*** bfournie has left #openstack-ironic | 00:01 | |
*** zackf has joined #openstack-ironic | 00:09 | |
*** mtanino has joined #openstack-ironic | 00:10 | |
*** zackf has quit IRC | 00:14 | |
*** zackf has joined #openstack-ironic | 00:24 | |
*** zackf has quit IRC | 00:29 | |
openstackgerrit | Merged openstack/bifrost: Cleanup: adding linebreaks for readability https://review.openstack.org/386775 | 00:36 |
---|---|---|
*** hoangcx has joined #openstack-ironic | 00:39 | |
*** bfournie has joined #openstack-ironic | 00:46 | |
*** mtanino has quit IRC | 00:53 | |
*** rama_y has joined #openstack-ironic | 01:18 | |
*** gcb has joined #openstack-ironic | 01:24 | |
*** tiendc has joined #openstack-ironic | 01:25 | |
*** tiendc has quit IRC | 01:25 | |
*** tiendc has joined #openstack-ironic | 01:26 | |
openstackgerrit | Merged openstack/ironic: Replace parse_root_device_hints with the ironic-lib version one https://review.openstack.org/386204 | 01:28 |
*** jcoufal has quit IRC | 01:38 | |
*** tiendc has quit IRC | 01:49 | |
*** tuanluong has joined #openstack-ironic | 01:51 | |
openstackgerrit | Tang Chen proposed openstack/python-ironicclient: Add missing unit tests for provision state commands https://review.openstack.org/377607 | 02:19 |
openstackgerrit | Tang Chen proposed openstack/python-ironicclient: Trivial: Rename and reorder deploy baremetal test https://review.openstack.org/378140 | 02:19 |
*** getvasanth has joined #openstack-ironic | 02:21 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic power interface for soft reboot and soft power off https://review.openstack.org/216730 | 02:25 |
*** Sukhdev has joined #openstack-ironic | 02:28 | |
*** phuongnh has joined #openstack-ironic | 02:29 | |
*** baoli has quit IRC | 02:35 | |
*** portdirect has quit IRC | 02:46 | |
*** gcb has quit IRC | 02:46 | |
*** vishwanathj has joined #openstack-ironic | 02:53 | |
*** vishwanathj has quit IRC | 02:53 | |
*** MattMan has quit IRC | 02:54 | |
*** MattMan has joined #openstack-ironic | 02:54 | |
*** gcb has joined #openstack-ironic | 02:58 | |
*** vikrant has joined #openstack-ironic | 03:17 | |
*** baoli has joined #openstack-ironic | 03:28 | |
*** chlong has joined #openstack-ironic | 03:30 | |
openstackgerrit | octopuszhang proposed openstack/ironic: Add the way to get the deploy ram disks https://review.openstack.org/387082 | 03:36 |
*** baoli has quit IRC | 03:41 | |
*** tiendc has joined #openstack-ironic | 03:47 | |
*** zhangyufei has joined #openstack-ironic | 03:47 | |
*** pmannidi has quit IRC | 03:52 | |
*** yufei_ has joined #openstack-ironic | 03:53 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/386860 | 04:05 |
*** nicodemos has joined #openstack-ironic | 04:09 | |
*** xavierr has quit IRC | 04:14 | |
*** xavierr has joined #openstack-ironic | 04:14 | |
*** vicky has joined #openstack-ironic | 04:20 | |
*** vikrant has quit IRC | 04:23 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Ipmitool power driver for soft reboot and soft power off https://review.openstack.org/216738 | 04:24 |
*** zackf has joined #openstack-ironic | 04:39 | |
*** zackf has quit IRC | 04:44 | |
*** pmannidi has joined #openstack-ironic | 04:50 | |
hogepodge | TheJulia: have you seen an issue on xenial where if you use the system pip the install scripts fail? | 05:08 |
*** zackf has joined #openstack-ironic | 05:14 | |
*** rama_y has quit IRC | 05:15 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic management I/F for Inject NMI https://review.openstack.org/348191 | 05:15 |
*** zackf has quit IRC | 05:19 | |
*** ChubYann has quit IRC | 05:38 | |
openstackgerrit | octopuszhang proposed openstack/ironic: Add the way to get the deploy ram disks https://review.openstack.org/387082 | 05:43 |
*** jaosorior has joined #openstack-ironic | 05:48 | |
*** zackf has joined #openstack-ironic | 05:52 | |
*** vicky has quit IRC | 05:53 | |
*** vicky has joined #openstack-ironic | 05:54 | |
*** zackf has quit IRC | 05:57 | |
*** rcernin has joined #openstack-ironic | 06:06 | |
*** mjura has joined #openstack-ironic | 06:20 | |
*** yibo has joined #openstack-ironic | 06:25 | |
*** zackf has joined #openstack-ironic | 06:26 | |
*** mjura has quit IRC | 06:27 | |
*** mjura has joined #openstack-ironic | 06:28 | |
*** jtomasek_ has joined #openstack-ironic | 06:31 | |
*** zackf has quit IRC | 06:31 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add a missing error check in ipmitool driver's reboot https://review.openstack.org/387181 | 06:33 |
*** skazi has joined #openstack-ironic | 06:38 | |
*** skazi has quit IRC | 06:45 | |
*** skazi has joined #openstack-ironic | 06:45 | |
*** jtomasek_ has quit IRC | 06:47 | |
openstackgerrit | octopuszhang proposed openstack/ironic: Remove unnecessary '.' before ':' in ironic rst https://review.openstack.org/387188 | 06:50 |
*** e0ne has joined #openstack-ironic | 06:53 | |
*** fxpester has joined #openstack-ironic | 06:54 | |
*** pester has joined #openstack-ironic | 06:58 | |
*** fxpester has quit IRC | 06:59 | |
*** Sukhdev has quit IRC | 07:01 | |
*** pmannidi has quit IRC | 07:01 | |
*** tesseract has joined #openstack-ironic | 07:03 | |
*** tesseract is now known as Guest21077 | 07:04 | |
*** skazi has quit IRC | 07:12 | |
*** hshiina has joined #openstack-ironic | 07:15 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip create_ovs_taps() for multitenancy case https://review.openstack.org/382360 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update iptables rules and services IPs for multinode https://review.openstack.org/368648 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add devstack setup_vxlan_network() https://review.openstack.org/368647 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip some steps for multinode case https://review.openstack.org/368646 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update ironic node names for multinode case https://review.openstack.org/368645 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update devstack provision net config for multihost https://review.openstack.org/368644 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip db configuration on subnodes https://review.openstack.org/353303 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Testing multinode stuff https://review.openstack.org/368173 | 07:15 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ignore required_services for multinode topology https://review.openstack.org/352793 | 07:15 |
*** e0ne has quit IRC | 07:23 | |
pas-ha | wow :) | 07:23 |
pas-ha | morning Iornic | 07:23 |
*** openstackgerrit has quit IRC | 07:34 | |
*** openstackgerrit has joined #openstack-ironic | 07:34 | |
*** jpich has joined #openstack-ironic | 07:35 | |
*** skazi has joined #openstack-ironic | 07:36 | |
*** dtantsur|afk is now known as dtantsur | 07:42 | |
dtantsur | Morning Ironic, morning pas-ha | 07:42 |
tangchen | Morning, dtantsur, pas-ha, and Ironic. | 07:43 |
*** amotoki has joined #openstack-ironic | 07:46 | |
*** ohamada has joined #openstack-ironic | 07:55 | |
*** zzzeek has quit IRC | 08:00 | |
*** pmannidi has joined #openstack-ironic | 08:00 | |
*** zzzeek has joined #openstack-ironic | 08:01 | |
*** Romanenko_K has joined #openstack-ironic | 08:18 | |
*** bcornec1 has quit IRC | 08:19 | |
*** bcornec1 has joined #openstack-ironic | 08:22 | |
*** e0ne has joined #openstack-ironic | 08:24 | |
*** athomas has joined #openstack-ironic | 08:25 | |
*** electrofelix has joined #openstack-ironic | 08:26 | |
*** ralonsoh has joined #openstack-ironic | 08:32 | |
*** portdirect has joined #openstack-ironic | 08:38 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-ui: Imported Translations from Zanata https://review.openstack.org/385832 | 08:50 |
*** skazi has quit IRC | 08:50 | |
*** ccamacho has joined #openstack-ironic | 08:51 | |
*** lindycoder has joined #openstack-ironic | 08:53 | |
*** lindycoder has quit IRC | 08:57 | |
*** derekh has joined #openstack-ironic | 08:59 | |
*** milan has joined #openstack-ironic | 08:59 | |
milan | morning Ironic! :) | 08:59 |
*** appu103_ has joined #openstack-ironic | 09:00 | |
*** lindycoder has joined #openstack-ironic | 09:00 | |
*** appu103_ is now known as aparnav | 09:00 | |
*** portdirect has quit IRC | 09:02 | |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-python-agent: Remove old lookup/heartbeat from IPA https://review.openstack.org/386052 | 09:03 |
*** skazi has joined #openstack-ironic | 09:07 | |
*** pcaruana has joined #openstack-ironic | 09:08 | |
*** pmannidi has quit IRC | 09:08 | |
aarefiev | morning ! | 09:10 |
*** zackf has joined #openstack-ironic | 09:11 | |
*** lindycoder has quit IRC | 09:11 | |
*** skazi has quit IRC | 09:12 | |
*** zackf has quit IRC | 09:16 | |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: Help a user to enable console redirection https://review.openstack.org/384578 | 09:17 |
*** getvasanth_ has joined #openstack-ironic | 09:18 | |
*** getvasanth has quit IRC | 09:19 | |
*** xuao has joined #openstack-ironic | 09:22 | |
milan | morning aarefiev! :) | 09:24 |
xuao | morning milan~ aarefiev~ | 09:24 |
*** skazi has joined #openstack-ironic | 09:25 | |
milan | morning xuao! :) | 09:25 |
*** chlong has quit IRC | 09:26 | |
aarefiev | morning milan, thanks for pointing to discussion , I've replied | 09:26 |
aarefiev | I didn't see new version, hope will take a look in 1h, long weekends ) | 09:27 |
milan | aarefiev, I saw, thanks! | 09:27 |
aarefiev | morning xuao | 09:27 |
milan | aarefiev, no new version, I just split the patch based on dtantsur 's suggestion | 09:27 |
*** zackf has joined #openstack-ironic | 09:27 | |
milan | aarefiev, I'd like to settle down the discussion first but it's still my focus to push it forward | 09:28 |
milan | * the endpoint as a whole | 09:28 |
aarefiev | milan: ok, on new patch set then, I see new patch | 09:29 |
aarefiev | aha, ok | 09:29 |
milan | aarefiev, dtantsur wanted to have the same data on both the endpoints (get&list) I think it's a lgtm concern | 09:30 |
xuao | I have a question about the ironic console. If the ipmitool support to start several sol consoles? And users are able to use consoles in the same time. | 09:31 |
*** zackf has quit IRC | 09:32 | |
sambetts | Morning all | 09:37 |
milan | morning sambetts! :) | 09:39 |
*** jaosorior has quit IRC | 09:41 | |
sambetts | Hey milan! | 09:41 |
*** jaosorior has joined #openstack-ironic | 09:42 | |
*** skazi has quit IRC | 09:42 | |
dtantsur | morning milan, aarefiev, sambetts | 09:42 |
dtantsur | xuao, I don't know a precise answer, but I suspect it's "no" | 09:43 |
*** skazi has joined #openstack-ironic | 09:43 | |
xuao | dtantsur, thanks so much!!!! And I've tried many configurations for it but no use~~~ | 09:46 |
openstackgerrit | Merged openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/385746 | 09:46 |
milan | aarefiev, dtantsur, sambetts if you have time, please check also: https://review.openstack.org/#/c/386703/ thanks! | 09:48 |
patchbot | patch 386703 - ironic-inspector - Enhance introspection status | 09:48 |
dtantsur | unfortunately, due to downstream reasons you should be aware of, I won't probably have much time today.. will try though | 09:49 |
tuanluong | hello ironic team | 09:51 |
milan | morning tuanluong! :) | 09:53 |
sambetts | milan: I'll take a look thanks :) | 09:58 |
vdrok | good morning everyone | 10:02 |
aarefiev | morning dtantsur, sambetts, tuanluong, vdrok | 10:03 |
*** tiendc has quit IRC | 10:07 | |
sambetts | milan: QQ in this patch the iso8601 library is only used for tests, perhaps we should only add it in the test-requirements.txt file | 10:10 |
openstackgerrit | Tang Chen proposed openstack/python-ironicclient: Add missing unit tests for provision state commands https://review.openstack.org/377607 | 10:14 |
openstackgerrit | Tang Chen proposed openstack/python-ironicclient: Trivial: Rename and reorder deploy baremetal test https://review.openstack.org/378140 | 10:14 |
*** zackf has joined #openstack-ironic | 10:14 | |
*** zackf has quit IRC | 10:18 | |
milan | sambetts, correct, I can fix it. In follow-ups that would change, though (filtering on timestamps will have to parse it, so we'll be moving that back) | 10:22 |
milan | morning vdrok! :) | 10:23 |
*** hoangcx has quit IRC | 10:27 | |
milan | sambetts, also do you think it's a -1 thing? ;) | 10:28 |
*** getvasanth_ has quit IRC | 10:29 | |
sambetts | milan: I was juist surprised that it wasn't being used anywhere else, the datetime std lib can output iso time I'm surprised it can't import them | 10:30 |
*** mgould has joined #openstack-ironic | 10:31 | |
mgould | morning Ironic! | 10:32 |
mgould | is this backport still mergeable? https://review.openstack.org/#/c/367346 | 10:32 |
patchbot | patch 367346 - python-tripleoclient (stable/mitaka) - Update baremetal ready state command | 10:32 |
milan | sambetts, yeah, maybe I've missed some date time lib part but I think python has an issue with time handling spread across 3+ libraries | 10:32 |
* mgould winces at the commit message | 10:32 | |
milan | mgould, morning! :) | 10:33 |
dtantsur | mgould, better ask on #tripleo | 10:33 |
mgould | morning milan sambetts dtantsur | 10:33 |
mgould | dtantsur: good point, thanks | 10:33 |
*** xuao has quit IRC | 10:34 | |
sambetts | milan: I see a number of different ways to handle it, none of them are standardised :-P I see a lot of people recomending the dateutils library though | 10:40 |
*** mmedvede has quit IRC | 10:41 | |
sambetts | milan: I'm so surprised that there is no std lib for it, but oh well | 10:41 |
milan | sambetts, tell me about it :D | 10:42 |
milan | I had my portion of fun already with that :D | 10:42 |
mgould | milan: what are you trying to do? | 10:43 |
milan | sambetts, cool, so lemme rephrase that in the terms of dateutil | 10:43 |
milan | mgould, parse&dump datetime to/from ISO8601 strings | 10:44 |
mgould | milan: doesn't datetime.strptime do that? | 10:45 |
* mgould had to do this over the weekend to fix up EXIF data from his .kg photos | 10:45 | |
mgould | turns out all three of our cameras were set to different time zones... | 10:46 |
milan | mgould, was there a pattern? or all the cameras had the same filename conventions | 10:47 |
milan | mgould, datetime.parser.parse | 10:47 |
milan | strptime would require me to specify the format, which I'm too lazy to do :D | 10:47 |
milan | parser should parse ISO OK | 10:48 |
milan | I'm not sure why I've not used that right away :-/ | 10:48 |
milan | on a more positive note: https://review.openstack.org/#/c/383862/ has 2x+1 maybe we'll have a decision this week | 10:49 |
patchbot | patch 383862 - api-wg - Specify time intervals based filtering queries | 10:49 |
sambetts | \o/ | 10:51 |
*** zhangyufei has quit IRC | 10:52 | |
*** tuanluong has quit IRC | 10:53 | |
mgould | milan: \o/ | 10:53 |
milan | :D | 10:53 |
mgould | milan: nah, they all used different filename conventions | 10:54 |
sambetts | milan: it would be nice to get a standard iso parser built into oslo.utils or something (if there isn't already one) | 10:54 |
mgould | I wrote a script to read EXIF data, fix up the timestamp, then rename the file to the new timestamp | 10:54 |
sambetts | mgould: Nice! | 10:54 |
milan | mgould, at least there was a pattern then, wasn't it | 10:54 |
mgould | milan: there was when I'd finished with them :-) | 10:55 |
*** mmedvede has joined #openstack-ironic | 10:55 | |
milan | sambetts, there is actually: http://paste.openstack.org/show/585922/ | 10:56 |
* milan facepalm | 10:56 | |
milan | mgould, ah :D | 10:56 |
*** phuongnh has quit IRC | 10:59 | |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: IPMI command should depend on console type https://review.openstack.org/363647 | 11:01 |
sambetts | milan: ;) oops | 11:07 |
milan | sambetts, -1 please | 11:08 |
milan | I shall correct that | 11:08 |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: IPMI command should depend on console type https://review.openstack.org/363647 | 11:08 |
*** narasimha_SV has joined #openstack-ironic | 11:31 | |
narasimha_SV | https://bugs.launchpad.net/ironic/+bug/1633575 | 11:31 |
openstack | Launchpad bug 1633575 in Ironic "unable to reboot the server after imaging it" [Undecided,New] | 11:31 |
*** hoangcx has joined #openstack-ironic | 11:38 | |
*** dprince has joined #openstack-ironic | 11:43 | |
openstackgerrit | Deepak proposed openstack/ironic-ui: Update .coveragerc after the removal of openstack directory https://review.openstack.org/387380 | 11:48 |
nicodemos | good morning, Ironicers! | 12:00 |
sambetts | narasimha_SV: there are two possiblities with that bug, either your server isn't responding to the power on command and therefore isn't powering back on, or its taking longer than the configured timeout to turn on | 12:01 |
sambetts | narasimha_SV: is the physical machine doing the requested reboot? | 12:02 |
*** pece has joined #openstack-ironic | 12:02 | |
*** bfournie has quit IRC | 12:03 | |
narasimha_SV | sambetts: not sure about that | 12:05 |
narasimha_SV | i even tried manually sometimes it is not able to power on by manual commands | 12:05 |
narasimha_SV | i tried the same command from the log | 12:06 |
narasimha_SV | what is the default timeout to check turn on ? | 12:06 |
narasimha_SV | because i logs i can see it is trying for more times to check the status for power on | 12:07 |
lucasagomes | morning all | 12:08 |
*** jcoufal has joined #openstack-ironic | 12:08 | |
mgould | morning lucasagomes nicodemos narasimha_SV | 12:09 |
narasimha_SV | good morning mgould | 12:10 |
sambetts | [ipmi] retry_timeout and min_command_interval, are two settings you can increase to increase the amount of time Ironic will try to talk to your server, and increase the time between calls to your server incase your server can't handle the amount of calls being made | 12:11 |
narasimha_SV | sambetts: ipmitool works for all kind of servers right ? i mean like HP,DELL and other | 12:11 |
narasimha_SV | because I am using oracle sun servers | 12:11 |
dtantsur | lucasagomes, morning! do we need more changes to https://bugs.launchpad.net/ironic/+bug/1561137 before we close it? | 12:12 |
openstack | Launchpad bug 1561137 in ironic-python-agent "[RFE] Add support for operators on root device hints" [Wishlist,In progress] - Assigned to Lucas Alvares Gomes (lucasagomes) | 12:12 |
lucasagomes | dtantsur, only docs, there's a patch up for it | 12:13 |
lucasagomes | dtantsur, but we can close it for IPA and ironic-lib | 12:13 |
dtantsur | lucasagomes, k, please do | 12:13 |
lucasagomes | ack | 12:14 |
sambetts | narasimha_SV: ipmitool speaks the IPMI standard, which is supported by many vendors, but we have found that the hardware that some vendors provide for handling IPMI on there severs can sometime break when hit with to many commands, so thats why we provide the configurations so that you can increase the timeouts and time between IPMI calls | 12:16 |
*** vicky has quit IRC | 12:16 | |
*** gcb has quit IRC | 12:17 | |
narasimha_SV | ok sure i will try by increasing timeouts and will verify | 12:17 |
*** sai has quit IRC | 12:19 | |
*** rook_ has quit IRC | 12:20 | |
narasimha_SV | what is the default time for the ipmicalls here ? | 12:20 |
TheJulia | hogepodge: how do you mean use the system pip? As in not use env-setup and attempt to directly execute the playbooks? | 12:20 |
*** Guest23990 is now known as zigo | 12:21 | |
sambetts | the default is a 60 second timeout and a 5 second command interval | 12:21 |
narasimha_SV | ok | 12:21 |
TheJulia | hogepodge: I'm on Xenial on my dev machine now and haven't had any pip issues, but I also use env-setup and execute the test scripts which run env-setup | 12:22 |
*** zackf has joined #openstack-ironic | 12:22 | |
xavierr | good morning Ironic | 12:24 |
dtantsur | morning TheJulia, xavierr | 12:24 |
TheJulia | good morning dtantsur xavierr | 12:25 |
*** rook has joined #openstack-ironic | 12:26 | |
*** rook is now known as Guest40693 | 12:26 | |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Do not pass ipa-driver-name as kernel parameter https://review.openstack.org/387402 | 12:26 |
TheJulia | narasimha_SV: Is it a single host with a single ipmi endpoint, no bridging to nodes, such as with a blade chassis? | 12:26 |
*** sai has joined #openstack-ironic | 12:27 | |
jroll | morning y'all | 12:27 |
sambetts | o/ jroll, TheJulia | 12:27 |
TheJulia | narasimha_SV: If your not already running with debug logging, you may want to. We've seen some fun error messages come back from chassis. :) | 12:28 |
TheJulia | good morning jroll sambetts | 12:28 |
nicodemos | morning mgould, dtantsur, jroll, TheJulia. | 12:28 |
TheJulia | good morning nicodemos | 12:28 |
*** dtantsur is now known as dtantsur|mtg | 12:30 | |
mgould | morning TheJulia xavierr jroll | 12:30 |
TheJulia | good morning mgould | 12:32 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Remove workaround for test_network_basic_ops https://review.openstack.org/387403 | 12:32 |
dtantsur|mtg | checking if this works again by chance ^^^ | 12:32 |
lucasagomes | xavierr, TheJulia morning | 12:34 |
TheJulia | good morning lucasagomes | 12:39 |
mat128 | good morning ironic | 12:41 |
*** trown is now known as trown|brb | 12:43 | |
*** bfournie has joined #openstack-ironic | 12:43 | |
*** baoli has joined #openstack-ironic | 12:46 | |
gabriel-bezerra | Hi. Is there any restriction for the kind of neutron network driver to be used with ironic? I could only find docs mentioning ovs, but the current default for neutron is linuxbridge. | 12:46 |
sambetts | gabriel-bezerra: either of those network drivers will work for your controller/network node and Ironic configured to use flat networking, however if you want to use Ironic multitenancy then you need an Ml2 driver that supports it | 12:47 |
gabriel-bezerra | I see. I'm having issues with some ironic-neutron integration in my deployment. I'll look further. | 12:49 |
gabriel-bezerra | thank you, sambetts | 12:50 |
*** srobert has joined #openstack-ironic | 12:50 | |
jroll | lucasagomes: jfyi, apparently unofficial projects are meant to manually tag, I just learned https://review.openstack.org/#/c/386734 | 12:51 |
patchbot | patch 386734 - releases - remove ironic-staging-drivers | 12:51 |
sambetts | jroll: yup :( | 12:51 |
lucasagomes | jroll, ouch :-/ | 12:52 |
sambetts | I have to do it all for networking-cisco | 12:52 |
lucasagomes | jroll, it was easier to just add a patch to that repo | 12:53 |
jroll | lucasagomes: see -release channel, they want to make a way to do it (or use scripts manually), it just isn't done yet | 12:53 |
lucasagomes | jroll, thanks for the heads-up | 12:53 |
* lucasagomes joins the channel | 12:53 | |
sambetts | same | 12:53 |
jroll | lucasagomes: you were already there when I said that :P | 12:54 |
jroll | sambetts: was just a quick conversation, see eavesdrop if you want to read all of 10 lines | 12:54 |
lucasagomes | jroll, yeah, just read the scrollback | 12:54 |
sambetts | jroll: ah thanks :) | 12:54 |
*** strigazi1AFK has joined #openstack-ironic | 12:55 | |
*** rloo has joined #openstack-ironic | 12:57 | |
*** strigazi_AFK has quit IRC | 12:58 | |
*** ppiela has joined #openstack-ironic | 13:01 | |
*** Goneri has joined #openstack-ironic | 13:02 | |
lucasagomes | vdrok, heads up, let's release ironic-staging-drivers after the summit (or even end of this week) ? The ansible driver is in now, we can try to merge whatever is left this week | 13:02 |
vdrok | lucasagomes: morning, yup, agree | 13:02 |
vdrok | I'll take a look at what's in review queue today | 13:02 |
vdrok | iirc there are some docs that would be good to merge too | 13:03 |
openstackgerrit | Mark Hamzy proposed openstack/ironic-specs: Add initial specification for MoltenIron https://review.openstack.org/387418 | 13:06 |
lucasagomes | ++ | 13:07 |
TheJulia | interesting... | 13:13 |
*** lindycoder has joined #openstack-ironic | 13:13 | |
*** lindycoder has quit IRC | 13:13 | |
*** trown|brb is now known as trown | 13:14 | |
jroll | does... does that need a spec? | 13:15 |
jroll | mjturek: is that a proposal to put molteniron in the ironic governance or? | 13:16 |
sambetts | lucasagomes: have you seen this before? http://logs.openstack.org/08/377608/4/check/gate-tempest-dsvm-ironic-ipa-wholedisk-pxe_ssh-tinyipa-nv/2a2a65e/logs/screen-ir-cond.txt.gz#_2016-10-05_14_51_03_458 | 13:16 |
jroll | sambetts: that should be fixed by latest ironic-lib, give it a recheck | 13:17 |
sambetts | lucasagomes: it looks like its coming from the create_config_drive_partition function in ironic-lib | 13:17 |
sambetts | jroll: oh cool thanks :) | 13:17 |
jroll | :) | 13:17 |
mjturek | jroll: yep, we're proposing molteniron as a sub-project | 13:17 |
openstackgerrit | Stephanie Miller proposed openstack/bifrost: Enable inspector discovery by default https://review.openstack.org/386701 | 13:18 |
*** pcaruana has quit IRC | 13:18 | |
sambetts | jroll: its really hard to follow some of those tracebacks when we catch it and raise a new exception without logging the orginal exception anywhere | 13:18 |
mjturek | jroll: is that the wrong place for something like that? | 13:19 |
jroll | mjturek: we've never done a spec for that before ¯\_(ツ)_/¯ | 13:19 |
jroll | mjturek: usually it's just a poll of the core team in a meeting or something | 13:19 |
jroll | mjturek: what benefit does molteniron get from being a sub-project, as opposed to just an unofficial project? | 13:20 |
mjturek | jroll: well, we'd like to be under the openstack governance and thought ironic would be the right place to propose it. | 13:22 |
jroll | mjturek: don't disagree, just curious if/why it should be under openstack governance? (note that it can be in openstack/ git tree, use openstack CI, etc without being under openstack governance) | 13:24 |
sambetts | if I remember correctly being under another projects tree means that its owned by that projects core team, the goverence umbrellas don't just indicate "related too", they indicate ownership and responsiblity | 13:25 |
jroll | right | 13:25 |
sambetts | jroll: advantages on being offical, include things like being able to publish docs to docs.openstack | 13:25 |
jroll | good point | 13:26 |
sambetts | literally the main reason I'm pushing for networking-cisco to regain offical status | 13:26 |
sambetts | is for docs | 13:26 |
jroll | I wonder if folks could revisit that | 13:26 |
openstackgerrit | Kyrylo Romanenko proposed openstack/ironic: Test ML2 Neutron Events feature https://review.openstack.org/370016 | 13:27 |
jroll | (TC etc) | 13:27 |
mjturek | sambetts: yeah very good point :) | 13:27 |
dtantsur|mtg | morning jroll | 13:28 |
jroll | \o dtantsur|mtg | 13:28 |
mjturek | jroll: sambetts: I was under the impression that we couldn't use official gerrit, launchpad, etc without being under the governance as well. Am I wrong? | 13:30 |
jroll | mjturek: yes, that's incorrect | 13:31 |
mjturek | ahhh okay, didn't realize that | 13:31 |
*** makowals_ has joined #openstack-ironic | 13:31 | |
jroll | lemme see if I can find a list | 13:31 |
mjturek | right now, we're living in the third-party-ci-tools repo. Ideally we'd want to be our own | 13:31 |
mjturek | thanks :) | 13:31 |
*** makowals has quit IRC | 13:32 | |
sambetts | nope goverence is completely separated from the git repos and launchpad etc, the only things you miss out on are publishing to docs.openstack.org and publishing your releases to release.openstack.org | 13:34 |
mjturek | oh wow | 13:35 |
jroll | also specs.o.o | 13:35 |
sambetts | ah yes bascially *.openstack.org publishing | 13:36 |
*** aparnav has quit IRC | 13:38 | |
jroll | mjturek: this is the best I have: http://docs.openstack.org/infra/manual/creators.html | 13:39 |
mjturek | jroll: yeah actually looking at that now, thank you! | 13:39 |
jroll | mjturek: if you ctrl-f 'official' there, you'll see the few exceptions that are 'official only' | 13:39 |
jroll | e.g. docs, governance patch, etc | 13:39 |
sambetts | ++ | 13:39 |
krtaylor | jroll, we were thinking about a separate project to get it out from under third-party-ci repo so that we'd get infra testing etc, especially now that other teams are using it for their test systems | 13:40 |
jroll | krtaylor: yeah, separate project for sure, just curious if it needs to be in ironic governance or not | 13:41 |
*** trown is now known as trown|brb | 13:41 | |
krtaylor | from the early discussions, we felt that it made more sense to be under the ironic tree since it was for baremetal | 13:41 |
sambetts | jroll, mjturek: http://docs.openstack.org/infra/manual/drivers.html | 13:41 |
sambetts | thats another useful guide ^ | 13:41 |
*** rama_y has joined #openstack-ironic | 13:42 | |
krtaylor | it would give it a broader audience for enhancements, reviews etc | 13:42 |
krtaylor | which would be harder to get if on its own | 13:42 |
jlvillal | Good morning Ironic :) | 13:42 |
mjturek | krtaylor: right but these are things we might be able to accomplish without openstack governance. It seems that the main benefit of being an official project is having docs on docs.openstack.org | 13:44 |
jroll | russellb reminded me about ATC/PTG space | 13:44 |
openstackgerrit | Hugo Nicodemos proposed openstack/python-oneviewclient: Refactor the validation to Server Profile Template https://review.openstack.org/377680 | 13:45 |
jlvillal | jroll: And the email :) | 13:45 |
jroll | email? | 13:45 |
jlvillal | mgould: That 'hypothesis" thing looks interesting. | 13:45 |
jlvillal | jroll: Email to openstack-dev to say Ironic will be attending PTG. | 13:45 |
mgould | jlvillal: yeah, it's very cool | 13:46 |
jroll | jlvillal: http://i.imgur.com/GtY9CiC.png | 13:46 |
sambetts | haha | 13:46 |
jlvillal | jroll: heh | 13:46 |
mgould | I've used it on a couple of small projects, but don't really know where we could start applying it in Ironic | 13:46 |
* jlvillal needs to figure out a good todo tracker and then actually use it | 13:46 | |
mgould | randomly hit API endpoints and assert that we never see a 5xx error | 13:46 |
jroll | jlvillal: I've bounced around different tools quite a bit, this has been working okay for me: https://www.rememberthemilk.com/ | 13:47 |
jlvillal | jroll: Thanks. I'll take a look. | 13:47 |
jlvillal | mgould: I think it could be used in the unit testing area maybe. | 13:48 |
jroll | jlvillal: + setting a 'start date' on things, and a view to only show me things where I'm past the start date, so I don't load up a million things in one view | 13:48 |
charlledaniel | Morning all | 13:48 |
jroll | that was always my problem, put everything in the list then get overwhelmed | 13:49 |
jroll | morning charlledaniel | 13:49 |
jlvillal | jroll: Ah, okay. I'm looking at the website now. | 13:49 |
*** snozhka has joined #openstack-ironic | 13:49 | |
charlledaniel | May someone review my patch https://review.openstack.org/#/c/372762/? | 13:50 |
patchbot | patch 372762 - ironic - Ensures that OneView nodes are free for use by Ironic | 13:50 |
* jlvillal discovers that work is very quiet at 6am | 13:51 | |
mrtenio | Morning! :D | 13:51 |
snozhka | Hi guys, I have a question regarding ironic-inspector tempest test, could anyone help me with it ? | 13:51 |
jlvillal | snozhka: Probably depends on the question ;) Though I am unlikely to be able to help :( | 13:52 |
jlvillal | Others might. | 13:52 |
jlvillal | snozhka: So I recommend asking the question and hopefully someone can help. | 13:53 |
snozhka | The problem is: while running InspectorBasicTest test on my local env, it fails during boot with the following exception - Details: Timed out waiting for node 7c160d83-916b-4318-a91d-2d0f65933d15 to reach power_state state(s) ['power on'] | 13:53 |
*** makowals_ has quit IRC | 13:54 | |
snozhka | after some investigation the following error was found in ironic-conductor.log | 13:54 |
snozhka | 2016-10-13 11:09:48.892 2979 ERROR ironic.conductor.manager [req-1d575adb-4d36-40c7-bde9-7296cec67ac9 - - - - -] Error while preparing to deploy to node 7c160d83-916b-4318-a91d-2d0f65933d15: No VIFs found for node 7c160d83-916b-4318-a91d-2d0f65933d15 when attempting to update DHCP BOOT options. | 13:55 |
snozhka | it's weird because i can boot instance manually, but test fails | 13:55 |
krtaylor | jroll, can we put the molteniron topic on the agenda for todays meeting? | 13:57 |
jroll | krtaylor: sure | 13:58 |
krtaylor | thx, mjturek ^^ | 13:58 |
*** portdirect has joined #openstack-ironic | 13:58 | |
*** makowals has joined #openstack-ironic | 13:59 | |
mjturek | throwing it under RFE review for now | 13:59 |
* krtaylor is afk for a bit | 14:00 | |
mgould | jlvillal: what kind of properties could we unit-test? | 14:00 |
*** anush has joined #openstack-ironic | 14:00 | |
*** rama_y has quit IRC | 14:01 | |
*** dtantsur|mtg is now known as dtantsur | 14:01 | |
*** rama_y has joined #openstack-ironic | 14:02 | |
*** hogepodge has quit IRC | 14:04 | |
*** krtaylor has quit IRC | 14:04 | |
mrtenio | Could someone take a look on my patch? Thank you. https://review.openstack.org/#/c/377073/ | 14:09 |
patchbot | patch 377073 - ironic - Adds another validation step when using dynamic al... | 14:09 |
jlvillal | mgould: Well when I was reading the docs they made it look easy :) But not exactly sure how to use it in the unit tests. | 14:10 |
*** anush has quit IRC | 14:10 | |
mgould | jlvillal: yeah :-) I'll have a think over the next week, and we can brainstorm some ideas at the summit | 14:11 |
*** trown|brb is now known as trown | 14:11 | |
jlvillal | mgould: Sounds good. Thanks. | 14:11 |
*** mtanino has joined #openstack-ironic | 14:12 | |
openstackgerrit | Hugo Nicodemos proposed openstack/python-oneviewclient: Refactor the validation to Server Profile Template https://review.openstack.org/377680 | 14:13 |
*** Goneri has quit IRC | 14:14 | |
*** hshiina has quit IRC | 14:15 | |
*** Goneri has joined #openstack-ironic | 14:15 | |
*** lucasagomes is now known as lucas-hungry | 14:18 | |
*** ohamada_ has joined #openstack-ironic | 14:20 | |
*** ohamada has quit IRC | 14:20 | |
*** skazi has quit IRC | 14:22 | |
*** skazi has joined #openstack-ironic | 14:23 | |
openstackgerrit | Szymon Borkowski proposed openstack/ironic-specs: WIP: Add VCA hardware type to Ironic https://review.openstack.org/369309 | 14:29 |
*** michauds has joined #openstack-ironic | 14:31 | |
*** srobert has quit IRC | 14:33 | |
*** ppiela has quit IRC | 14:33 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic-python-agent: Fix config drive writing script https://review.openstack.org/385957 | 14:35 |
vdrok | lucas-hungry: https://review.openstack.org/385957 this definitely works as expected now, but there is an ugly sleep there, any suggestions welcome :) | 14:39 |
patchbot | patch 385957 - ironic-python-agent - Fix config drive writing script | 14:39 |
openstackgerrit | Merged openstack/python-oneviewclient: Refactor the validation to Server Profile Template https://review.openstack.org/377680 | 14:45 |
dtantsur | aarefiev, around? mind helping snozhka figuring out why he can't run inspector tempest plugin? /me is jammed today, sorry | 14:48 |
aarefiev | I'm here | 14:49 |
*** dprince has quit IRC | 14:49 | |
aarefiev | hey snozhka, whats failures do you see | 14:49 |
snozhka | aarefiev, hi | 14:50 |
snozhka | <snozhka> The problem is: while running InspectorBasicTest test on my local env, it fails during boot with the following exception - Details: Timed out waiting for node 7c160d83-916b-4318-a91d-2d0f65933d15 to reach power_state state(s) ['power on'] | 14:50 |
snozhka | * makowals_ has quit (Quit: My MacBook Air has gone to sleep. ZZZzzz…) | 14:50 |
snozhka | <snozhka> after some investigation the following error was found in ironic-conductor.log | 14:50 |
snozhka | <snozhka> 2016-10-13 11:09:48.892 2979 ERROR ironic.conductor.manager [req-1d575adb-4d36-40c7-bde9-7296cec67ac9 - - - - -] Error while preparing to deploy to node 7c160d83-916b-4318-a91d-2d0f65933d15: No VIFs found for node 7c160d83-916b-4318-a91d-2d0f65933d15 when attempting to update DHCP BOOT options. | 14:50 |
snozhka | <snozhka> it's weird because i can boot instance manually, but test fails | 14:50 |
snozhka | aarefiev, ^^^^ | 14:50 |
aarefiev | reading | 14:53 |
aarefiev | hmm, strange it says there is no ports for this node | 14:55 |
snozhka | but it present | 14:56 |
aarefiev | snozhka: so if you says it works manually, did you configure tempest | 14:56 |
snozhka | is it requires to add port id into tempest.conf file ? | 14:56 |
sambetts | its says there is no VIF, that means no neutron port assigned to the Ironic port | 14:56 |
snozhka | aarefiev, yes i do, the introspection works file | 14:56 |
sambetts | but Ironic shouldn't be trying to configure that for an inspection if I remember correctly | 14:57 |
aarefiev | sambetts: but it works if run manually | 14:57 |
aarefiev | yeah | 14:57 |
dtantsur | snozhka, wait, what exactly are you running manually, introspection or deployment? | 14:58 |
aarefiev | sambetts : and when booting instance | 14:58 |
dtantsur | your tempest fails on deployment | 14:58 |
snozhka | dtantsur, i'm booting instance manually | 14:58 |
*** lucas-hungry is now known as lucasagomes | 14:58 | |
dtantsur | snozhka, via nova boot? | 14:58 |
snozhka | introspection works file from test itself | 14:58 |
snozhka | dtantsur, yes, correct | 14:58 |
dtantsur | hmm | 14:59 |
openstackgerrit | yolanda.robla proposed openstack/bifrost: Add instructions to deploy bifrost on virsh https://review.openstack.org/383792 | 15:00 |
aarefiev | snozhka: is it possible to post logs somewhere | 15:02 |
snozhka | aarefiev, sure, which logs you want ? | 15:02 |
aarefiev | inspector and ironic at least | 15:03 |
*** pcaruana has joined #openstack-ironic | 15:03 | |
sambetts | aarefiev, snozhka: If you want to upload the logs somewhere, I suggest opening a launchpad bug for inspector, and you can upload them as attachments | 15:03 |
aarefiev | yeah, good idea | 15:03 |
*** fxpester has joined #openstack-ironic | 15:04 | |
*** penick has quit IRC | 15:05 | |
*** mjura has quit IRC | 15:05 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Set workaround for test_network_basic_ops in grenade setting https://review.openstack.org/387403 | 15:06 |
*** pester has quit IRC | 15:07 | |
*** syed_ has joined #openstack-ironic | 15:08 | |
openstackgerrit | melissaml proposed openstack/ironic: Fix typos in documents https://review.openstack.org/387524 | 15:12 |
snozhka | aarefiev, sambetts - https://bugs.launchpad.net/ironic/+bug/1634175 | 15:19 |
openstack | Launchpad bug 1634175 in Ironic "Ironic-inspector tempest test fails while instance boot" [Undecided,New] | 15:19 |
sambetts | moved it to ironic-inspectors bug tracker | 15:20 |
sambetts | I have * | 15:20 |
snozhka | sambetts, thank you! | 15:21 |
aarefiev | thanks, I'm looking | 15:22 |
snozhka | aarefiev, thanks! | 15:23 |
*** rcernin has quit IRC | 15:26 | |
sambetts | snozhka: quick question, what driver are you using for the node in this test environment? | 15:29 |
snozhka | sambetts, pxe_ssh | 15:29 |
sambetts | snozhka: I don't see the error logs in the bug report description in the ironic-conductor or tempest logs, and I see log messages about the drac driver in the ironic logs | 15:32 |
snozhka | min | 15:32 |
aarefiev | I see Found network None available for tenant | 15:32 |
*** hoangcx has quit IRC | 15:34 | |
sambetts | aarefiev: that log appears in a passing tempest run here: http://logs.openstack.org/03/386703/1/check/gate-tempest-dsvm-ironic-inspector/06a50dd/logs/tempest.txt.gz | 15:34 |
sambetts | aarefiev: so I don't think thats an issue | 15:34 |
snozhka | sambetts, aarefiev uploaded full logs | 15:35 |
aarefiev | yeah, I see now it's not mandatory field | 15:35 |
*** gcb has joined #openstack-ironic | 15:37 | |
*** fxpester has quit IRC | 15:37 | |
*** krtaylor has joined #openstack-ironic | 15:37 | |
*** jvgrant has joined #openstack-ironic | 15:39 | |
JayF | vdrok: would the bug you're fixing here: https://review.openstack.org/#/c/385957/ also exist in ironic-lib? | 15:39 |
patchbot | patch 385957 - ironic-python-agent - Fix config drive writing script | 15:39 |
openstackgerrit | Mario Villaplana proposed openstack/ironic: Add power state change notifications https://review.openstack.org/321865 | 15:39 |
JayF | vdrok: if not, I'm really tempted to say the best fix is just to delete the script and use ironic-lib instead | 15:39 |
mgould | hi everyone | 15:39 |
mariojv | JayF: ^ there's the power state notifications patch with the logging changes if you wanted to review | 15:39 |
vsaienk0 | jroll, sambetts: mriedem requested a spec for static portgroups support https://review.openstack.org/#/c/387534/ | 15:40 |
patchbot | patch 387534 - nova-specs - Proposal for ironic static portgroups support | 15:40 |
dtantsur | who doesn't like writing nova specs? | 15:40 |
* dtantsur ducks | 15:40 | |
jroll | heh | 15:40 |
mgould | we're getting the error "No VIFs found for node [uuid] when attempting to update DHCP BOOT options" when running the InspectorBasicTest tempest test | 15:40 |
mgould | but we can deploy nodes manually without problems | 15:40 |
dtantsur | snozhka, mgould, aarefiev, I think it would be cool to get nova compute logs as well | 15:41 |
jroll | vsaienk0: thanks, I suspect it will need some detail on the configdrive changes, maybe sambetts can help you with that | 15:41 |
snozhka | dtantsur, adding... | 15:41 |
*** harlowja_at_home has joined #openstack-ironic | 15:41 | |
vsaienk0 | jroll: vdrok is working on configdrive part, I will ask him to add details to the spec | 15:42 |
mgould | does anyone know how tempest sets up the network for that test, or how it's meant to create a VIF for a node? | 15:42 |
jroll | vsaienk0: thanks | 15:42 |
aarefiev | btw does it devstack deployment? | 15:43 |
snozhka | dtantsur, nova-compute log uploaded | 15:43 |
dtantsur | jroll, marking http://specs.openstack.org/openstack/ironic-specs/specs/6.2/add-infiniband-support.html as complete was a mistake IMO | 15:45 |
dtantsur | jroll, at the very least, we have https://review.openstack.org/#/c/334816/ open | 15:45 |
patchbot | patch 334816 - ironic-python-agent - Infiniband Support for IPA | 15:45 |
jroll | dtantsur: oh, I saw the release note in ironic and thought it was done :( | 15:45 |
jroll | so | 15:45 |
jroll | it is done for ironic, but not inspector, yes? | 15:45 |
*** jistr is now known as jistr|biab | 15:46 | |
mgould | dtantsur, jroll yep: https://review.openstack.org/#/c/264257/ | 15:46 |
patchbot | patch 264257 - ironic-inspector - Adding InfiniBand Support | 15:46 |
dtantsur | maybe? I'm not sure if IPA part is only for inspector | 15:46 |
JayF | I'll review that right meow | 15:46 |
dtantsur | also lucasagomes just told me, it does not work :) | 15:46 |
jroll | dtantsur: the IPA patch seems to only collect info about infiniband nics | 15:46 |
jroll | dtantsur: oh, that's good | 15:46 |
lucasagomes | jroll, dtantsur what doesn't work is the validation I believe, we have a macaddress type in the API | 15:47 |
lucasagomes | which validates if the mac address has 6 octets | 15:47 |
snozhka | aarefiev, it doesn't | 15:47 |
lucasagomes | but the inifiniband GUID has more (idk how many gotta google it up) | 15:47 |
jroll | lucasagomes: oh, I thought they changed that :/ | 15:47 |
lucasagomes | jroll, lemme re-check, but I don't think so | 15:47 |
dtantsur | jroll, the spec touches upon inspector parts too, hence I'm not sure if we should close it yet | 15:48 |
lucasagomes | we have not https://github.com/openstack/ironic/blob/44bde8cd2d6e624cfa483cffc38d41571e5b9b6c/ironic/common/utils.py#L118 | 15:49 |
mgould | jroll, lucasagomes: I *definitely* remember reviewing code to validate MACs and GUIDs, not sure if it ever merged | 15:49 |
lucasagomes | which is called here: https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/types.py#L41 | 15:49 |
lucasagomes | mgould, strange | 15:49 |
openstackgerrit | Kyrylo Romanenko proposed openstack/ironic: Test ML2 Neutron Events feature https://review.openstack.org/370016 | 15:49 |
mgould | I guess it must be stuck in review limbo | 15:50 |
lucasagomes | probably | 15:50 |
* lucasagomes tries to find it | 15:50 | |
jroll | dtantsur: yeah, fair, it's weird because we base it on ironic version | 15:50 |
dtantsur | I see.. I'm fine either way | 15:50 |
jroll | dtantsur: we could add inspector/ipa to the rfe and mark it done for ironic | 15:51 |
*** athomas has quit IRC | 15:51 | |
jroll | which was already done for inspector :D | 15:51 |
dtantsur | cool :) | 15:52 |
*** david-lyle_ has joined #openstack-ironic | 15:52 | |
*** david-lyle has quit IRC | 15:53 | |
lucasagomes | mgould, dunno, it maybe a specific patch-set in a review as well | 15:53 |
lucasagomes | it's hard to track | 15:53 |
*** hogepodge has joined #openstack-ironic | 15:53 | |
mgould | lucasagomes: early versions of https://review.openstack.org/#/c/264257, looks like | 15:53 |
patchbot | patch 264257 - ironic-inspector - Adding InfiniBand Support | 15:53 |
openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add basic tests for OSC plugin baremetal node commands https://review.openstack.org/301687 | 15:54 |
lucasagomes | mgould, right | 15:54 |
mgould | holy cow, that patch has been in review since January | 15:54 |
lucasagomes | mgould, I will try to cook a patch for it | 15:54 |
openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Tests for OSC baremetal node create command with options https://review.openstack.org/382352 | 15:54 |
*** rama_y has quit IRC | 15:54 | |
dtantsur | yeah, maybe it got lost with the transition to new ramdisk API | 15:54 |
mgould | last patch set uploaded Oct 9 | 15:55 |
patchbot | 9 | 15:55 |
mgould | thanks, patchbot | 15:55 |
JayF | someone needs to patch patchbot | 15:56 |
JayF | I just feel like there's a bunch of vulns in there waiting to be exploited | 15:56 |
dtantsur | jroll, release management cross-project session overlaps with our neutron events session.. I guess I'll go to the former, and you - to the latter? | 15:56 |
*** athomas has joined #openstack-ironic | 15:56 | |
JayF | from how randomly it responses sometimes to strange characters | 15:56 |
*** hshiina has joined #openstack-ironic | 15:57 | |
jroll | dtantsur: sure | 15:57 |
dtantsur | k | 15:57 |
sambetts | dtantsur: I'm going to a bunch of the cross-project sessions too | 15:57 |
vdrok | JayF: morning, afk currently but I'll take a look at ironic lib today too | 15:57 |
vdrok | Re config drive change | 15:58 |
*** mandre has quit IRC | 15:59 | |
jroll | sambetts: sukhdev is out today, wanna run the neutron meeting? | 16:00 |
sambetts | jroll: he asked me too already, just heading over now | 16:00 |
jroll | sorry I saw him say that friday and literally just remembered | 16:00 |
jroll | ok | 16:00 |
sambetts | np | 16:03 |
*** pshedimb has joined #openstack-ironic | 16:06 | |
*** rcernin has joined #openstack-ironic | 16:10 | |
*** ChubYann has joined #openstack-ironic | 16:11 | |
*** rbrud has quit IRC | 16:12 | |
*** rbrud has joined #openstack-ironic | 16:12 | |
*** Vinsh_ has joined #openstack-ironic | 16:12 | |
*** Vinsh has quit IRC | 16:12 | |
openstackgerrit | Merged openstack/ironic: Clean up provision ports when reattempting deploy https://review.openstack.org/348392 | 16:13 |
jlvillal | Over on infra I saw this | 16:13 |
jlvillal | <anteaya> if folks are asking about how to use nodepool with ironic, wznoinsk shared his experience in the third party meeting earlier today: http://eavesdrop.openstack.org/meetings/third_party/2016/third_party.2016-10-17-15.00.log.html | 16:13 |
jlvillal | <anteaya> the ssh timing needs to be reconfigured for bare metal | 16:13 |
jlvillal | as an FYI | 16:13 |
wznoinsk | ping me if needs be | 16:14 |
openstackgerrit | Merged openstack/ironic: Remove unnecessary '.' before ':' in ironic rst https://review.openstack.org/387188 | 16:14 |
*** dprince has joined #openstack-ironic | 16:14 | |
*** mandre has joined #openstack-ironic | 16:15 | |
*** jpich has quit IRC | 16:18 | |
*** skazi has quit IRC | 16:20 | |
*** Vinsh has joined #openstack-ironic | 16:21 | |
*** Vinsh_ has quit IRC | 16:21 | |
*** jistr|biab is now known as jistr | 16:22 | |
*** Vinsh_ has joined #openstack-ironic | 16:23 | |
*** Vinsh has quit IRC | 16:23 | |
*** Vinsh_ has quit IRC | 16:23 | |
*** david-lyle_ is now known as david-lyle | 16:24 | |
*** Vinsh has joined #openstack-ironic | 16:24 | |
*** gcb has quit IRC | 16:24 | |
sambetts | wznoinsk: so that is for using baremetal as slaves for zuul? | 16:24 |
sambetts | wznoinsk: so you can do DSVM on bm ? | 16:25 |
*** gcb has joined #openstack-ironic | 16:25 | |
*** Vinsh has quit IRC | 16:26 | |
*** Vinsh has joined #openstack-ironic | 16:26 | |
wznoinsk | sambetts: yes, as long as you have local boot enabled for ironic nodes, otherwise you maybe booting a different kernel with different kernel package that are in the image (by dib) | 16:28 |
sambetts | wznoinsk: oh awesome :D thats some cool stuff! | 16:28 |
*** jtomasek has quit IRC | 16:28 | |
krtaylor | mjturek, ^^^ | 16:28 |
*** dprince has quit IRC | 16:28 | |
*** dprince has joined #openstack-ironic | 16:29 | |
*** Vinsh has quit IRC | 16:29 | |
sambetts | wznoinsk: is it possible to use node pool to reserve a BM node without the nodepool image being installed on it? | 16:29 |
wznoinsk | sambetts: correct, nodepool talks to a cloud provider, here' the provider is ironic which powers the physical server up and images the disk remotely (over iscsi with pxe_ipmitool driver) with the image from openstack (ironic) controller's glance during nova boot | 16:31 |
wznoinsk | openstack BM (ironic) controller is a VM with a few interfaces as hostdev from the host for things like 10Gbps network and one that is on the IPMI l2 with the physical servers | 16:32 |
wznoinsk | http://intel-openstack-ci-logs.ovh/98/356398/8/test/test-nova/e219dde/console.html | 16:33 |
sambetts | wznoinsk: I was just thinking for Ironic third party CI case, we have to manage a pool of resources to be given to the node running the tests, we have a couple of solutions for this already, one called molton iron, and other ones are DB based, just wondered if node pool could provide us with another | 16:33 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Account for Infiniband GID when validating an address https://review.openstack.org/387559 | 16:33 |
lucasagomes | sai, ^ | 16:33 |
*** rama_y has joined #openstack-ironic | 16:34 | |
sai | lucasagomes: ack, so should we be pulling in the changes and testing it? | 16:35 |
sai | or its still wip right | 16:35 |
lucasagomes | sai, yeah, I marked it as WIP until we conclude the tests | 16:35 |
sai | gotcha | 16:35 |
lucasagomes | since you have the envinronemnt, mind giving it a go ? | 16:35 |
sai | heading out to lunch lucasagomes lemme know if u need something from the env | 16:35 |
lucasagomes | sai, cool, ping me when you back | 16:35 |
wznoinsk | nodepool is talking to the API of a cloud controller and requesting specific image/flavor/network etc., you could probably wrap it with something that does things before/afterwards | 16:36 |
sai | rgr | 16:36 |
wznoinsk | sambetts: ^ | 16:36 |
sambetts | wznoinsk: so our case is that we want a VM running devstack, to be able to ask for a baremetal to provision for testing Ironic, so its kind of like multinode, except 1 node is a VM with nodepool on, and the second node is a blank baremetal | 16:38 |
sambetts | VM with the nodepool image on* | 16:38 |
sambetts | not nodepool itself | 16:38 |
*** makowals has quit IRC | 16:39 | |
*** makowals has joined #openstack-ironic | 16:40 | |
wznoinsk | nodepool is only to keep a certain amount of nodes up available for use, in your case you want to be starting devstack VM with ironic enabled (nova_compute disabled) and add an image you want to deploy onto BM in IMAGE_URLS in devstack I guess | 16:41 |
JayF | dtantsur: if you have a moment, https://review.openstack.org/#/c/369245/ seems like it should be an easy review, and I'd like someone more inspector-y to review it before it lands | 16:42 |
wznoinsk | further, ironic needs to know the details of the BM it should talk to (I use IPMI), i.e.: ipmi ip/user/password to manage the BM remotely (i.e.: power it on/off) | 16:42 |
patchbot | patch 369245 - ironic-python-agent - Get numa_node info when collecting pci devices info | 16:42 |
wznoinsk | from then on it's a normal nova boot | 16:42 |
sambetts | wznoinsk: exactly so we are having to do our own pool managment for the baremetals that we are going to boot | 16:42 |
*** trown is now known as trown|lunch | 16:43 | |
sambetts | wznoinsk: e.g. I have 6 BM reserved for the tests to use, I use a DB entry with a reserved flag to manage when a test is using one of the nodes | 16:43 |
sambetts | wznoinsk: I wondered if nodepool could take its place | 16:44 |
*** derekh has quit IRC | 16:44 | |
openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Enhance introspection status https://review.openstack.org/386703 | 16:46 |
wznoinsk | you can create 2 slave nodes, you'd have to have a separate (from nova_compute/libvirt) controller (like a VM) with ironic inside | 16:46 |
*** penick has joined #openstack-ironic | 16:46 | |
*** jheroux has joined #openstack-ironic | 16:47 | |
wznoinsk | and have ironic controller as another provider in nodepool | 16:47 |
wznoinsk | never tried to mix it tho, better to confirm with infra | 16:47 |
sambetts | wznoinsk: ok, thanks for the info :D | 16:48 |
*** Goneri has quit IRC | 16:49 | |
*** gcb has quit IRC | 16:51 | |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation https://review.openstack.org/353102 | 16:52 |
*** rama_y has quit IRC | 16:53 | |
mmedvede | wznoinsk: wouldn't ironic controller plugged to nodepool would basically be a pool of booted up BM servers? | 16:54 |
*** Romanenko_K has quit IRC | 16:55 | |
*** harlowja_at_home has quit IRC | 16:55 | |
jroll | meeting in 5 minutes, whee | 16:55 |
wznoinsk | mmedvede: not booted, powered off physical servers and booted only when you nova boot , sambetts wanted to have VM + BM managed as two-node build nodes | 16:55 |
mmedvede | wznoinsk: I did consider nodepool about a year ago, and at the time, it seemed easier to write your own bare-bones pool management | 16:55 |
mmedvede | wznoinsk: ah, ok | 16:56 |
mmedvede | a little different use case from ours | 16:56 |
mmedvede | wznoinsk: if you use BM node same way as you would VM in the context of nodepool, then nodepool should be able to do it indeed | 16:57 |
*** jaosorior has quit IRC | 16:57 | |
*** penick has quit IRC | 16:57 | |
wznoinsk | yes, I use BMs as the jenkins slaves themselves and run DSVM on them | 16:57 |
mat128 | wznoinsk: can't you have nodepool talk to nova like the regular use case? | 16:58 |
mat128 | wznoinsk: instead of talking to Ironic directly | 16:58 |
mmedvede | to make it clear: our usecase is to use BM nodes as second-level VM during tempest, to test ipmi boot for example | 16:58 |
dtantsur | JayF, ack | 16:58 |
wznoinsk | mat128: nodepool talks to nova with ironic as 'driver' | 16:58 |
*** rpioso has joined #openstack-ironic | 16:59 | |
mat128 | wznoinsk: so it should just work, no? | 16:59 |
wznoinsk | mat128: yes, it had a bit of things I had to tweak, one of them was sshd returning messages nodepool couldn't handle correctly (which now it is but I don't have that new nodepool installed) | 17:00 |
mat128 | wznoinsk: interesting | 17:01 |
*** mgould is now known as mgould|afk | 17:01 | |
*** beekneemech has quit IRC | 17:02 | |
*** penick has joined #openstack-ironic | 17:03 | |
*** electrofelix has quit IRC | 17:03 | |
*** rbrud has quit IRC | 17:03 | |
*** jtomasek has joined #openstack-ironic | 17:04 | |
*** ralonsoh has quit IRC | 17:05 | |
*** jaosorior has joined #openstack-ironic | 17:06 | |
*** jtomasek has quit IRC | 17:06 | |
*** jtomasek has joined #openstack-ironic | 17:06 | |
*** pmannidi has joined #openstack-ironic | 17:07 | |
*** bnemec has joined #openstack-ironic | 17:07 | |
*** ohamada_ has quit IRC | 17:09 | |
*** pmannidi has quit IRC | 17:11 | |
*** pece has quit IRC | 17:12 | |
*** srobert has joined #openstack-ironic | 17:19 | |
*** vinsh has joined #openstack-ironic | 17:29 | |
*** trown|lunch is now known as trown | 17:31 | |
*** skazi has joined #openstack-ironic | 17:34 | |
*** wajdi has joined #openstack-ironic | 17:36 | |
*** hshiina has quit IRC | 17:40 | |
*** jaosorior has quit IRC | 17:41 | |
*** jaosorior has joined #openstack-ironic | 17:42 | |
*** dprince has quit IRC | 17:42 | |
*** dprince has joined #openstack-ironic | 17:43 | |
openstackgerrit | Merged openstack/ironic: Move build_instance_info_for_deploy to deploy_utils https://review.openstack.org/377608 | 17:43 |
*** Goneri has joined #openstack-ironic | 17:46 | |
*** hj-hpe has joined #openstack-ironic | 17:48 | |
*** e0ne has quit IRC | 17:49 | |
*** jaosorior has quit IRC | 17:51 | |
*** Sukhdev has joined #openstack-ironic | 17:58 | |
*** penick has quit IRC | 17:59 | |
* sambetts -> home | 18:00 | |
openstackgerrit | Merged openstack/ironic: Add power state change notifications https://review.openstack.org/321865 | 18:01 |
sambetts | night all! | 18:01 |
*** sambetts is now known as sambetts|afk | 18:01 | |
*** rama_y has joined #openstack-ironic | 18:01 | |
dtantsur | see you tomorrow, folks! | 18:02 |
*** dtantsur is now known as dtantsur|afk | 18:02 | |
JayF | o/ | 18:02 |
*** penick has joined #openstack-ironic | 18:03 | |
lucasagomes | dtantsur|afk, see ya | 18:03 |
*** hj-hpe has quit IRC | 18:05 | |
Sukhdev | sambetts|afk : hey Sam are you around | 18:06 |
*** vsaienko has joined #openstack-ironic | 18:10 | |
vdrok | JayF just tried with ironic-lib disk_utils.create_config_drive_partition, still getting this Stderr: u"dd: failed to open '/dev/vda2': Invalid argument\n". | 18:17 |
vdrok | JayF: do you think it's better to fix it only in ironic-lib? | 18:18 |
JayF | vdrok: well, there's two choices, right? 1) Fix it in ironic-lib and migrate IPA to use it or 2) Fix it two different places two different ways | 18:19 |
vdrok | or we can fix the bash script for now and switch to ironic lib later | 18:19 |
JayF | vdrok: I think #1 would likely be more efficient, especially given how simple the change to ironic-lib would be in code, but I'm OK with either approach. | 18:19 |
vdrok | JayF: the way to fix it will be the same :) | 18:19 |
JayF | vdrok: well, in bash vs in python, right? | 18:19 |
vdrok | yup, OK, I'll make it work properly first, and then propose fixes to both places, to let reviewers decide the better way | 18:20 |
*** rbrud has joined #openstack-ironic | 18:21 | |
*** mjturek has quit IRC | 18:22 | |
JayF | vdrok: the real problem is that different ramdisks have different versions of the utilities to do parsing in, if I read that review correctly, right? | 18:22 |
*** jcoufal has quit IRC | 18:23 | |
vdrok | JayF: the parsing problems seem to be solved with the changes proposed in the current patch, now the only one remaining is with waiting for partition to appear in /dev | 18:23 |
JayF | hm. okay | 18:24 |
JayF | vdrok: do we not test configdrive in the gate? should we? | 18:24 |
vdrok | JayF: we do, but not with gpt as far as i know | 18:24 |
vdrok | all these problems come from gpt | 18:24 |
JayF | aha, gotcha, was trying to understand what pieces were at play to cause this | 18:24 |
*** vsaienko has quit IRC | 18:25 | |
JayF | that makes sense, ty | 18:25 |
JayF | and ty for digging into fixing this | 18:25 |
vdrok | np :) | 18:25 |
*** vsaienko has joined #openstack-ironic | 18:26 | |
openstackgerrit | Annie Lezil proposed openstack/ironic: Adding Timing metrics for DRAC drivers. https://review.openstack.org/377957 | 18:28 |
*** Guest21077 has quit IRC | 18:29 | |
*** mjturek has joined #openstack-ironic | 18:31 | |
*** dprince has quit IRC | 18:32 | |
*** dprince has joined #openstack-ironic | 18:32 | |
*** Sukhdev has quit IRC | 18:32 | |
*** e0ne has joined #openstack-ironic | 18:35 | |
rloo | hi mariojv, wrt notifications (subteam report in our whilteboard). do you still want to be the contact person or should i put yuriy's name there? | 18:40 |
*** xhku has left #openstack-ironic | 18:50 | |
*** vsaienko has quit IRC | 18:51 | |
*** vsaienko has joined #openstack-ironic | 18:53 | |
*** rajinir has joined #openstack-ironic | 18:58 | |
*** xhku has joined #openstack-ironic | 18:58 | |
*** Sukhdev has joined #openstack-ironic | 18:59 | |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation https://review.openstack.org/353102 | 18:59 |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 19:02 |
*** penick has quit IRC | 19:04 | |
*** e0ne has quit IRC | 19:05 | |
*** e0ne has joined #openstack-ironic | 19:05 | |
*** anup has joined #openstack-ironic | 19:07 | |
*** anup has left #openstack-ironic | 19:07 | |
*** mjturek has quit IRC | 19:08 | |
*** anup has joined #openstack-ironic | 19:09 | |
*** anup is now known as Guest98289 | 19:09 | |
*** Guest98289 has quit IRC | 19:10 | |
*** makowals has quit IRC | 19:10 | |
*** makowals has joined #openstack-ironic | 19:11 | |
*** vsaienko has quit IRC | 19:13 | |
*** srobert has quit IRC | 19:15 | |
mariojv | rloo: i don't mind staying on there or having us both there - i think yuriy's going to be writing more of the new code though, so it might make sense to switch | 19:19 |
rloo | mariojv: yeah, i think your part is done. we just need some contact person, and i don't know if you want to keep track of the status of the rest of the work | 19:20 |
mariojv | rloo: yeah, let's switch if yuriy's OK with it, otherwise i don't mind tracking and helping when needed | 19:22 |
mariojv | definitely going to review everything though :) | 19:22 |
*** penick has joined #openstack-ironic | 19:23 | |
rloo | mariojv: i'll check with yuriy first. thx for reviewing! i think we are both going to be the main reviewers... | 19:23 |
*** vishwanathj has joined #openstack-ironic | 19:24 | |
*** Guest40693 is now known as rook | 19:25 | |
openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: WiP Dynamic iPXE with core+vendor API https://review.openstack.org/369438 | 19:30 |
*** vsaienko has joined #openstack-ironic | 19:31 | |
*** vsaienko has quit IRC | 19:34 | |
*** ccamacho has quit IRC | 19:34 | |
*** jcoufal has joined #openstack-ironic | 19:34 | |
*** vsaienko has joined #openstack-ironic | 19:35 | |
*** hoangcx has joined #openstack-ironic | 19:35 | |
*** bdeschenes has joined #openstack-ironic | 19:35 | |
lucasagomes | folks, I'm calling it a day | 19:39 |
lucasagomes | have a great evening all! | 19:39 |
*** shobhar_ has joined #openstack-ironic | 19:39 | |
*** portdirect has quit IRC | 19:42 | |
openstackgerrit | Ruby Loo proposed openstack/ironic: Remove 'agent_last_heartbeat' from node.driver_internal_info https://review.openstack.org/387652 | 19:42 |
JayF | oh, good call rloo | 19:43 |
*** lucasagomes is now known as lucas-afk | 19:43 | |
JayF | mariojv: ^ you probably want to look at that patch too | 19:43 |
rloo | JayF: i was going to just reply to mariojv's patch, and then thought i would look into it just for fun :) | 19:44 |
milan | Good night Ironic! :) | 19:48 |
*** joanna has joined #openstack-ironic | 19:48 | |
*** milan is now known as milan-afk | 19:48 | |
mariojv | rloo: looks fine to me - thanks | 19:51 |
*** ccamacho has joined #openstack-ironic | 19:51 | |
rloo | thx mariojv! | 19:51 |
*** hj-hpe has joined #openstack-ironic | 19:56 | |
*** Sukhdev has quit IRC | 19:56 | |
*** anup has joined #openstack-ironic | 19:57 | |
*** anup is now known as Guest95387 | 19:58 | |
openstackgerrit | Jay Faulkner proposed openstack/ironic: Devstack will die if default driver is not enabled https://review.openstack.org/387657 | 19:59 |
JayF | aslezil: ^ that's half of the bugfix for the error you got on Friday | 19:59 |
*** trown is now known as trown|afk | 20:03 | |
openstackgerrit | Ruby Loo proposed openstack/ironic: Remove ..agent.build_instance_info_for_deploy() in Pike https://review.openstack.org/387661 | 20:06 |
*** narasimha_SV has quit IRC | 20:08 | |
JayF | Do we have a separate bugtracker for ironic-inspector client? | 20:09 |
JayF | Or should those go in ironic-inspector? | 20:09 |
*** vsaienko has quit IRC | 20:16 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic-lib: Wait for a config drive device to be available https://review.openstack.org/387664 | 20:20 |
*** Sukhdev has joined #openstack-ironic | 20:22 | |
*** skazi has quit IRC | 20:23 | |
*** skazi has joined #openstack-ironic | 20:23 | |
*** pcaruana has quit IRC | 20:25 | |
*** trown|afk is now known as trown | 20:28 | |
*** e0ne has quit IRC | 20:29 | |
rloo | JayF: yes, there's a sep bug tracker for inspector | 20:30 |
JayF | rloo: for inspector, yes, for inspector client is what I was asking about | 20:30 |
JayF | although I now don't have a bug to file there, as it was a bad report, I'm still curious for future | 20:30 |
JayF | I couldn't find a bugracker for the inspector client | 20:30 |
rloo | JayF: oh. i missed that word. good question. the inspector client isn't listed here: https://wiki.openstack.org/wiki/Ironic | 20:31 |
rloo | JayF: although if dtantsur|afk is involved, I'd think there would be a sep bug tracker | 20:31 |
*** bfournie has quit IRC | 20:33 | |
aslezil | Thanks JayF | 20:35 |
*** baoli has quit IRC | 20:35 | |
*** skazi has quit IRC | 20:38 | |
openstackgerrit | Jay Faulkner proposed openstack/ironic: Devstack will die if default driver is not enabled https://review.openstack.org/387657 | 20:53 |
openstackgerrit | Annie Lezil proposed openstack/ironic: Adding Timing metrics for DRAC drivers. https://review.openstack.org/377957 | 20:53 |
*** trown is now known as trown|outtypewww | 20:55 | |
rloo | jroll: wrt the summit sessions, do you know who is leading them? eg, neutron events etherpad has nothing on it except the title. not sure who to ask... | 20:56 |
openstackgerrit | Vladyslav Drok proposed openstack/ironic-python-agent: Fix config drive writing script https://review.openstack.org/385957 | 20:59 |
JayF | vdrok: did you test and confirm ^ that fixed the issue? | 21:01 |
*** krtaylor has quit IRC | 21:01 | |
vdrok | JayF: yup, just booted the instance with this one, configdrive properly written | 21:01 |
JayF | nice | 21:01 |
JayF | do you have a gpt image I could test this with real quick in my devstack? | 21:01 |
vdrok | this is why msdos case worked :) | 21:01 |
vdrok | JayF: I have a gpt cirros, let me share it somewhere | 21:03 |
JayF | vdrok: well, looking at the change now ... it's a lot more simple than it was before | 21:03 |
*** vsaienko has joined #openstack-ironic | 21:03 | |
vdrok | yup | 21:03 |
JayF | why the parted/partprobe change early? | 21:03 |
JayF | is that a vestige of the old fix or is it still needed too? | 21:03 |
vdrok | JayF: partprobe does not have any print stuff | 21:06 |
vdrok | I think it could not work from the beginning on any distro | 21:06 |
vdrok | JayF: https://drive.google.com/open?id=0B2h-YW8cLPRVMWtGNnBiZ2xBMzg | 21:06 |
JayF | oh. OH. | 21:06 |
vdrok | that is the gpt one ^^ | 21:07 |
JayF | as long as CI passes, I think I'm OK +2'ing this without seeing it in my own devstack | 21:07 |
JayF | b/c I understand the full change now | 21:07 |
*** portdirect has joined #openstack-ironic | 21:07 | |
JayF | We really need to get a GPT deploy image into testing :/ | 21:07 |
vdrok | I'll also add these changes to ironic-lib. I think we did not notice this stuff as configdrive code from there is not used anywhere | 21:09 |
JayF | that's not true; I think it's used for iscsi-style deploys already | 21:09 |
vdrok | oh | 21:09 |
*** jcoufal has quit IRC | 21:09 | |
vdrok | JayF: but I think in iscsi case, there is a different mechanism to check if the partition is ready to be written to | 21:10 |
JayF | I'm looking for it now | 21:10 |
vdrok | at least for agent case, it has the same problem as in that shell script | 21:10 |
jroll | rloo: I'm going to assume I'll run them all with help from the proposers (or defer to them if they are comfortable) | 21:15 |
JayF | vdrok: https://github.com/openstack/ironic/blob/914f19ecfc8699817ec7aa7156dbf83f7b3b4de2/ironic/drivers/modules/deploy_utils.py#L378 (whole disk deploy with iscsi) uses https://github.com/openstack/ironic-lib/blob/master/ironic_lib/disk_utils.py#L679 (create config drive in ironic-lib) | 21:15 |
JayF | vdrok: so it appears whole disk image + iscsi uses the existing ironic-lib configdrive path | 21:15 |
jroll | JayF: https://bugs.launchpad.net/python-ironic-inspector-client | 21:16 |
JayF | jroll: ty | 21:16 |
jroll | np | 21:16 |
JayF | the - is ... inconsitent | 21:16 |
jroll | yeah | 21:16 |
jroll | but oh so sensible :D | 21:16 |
jroll | it does match the project name though | 21:17 |
JayF | yeah; it took me longer than I'd like to admit to find the project name too | 21:17 |
JayF | lol | 21:17 |
*** dprince has quit IRC | 21:17 | |
vdrok | JayF: I'm not sure if this is the reason why it works in iscsi - https://github.com/openstack/ironic-lib/blob/master/ironic_lib/disk_partitioner.py#L114 | 21:19 |
vdrok | as these options refer only to iscsi - https://github.com/openstack/ironic-lib/blob/master/ironic_lib/disk_partitioner.py#L29-L42 | 21:19 |
vdrok | but anyway, I'll look into this tomorrow :) | 21:20 |
vdrok | good night everyone! | 21:20 |
JayF | that's just checking to make sure it's not open by another process | 21:20 |
JayF | that's all fuser does for you | 21:20 |
JayF | weird. Seems like there'll be more to unravel than I thought in making IPA use ironic-lib disk-utils | 21:20 |
vdrok | JayF: then I'm out of options why it works in one case and does not in another | 21:20 |
JayF | it may just come directly down to timing | 21:21 |
JayF | or something the iscsi daemons itself provide | 21:21 |
JayF | i.e. maybe just exporting a block device as iscsi forces the partition table to update more quickly | 21:22 |
*** athomas has quit IRC | 21:23 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic-python-agent: Fix config drive writing script https://review.openstack.org/385957 | 21:28 |
*** vishwanathj has quit IRC | 21:38 | |
*** jheroux has quit IRC | 21:42 | |
*** bdeschenes has quit IRC | 22:04 | |
*** rcernin has quit IRC | 22:06 | |
*** vsaienko has quit IRC | 22:20 | |
openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic-lib: Changed message attribute to _msg_fmt https://review.openstack.org/386251 | 22:26 |
*** pmannidi has joined #openstack-ironic | 22:27 | |
*** shobhar_ has quit IRC | 22:28 | |
*** shobhar_ has joined #openstack-ironic | 22:29 | |
*** rbrud has quit IRC | 22:34 | |
hogepodge | TheJulia: If I use the system installed pip, version 8.1.1 on xenial, I get an error on almost everything I try to install. Manually upgrading with pip install --upgrade pip gets me a working version | 22:36 |
TheJulia | Well... that is not good | 22:39 |
TheJulia | hogepodge: come to think about it, I did hit an issue with installing diskimage-builder prior to running the bifrost test script | 22:41 |
TheJulia | I think the answer was pretty much "upgrade to the latest pip" | 22:42 |
hogepodge | Something we would want to document? | 22:42 |
JayF | I wonder if there needs to be some fix backported into xenial, if pip is broken so badly nothing can install | 22:42 |
JayF | like that seems like a bug ubuntu would care about, no? | 22:42 |
hogepodge | File a bug on? Just wondering. It's kind of widespread nonsense from a bad canonical build of pip | 22:42 |
TheJulia | I think it was something like something got pulled from the openstack project repos and the older pip couldn't install it anymore | 22:42 |
JayF | Do we have any idea why 8.1.1 doesn't work, but pypi version does? | 22:42 |
hogepodge | JayF: From what I've read some changes from 8.1.2 were cherry picked back that broke internal apis | 22:43 |
TheJulia | hogepodge: cherrypicked by who? | 22:43 |
hogepodge | I'm really regretting haven chosen xenial for this work, but it's too late to switch to centos | 22:43 |
hogepodge | TheJulia: whoever packaged pip | 22:44 |
TheJulia | hogepodge: So, yeah, in that case an ubuntu bug seems to be the most appropriate thing | 22:44 |
hogepodge | They also released new ceph packages over the weekend that broke kolla builds. | 22:44 |
TheJulia | The Joy of distributions. :( | 22:45 |
hogepodge | (a fix for it was just merged upsteam in kolla) | 22:45 |
hogepodge | I can't say that centos packaging hasn't broken for me in clutch situations. "Why did all my dependencies break in the last hour...?" | 22:45 |
*** bdeschenes has joined #openstack-ironic | 22:46 | |
JayF | Honestly, this is why I'd never run an openstack service not installed from pip | 22:46 |
JayF | and now maybe add the caveat installed /from upstream/ pip | 22:46 |
JayF | lol | 22:46 |
openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic: Making IronicException to subclass from ironic_lib https://review.openstack.org/387703 | 22:46 |
TheJulia | JayF: idea w/r/t partition table, direct IO | 22:46 |
JayF | TheJulia: what are you talking about? | 22:46 |
hogepodge | xenial also has a bad kernel that breaks with the display driver in the nuc I'm using for these installs. So bring extra network cables to ssh into the thing through a dumb switch... | 22:48 |
TheJulia | Basically, non-direct IO goes through the caching subsystem of the kernel. Direct IO does not. I'm wondering if stuff is seeing cached data when operations were direct... | 22:48 |
JayF | TheJulia: Oh, for vdrok's patch, if you look at his latest, the real issue was something pretty minor | 22:48 |
TheJulia | oh, nevermind then | 22:48 |
JayF | TheJulia: basically it only waited for the configdrive partition to come up on the MBR codepath, not on GPT codepath | 22:48 |
JayF | TheJulia: you should totally review it though :D | 22:48 |
TheJulia | I saw the discussion and I was curious | 22:48 |
TheJulia | so I started reading back :) | 22:49 |
TheJulia | Ahh, there we go, found it | 22:51 |
openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic: Making IronicException to subclass from ironic_lib https://review.openstack.org/387703 | 22:56 |
TheJulia | JayF: Ahh, yeah, that makes too much sense | 22:58 |
JayF | TheJulia: the real wtf is how we ever thought that'd work | 22:59 |
JayF | TheJulia: partprobe doesn't give output at all aiui | 22:59 |
TheJulia | JayF: likely distributions carrying extra features | 22:59 |
openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic: Making IronicException to subclass from ironic_lib https://review.openstack.org/387703 | 22:59 |
TheJulia | and limited verification testing | 22:59 |
JayF | TheJulia: yeah, I'd buy that | 22:59 |
TheJulia | like "Oh, it worked on this distro!" \o/ | 22:59 |
JayF | TheJulia: I was talking to jroll today about how we gotta figure out how to increase the test matrix for stuff like this | 22:59 |
JayF | so much complexity and our tests are only covering a small % of it | 22:59 |
*** michauds has quit IRC | 23:00 | |
TheJulia | JayF: does anyone publish a gpt based cloud image? | 23:01 |
JayF | TheJulia: idk; vdrok linked a GPT cirros in here earlier | 23:02 |
TheJulia | Ahh... time opportunistic tempest job time? | 23:02 |
openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic: Making IronicException to subclass from ironic_lib https://review.openstack.org/387703 | 23:02 |
TheJulia | That doesn't use nova perhaps | 23:02 |
JayF | I'm under the impression he just locally built it | 23:03 |
JayF | i.e. the link he gave was google drive, not like an official cirros image or anything | 23:03 |
TheJulia | Well, shouldn't be THAT hard to rebuild cirros as a partition in a GPT disk | 23:03 |
JayF | yeah. I don't think any of the pieces are hard. | 23:04 |
JayF | Finding time to do it in a reproducable, openstack-acceptable way is :) | 23:04 |
*** baoli has joined #openstack-ironic | 23:10 | |
TheJulia | That... could be a neverending task | 23:15 |
*** baoli has quit IRC | 23:15 | |
*** penick has quit IRC | 23:21 | |
openstackgerrit | Stephane Robert proposed openstack/ironic: Baytech MRP27 PDU support https://review.openstack.org/387718 | 23:30 |
*** amotoki has quit IRC | 23:31 | |
*** rama_y has quit IRC | 23:33 | |
*** srobert has joined #openstack-ironic | 23:34 | |
*** jralph has quit IRC | 23:35 | |
JayF | jroll: or others: What's our policy for taking changes on the deprecated drivers; i.e. ^ that patch adding support for another pdu to the snmp driver | 23:35 |
*** rajinir has quit IRC | 23:36 | |
*** Sukhdev has quit IRC | 23:36 | |
*** awiddersheim has quit IRC | 23:37 | |
*** jralph has joined #openstack-ironic | 23:38 | |
*** awiddersheim has joined #openstack-ironic | 23:39 | |
*** chlong has joined #openstack-ironic | 23:40 | |
mtanino | Hi, | 23:43 |
mtanino | I'm trying to create Ironic env using latest devstack with neutron(not flat network). After building the devstack envirnment, ironic node has http://10.0.0.x:9999 in agent_url of driver_internal_info. The http address can't access directly because of neutron L3. When I deploy nova baremetal instance, I saw ironic-conductor tries to access that agent_url to know whether the deploy is succeeded or not, but this access always fails be | 23:44 |
mtanino | cause of neutron L3. | 23:44 |
*** amotoki has joined #openstack-ironic | 23:44 | |
mtanino | Is this my configuration issue? | 23:44 |
JayF | mtanino: I deployed a fresh multitenant ironic devstack, using with little modification the config here: http://docs.openstack.org/developer/ironic/dev/ironic-multitenant-networking.html and my conductor can talk to agents just fine | 23:45 |
JayF | mtanino: I literally built it on Friday, so I know it at least works in some cases. Additionally we test this in the gate. So more information is needed about your configuration to determine what, if anything, is up | 23:45 |
mtanino | JayF: Thank you. Let me try using that local,conf. | 23:46 |
openstackgerrit | Stephane Robert proposed openstack/ironic: Baytech MRP27 PDU support https://review.openstack.org/387718 | 23:46 |
mtanino | JayF: I'm following this document. http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html | 23:47 |
JayF | mtanino: yep, I basically did that except using teh config for multitenant to get my multitenant environment working. | 23:47 |
*** wajdi has quit IRC | 23:51 | |
*** wajdi has joined #openstack-ironic | 23:51 | |
*** wajdi has quit IRC | 23:52 | |
*** bfournie has joined #openstack-ironic | 23:52 | |
mtanino | JayF: In your environment, can you ping to 10.1.0.x address from the machine which conductor is working on? | 23:53 |
*** Goneri has quit IRC | 23:53 | |
mtanino | ping or curl. | 23:53 |
JayF | mtanino: I haven't tried to ping or curl an agent while deploying; it's not something that I've had to do. | 23:53 |
JayF | mtanino: you do realize that URL will only be accessible while the agent is actively booted and doing work; not at all times the machine is powered on? | 23:54 |
mtanino | JayF: Ah, I see. | 23:54 |
JayF | Generally speaking, when using ironic, stuff in driver_internal_info, including the agent_url, is sorta... non-user-operable parts :D | 23:55 |
JayF | Now, if your deploys are failing with connection errors to the agent, obviosuly troubleshooting there makes sense | 23:55 |
JayF | but don't go looking for trouble :P | 23:55 |
JayF | Honestly, unless your local.com is almost identical to the one documented on our multitenant page, I'd strongly suggest doing an unstack, replacing your config with our reccomended for multitenant testing, and re-stacking | 23:56 |
JayF | *local.conf | 23:56 |
mtanino | JayF: so conductor must have an access to the agent URL using http during instance deploy, right? | 23:57 |
JayF | Yes. | 23:57 |
JayF | in the case of devstack though ... that's not something you should have to explicitly configure | 23:57 |
JayF | if that makes sense? | 23:57 |
*** rloo has quit IRC | 23:58 | |
mtanino | JayF: Yes, may be I'm missing somthing of basic network configuration. | 23:59 |
JayF | Yeah; I'd do what I'd suggest and unstack/restack | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!