openstack/tripleo-image-elements/blob/master/elements/nova/os-apply-config/etc/nova/nova.conf#L167 | 19:26 |
| pensu | NobodyCam: no, booting through a BM through virtual machine would be a challenge, I am thinking about providing a public IP to my VM, that should do the trick.... | 19:26 |
| NobodyCam | pensu: I boot BM node from the svm seed all the time | 19:27 |
| NobodyCam | s/svm/vm/ | 19:27 |
| pensu | NobodyCam: wow! that works fine? I actually tried that with TripleO, I had a lot of network issues, may be something else then.... | 19:28 |
| pensu | NobodyCam: so, basically I need to switch to baremetal host manager, let's see if that works out.....thanks...:) | 19:29 |
| NobodyCam | pensu: the trick is to add the correct interface to the brbm bridge: (ie. ovs-vsctl add-port brbm eth#) | 19:29 |
| NobodyCam | pensu: yep nova can not run vm and bm from a single compute node | 19:31 |
| pensu | NobodyCam: hmmm, that sounds like better solution, gonna try that... | 19:31 |
| pensu | NobodyCam: yeah, that might be it... | 19:32 |
| *** marcoemorais1 has joined #openstack-ironic | 19:39 | |
| NobodyCam | pensu: you'll still need to set the correct host manager :) | 19:39 |
| * NobodyCam steps away to hunt down some food stuffs | 19:40 | |
| *** marcoemorais has quit IRC | 19:41 | |
| pensu | NobodyCam: yeah, and I guess driver too... | 19:45 |
| *** AJaeger has left #openstack-ironic | 19:49 | |
| NobodyCam | hehehe yep:) | 19:51 |
| openstackgerrit | A change was merged to openstack/ironic: Update ironic.conf.sample https://review.openstack.org/124090 | 19:51 |
| *** marcoemorais1 has quit IRC | 19:59 | |
| *** marcoemorais has joined #openstack-ironic | 19:59 | |
| *** marcoemorais has quit IRC | 19:59 | |
| *** marcoemorais has joined #openstack-ironic | 20:00 | |
| *** marcoemorais has quit IRC | 20:00 | |
| *** marcoemorais has joined #openstack-ironic | 20:01 | |
| *** marcoemorais has quit IRC | 20:01 | |
| *** marcoemorais has joined #openstack-ironic | 20:01 | |
| *** marcoemorais has quit IRC | 20:02 | |
| *** marcoemorais has joined #openstack-ironic | 20:02 | |
| pensu | NobodyCam: so, I ran ovs-vsctl add-port brbm eth# and my machine's whole network came down, can't ping, can't ssh... | 20:04 |
| pensu | NobodyCam: anything I should have had in my mind before doing that?! | 20:04 |
| NobodyCam | pensu: did you do eth#? | 20:04 |
| NobodyCam | or eth1 or eth2, etc | 20:04 |
| pensu | NobodyCam: No, I did eth1, I just copied your command there...:) | 20:06 |
| pensu | NobodyCam: okay, that was ambiguous! I ran the command with eth1, I copied the command as it is in chat! | 20:07 |
| *** marcoemorais has quit IRC | 20:07 | |
| NobodyCam | pensu: ok and eth1 is your access to the network? | 20:07 |
| *** foexle has quit IRC | 20:08 | |
| pensu | NobodyCam: yes, it is... | 20:08 |
| NobodyCam | so what I do with tripleo seed is use another eth port on the seeds host.. so for example I will take an unused ethernet port wire it to the BM network add an ip like 192.0.2.254 to it | 20:10 |
| NobodyCam | then add that ether net port to the brbm bridge | 20:11 |
| NobodyCam | but I think you've got more going on | 20:12 |
| NobodyCam | config wise | 20:12 |
| NobodyCam | s/think/feel/ | 20:12 |
| pensu | NobodyCam: okay, that makes sense. I actually had only one ethernet port, that might be the reason why my machine came down... | 20:15 |
| NobodyCam | yep | 20:15 |
| pensu | NobodyCam: I have actually been trying to figure it out for past few days... | 20:15 |
| *** mikedillion has quit IRC | 20:15 | |
| pensu | NobodyCam: Anyways, it's 2 in the morning here, gotta go to sleep, will try it tomorrow, hope it will work....thank you....:) | 20:16 |
| pensu | NobodyCam: good night...:) | 20:16 |
| NobodyCam | pensu: sure... let us know how it goes, and we're happy to help | 20:17 |
| NobodyCam | but will prob need some more info about the setup/config | 20:17 |
| pensu | NobodyCam: yeah, sure.....:) | 20:18 |
| pensu | NobodyCam: tomorrow a new day, a new VM, a new config! | 20:18 |
| NobodyCam | :) | 20:21 |
| *** igordcard has joined #openstack-ironic | 20:28 | |
| *** dhellmann is now known as dhellmann_ | 20:29 | |
| *** pcrews has quit IRC | 20:34 | |
| * NobodyCam makes a run the store... bbaifm | 20:34 | |
| *** linggao has quit IRC | 20:46 | |
| * NobodyCam is back | 20:52 | |
| *** HenryG has quit IRC | 20:53 | |
| *** comstud has quit IRC | 20:56 | |
| *** andreykurilin_ has joined #openstack-ironic | 21:00 | |
| *** comstud has joined #openstack-ironic | 21:01 | |
| *** pensu has quit IRC | 21:04 | |
| *** pcrews has joined #openstack-ironic | 21:05 | |
| *** jjohnson2 has quit IRC | 21:06 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-python-agent: Use LLDP to get switch port mapping https://review.openstack.org/92627 | 21:37 |
| adam_g | devananda, https://review.openstack.org/#/c/124162/ working great so far | 21:49 |
| NobodyCam | adam_g: w00t :) | 21:50 |
| adam_g | actually, works weel for the reboot case. not sure about power on /off | 21:57 |
| NobodyCam | brb | 22:04 |
| adam_g | still racey on stop/start | 22:07 |
| adam_g | not sure why, tho | 22:08 |
| *** marcoemorais has joined #openstack-ironic | 22:15 | |
| NobodyCam | humm | 22:16 |
| *** penick has quit IRC | 22:18 | |
| devananda | adam_g: what's racing? | 22:19 |
| adam_g | commenting in bug | 22:19 |
| adam_g | one sec | 22:19 |
| adam_g | er, review | 22:19 |
| * devananda looks | 22:20 | |
| adam_g | added | 22:22 |
| *** andreykurilin_ has quit IRC | 22:23 | |
| devananda | d'oh! copy/paste errors. lame of me | 22:29 |
| devananda | adam_g: so, the inverse race is a strong possibility in a loaded system, however | 22:30 |
| adam_g | devananda, yeah | 22:30 |
| devananda | oh, i see | 22:30 |
| devananda | yes, there's still a race -- it's just smaller than before | 22:30 |
| *** penick has joined #openstack-ironic | 22:31 | |
| devananda | instead of request 202 -> action complete, the race is now request 202 -> action start | 22:31 |
| adam_g | yeah. can't seem to figure out why reboot seems to be working okay, tho | 22:32 |
| devananda | ironic returns when a greenthread starts the work, and then changes the target_power_state in that greenthread,, but there is still a brief window where that thread has started but not yet updated the DB | 22:32 |
| *** romcheg1 has quit IRC | 22:32 | |
| devananda | solving this should be easy | 22:33 |
| devananda | we need t move changing node.target_power_state from conductor/utils to conductor/manager, keeping it inside the lock context | 22:34 |
| * devananda starts hacknig on that | 22:34 | |
| devananda | adam_g: fixes posted to nova review | 22:44 |
| openstackgerrit | A change was merged to openstack/ironic: Add parameter to override locale to utils.execute https://review.openstack.org/122743 | 22:48 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-python-agent: Adding support for decommissioning https://review.openstack.org/104379 | 22:50 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-python-agent: Adding support for decommissioning https://review.openstack.org/104379 | 22:53 |
| *** penick has quit IRC | 22:53 | |
| *** HenryG has joined #openstack-ironic | 22:59 | |
| *** HenryG has quit IRC | 23:04 | |
| *** HenryG has joined #openstack-ironic | 23:07 | |
| *** marcoemorais has quit IRC | 23:10 | |
| *** marcoemorais has joined #openstack-ironic | 23:10 | |
| *** marcoemorais has quit IRC | 23:10 | |
| *** yuanying has joined #openstack-ironic | 23:10 | |
| *** marcoemorais has joined #openstack-ironic | 23:11 | |
| *** marcoemorais has quit IRC | 23:11 | |
| *** penick has joined #openstack-ironic | 23:11 | |
| *** marcoemorais has joined #openstack-ironic | 23:11 | |
| *** ChuckC has quit IRC | 23:14 | |
| *** marcoemorais has quit IRC | 23:17 | |
| *** marcoemorais has joined #openstack-ironic | 23:17 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Conductor changes target_power_state before starting work https://review.openstack.org/124225 | 23:31 |
| devananda | adam_g: corresponding fix in Ironic ^ | 23:31 |
| adam_g | devananda, nice | 23:33 |
| *** marcoemorais has quit IRC | 23:33 | |
| *** marcoemorais has joined #openstack-ironic | 23:33 | |
| *** marcoemorais has quit IRC | 23:33 | |
| *** marcoemorais has joined #openstack-ironic | 23:33 | |
| *** HenryG has quit IRC | 23:37 | |
| *** igordcard has quit IRC | 23:39 | |
| adam_g | devananda, \o/ working much better for all power ops | 23:40 |
| NobodyCam | devananda: line 280 of https://review.openstack.org/#/c/124225/1/ironic/conductor/manager.py | 23:45 |
| *** boris-42 has quit IRC | 23:47 | |
| *** boris-42 has joined #openstack-ironic | 23:48 | |
| devananda | adam_g: awesome | 23:54 |
| devananda | NobodyCam: ? | 23:54 |
| devananda | typos... | 23:55 |
| NobodyCam | yep | 23:55 |
| NobodyCam | lol | 23:55 |
| NobodyCam | I +2 and nited it | 23:55 |
| devananda | already fixed, and rebased on master | 23:56 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Conductor changes target_power_state before starting work https://review.openstack.org/124225 | 23:56 |
| NobodyCam | Thank you :) | 23:56 |
| NobodyCam | when is flight back? | 23:56 |
| devananda | tomorrow | 23:59 |
| NobodyCam | :) still in meetings? :( | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!