| *** ChrisAusten has joined #openstack-ironic | 00:05 | |
| *** chlong has quit IRC | 00:05 | |
| *** aginwala has quit IRC | 00:09 | |
| *** aginwala has joined #openstack-ironic | 00:12 | |
| *** garthb_ has quit IRC | 00:13 | |
| *** rloo has quit IRC | 00:16 | |
| *** fragatin_ has quit IRC | 00:17 | |
| *** dtantsur has joined #openstack-ironic | 00:22 | |
| *** fragatina has joined #openstack-ironic | 00:29 | |
| *** fragatina has quit IRC | 00:33 | |
| *** aginwala has quit IRC | 00:36 | |
| *** rloo has joined #openstack-ironic | 00:40 | |
| *** aginwala has joined #openstack-ironic | 00:40 | |
| *** fragatina has joined #openstack-ironic | 00:41 | |
| *** rloo has quit IRC | 00:41 | |
| *** piet has joined #openstack-ironic | 00:42 | |
| *** fragatina has quit IRC | 00:42 | |
| *** afaranha has quit IRC | 00:47 | |
| *** fragatina has joined #openstack-ironic | 00:51 | |
| *** mtanino has quit IRC | 00:52 | |
| *** fragatin_ has joined #openstack-ironic | 00:52 | |
| *** fragatin_ has quit IRC | 00:53 | |
| *** fragatin_ has joined #openstack-ironic | 00:53 | |
| *** hoangcx has joined #openstack-ironic | 00:54 | |
| *** fragatina has quit IRC | 00:55 | |
| *** fragatina has joined #openstack-ironic | 00:56 | |
| *** fragatin_ has quit IRC | 00:57 | |
| *** fragatina has quit IRC | 01:01 | |
| *** piet has quit IRC | 01:01 | |
| *** baoli_ has quit IRC | 01:09 | |
| *** baoli has joined #openstack-ironic | 01:10 | |
| *** chlong has joined #openstack-ironic | 01:31 | |
| *** fragatina has joined #openstack-ironic | 01:34 | |
| *** sacharya has joined #openstack-ironic | 01:47 | |
| *** Goneri has joined #openstack-ironic | 01:48 | |
| *** sacharya has quit IRC | 01:51 | |
| *** jcoufal has joined #openstack-ironic | 01:57 | |
| *** baoli has quit IRC | 01:57 | |
| *** keedya has joined #openstack-ironic | 01:59 | |
| *** jcoufal has quit IRC | 02:03 | |
| openstackgerrit | Merged openstack/ironic-python-agent: Stop using git:// and be nice to people behind proxy servers https://review.openstack.org/313115 | 02:05 |
|---|---|---|
| *** aginwala has quit IRC | 02:13 | |
| openstackgerrit | Merged openstack/ironic: Pass environment through to create-node.sh https://review.openstack.org/313106 | 02:14 |
| openstackgerrit | Merged openstack/ironic: Update resources only for specific node during deletion. https://review.openstack.org/287498 | 02:15 |
| *** piet has joined #openstack-ironic | 02:20 | |
| *** joprovost has quit IRC | 02:22 | |
| *** fragatina has quit IRC | 02:24 | |
| *** keedya has quit IRC | 02:24 | |
| *** sacharya has joined #openstack-ironic | 02:35 | |
| *** piet has quit IRC | 02:37 | |
| *** absubram has joined #openstack-ironic | 02:53 | |
| *** absubram_ has joined #openstack-ironic | 02:54 | |
| *** piet has joined #openstack-ironic | 02:56 | |
| *** absubram has quit IRC | 02:58 | |
| *** absubram_ is now known as absubram | 02:58 | |
| *** daemontool_ has quit IRC | 03:04 | |
| *** chenke has joined #openstack-ironic | 03:05 | |
| chenke | have anybody use the inspur server as the baremetal server, i saw the server have loaded the deploy ramdisk and kernel but not execute it. Dose anybody solve this kind of issue? | 03:08 |
| *** piet has quit IRC | 03:13 | |
| *** fragatina has joined #openstack-ironic | 03:16 | |
| *** fragatina has quit IRC | 03:16 | |
| *** fragatina has joined #openstack-ironic | 03:16 | |
| *** absubram has quit IRC | 03:26 | |
| *** amit213 has quit IRC | 03:33 | |
| *** amit213 has joined #openstack-ironic | 03:33 | |
| *** sivaramakrishna has joined #openstack-ironic | 03:46 | |
| *** vmud213 has joined #openstack-ironic | 03:53 | |
| *** mmitchell has joined #openstack-ironic | 03:58 | |
| *** mmitchell has left #openstack-ironic | 03:58 | |
| *** links has joined #openstack-ironic | 04:02 | |
| *** hamzy has joined #openstack-ironic | 04:06 | |
| *** ankkumar5 has joined #openstack-ironic | 04:08 | |
| *** praneshp has quit IRC | 04:09 | |
| *** sacharya has quit IRC | 04:19 | |
| *** piet has joined #openstack-ironic | 04:21 | |
| *** aginwala has joined #openstack-ironic | 04:28 | |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 04:51 |
| *** wajdi has joined #openstack-ironic | 04:53 | |
| openstackgerrit | vinay kumar muddu proposed openstack/ironic-lib: Add support for BIOS local boot for GPT label https://review.openstack.org/311613 | 05:14 |
| *** mbound has joined #openstack-ironic | 05:18 | |
| *** sacharya has joined #openstack-ironic | 05:20 | |
| *** sacharya has quit IRC | 05:25 | |
| *** piet has quit IRC | 05:27 | |
| openstackgerrit | vinay kumar muddu proposed openstack/ironic-lib: Add support for BIOS local boot for GPT label https://review.openstack.org/311613 | 05:28 |
| *** ChrisAusten has quit IRC | 05:36 | |
| *** mbound has quit IRC | 05:43 | |
| *** yolanda has joined #openstack-ironic | 05:55 | |
| *** wajdi has quit IRC | 06:01 | |
| *** vishwanathj has quit IRC | 06:13 | |
| *** chenke has quit IRC | 06:19 | |
| *** e0ne has joined #openstack-ironic | 06:39 | |
| *** yolanda has quit IRC | 06:46 | |
| *** yolanda has joined #openstack-ironic | 06:47 | |
| *** sacharya has joined #openstack-ironic | 06:53 | |
| *** e0ne has quit IRC | 06:53 | |
| *** athomas has joined #openstack-ironic | 06:55 | |
| *** sacharya has quit IRC | 06:58 | |
| *** e0ne has joined #openstack-ironic | 06:59 | |
| *** e0ne has quit IRC | 07:00 | |
| *** yolanda has quit IRC | 07:02 | |
| openstackgerrit | Aparna proposed openstack/proliantutils: Using python 'subprocess' to execute hpssacli commands https://review.openstack.org/313254 | 07:05 |
| *** yolanda has joined #openstack-ironic | 07:06 | |
| *** hoangcx has quit IRC | 07:08 | |
| *** aginwala has quit IRC | 07:13 | |
| *** aginwala has joined #openstack-ironic | 07:14 | |
| *** e0ne has joined #openstack-ironic | 07:16 | |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add network drivers https://review.openstack.org/285852 | 07:20 |
| *** ifarkas has joined #openstack-ironic | 07:21 | |
| *** yolanda has quit IRC | 07:21 | |
| *** chopmann has joined #openstack-ironic | 07:27 | |
| *** yolanda has joined #openstack-ironic | 07:32 | |
| betherly | o/ morning all! | 07:35 |
| *** openstackstatus has quit IRC | 07:38 | |
| *** openstackstatus has joined #openstack-ironic | 07:39 | |
| *** ChanServ sets mode: +v openstackstatus | 07:39 | |
| *** ohamada has joined #openstack-ironic | 07:51 | |
| *** daemontool has joined #openstack-ironic | 07:53 | |
| *** chlong has quit IRC | 07:54 | |
| *** hoangcx has joined #openstack-ironic | 07:54 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:02 | |
| *** e0ne has quit IRC | 08:07 | |
| *** dmk0202 has joined #openstack-ironic | 08:08 | |
| *** lucas-afk is now known as lucasagomes | 08:15 | |
| lucasagomes | jlvillal, yeah I hit the resource creation error yesterday in grenade | 08:15 |
| lucasagomes | I will tkae a look at it | 08:15 |
| lucasagomes | betherly, morning | 08:15 |
| lucasagomes | morning all | 08:15 |
| betherly | o/ lucasagomes :) | 08:18 |
| *** e0ne has joined #openstack-ironic | 08:20 | |
| vmud213 | morning lucasagomes, betherly | 08:22 |
| *** aginwala has quit IRC | 08:28 | |
| *** chopmann has quit IRC | 08:29 | |
| betherly | o/ vmud213 | 08:30 |
| alineb | morning all | 08:32 |
| *** jistr has joined #openstack-ironic | 08:32 | |
| *** pcaruana has joined #openstack-ironic | 08:36 | |
| *** derekh has joined #openstack-ironic | 08:38 | |
| *** e0ne has quit IRC | 08:40 | |
| *** chenke has joined #openstack-ironic | 08:41 | |
| *** fragatina has quit IRC | 08:41 | |
| openstackgerrit | Aline Bousquet proposed openstack/ironic-lib: Add support for metrics https://review.openstack.org/301526 | 08:42 |
| *** e0ne has joined #openstack-ironic | 08:42 | |
| *** chopmann has joined #openstack-ironic | 08:45 | |
| lucasagomes | vmud213, alineb morning | 08:46 |
| lucasagomes | hi all, https://review.openstack.org/#/c/308965/ has 2+2 already someone wants to give it a final review? | 08:47 |
| *** fragatina has joined #openstack-ironic | 08:50 | |
| *** sacharya has joined #openstack-ironic | 08:54 | |
| *** tesseract has joined #openstack-ironic | 08:56 | |
| *** tesseract is now known as Guest42930 | 08:56 | |
| *** sacharya has quit IRC | 08:59 | |
| *** alex_xu has quit IRC | 09:03 | |
| *** alex_xu has joined #openstack-ironic | 09:05 | |
| *** electrofelix has joined #openstack-ironic | 09:07 | |
| *** chenke_ has joined #openstack-ironic | 09:08 | |
| *** mkovacik has joined #openstack-ironic | 09:09 | |
| mkovacik | morning Ironic! | 09:10 |
| vmud213 | lucasagomes: When you get a chance can you look into https://review.openstack.org/#/c/311613 | 09:11 |
| *** chenke has quit IRC | 09:12 | |
| *** irf has joined #openstack-ironic | 09:14 | |
| irf | morning Ironic !!! | 09:14 |
| sambetts|afk | Morning all | 09:14 |
| *** sambetts|afk is now known as sambetts | 09:14 | |
| irf | as per doc : http://docs.openstack.org/developer/ironic/deploy/troubleshooting.html#nova-returns-no-valid-host-was-found-error | 09:15 |
| irf | i have completed the first four points ... | 09:15 |
| irf | regarding fifth point ... i am not sure ... | 09:16 |
| irf | the output of command : "nova hypervisor-stats" should show the status of hypervisor for Ironic node ? | 09:18 |
| mkovacik | sambetts, hi, how about the HA spec ;) | 09:18 |
| mkovacik | aarefiev, morning! ^ :) | 09:18 |
| sambetts | irf the output of the command nova hypervisor-stats should show you the count of all the resources avaialble to it from Ironic | 09:19 |
| sambetts | mkovacik: will check it out this morning, I've been distracted by tinyIPA the last couple of days hehe | 09:19 |
| irf | ok .. so it means all the bare metal resources ? | 09:20 |
| mkovacik | sambetts, if not I'm reminding you on Monday for sure :P | 09:20 |
| sambetts | irf: so if the ironic node is availiable and usable by nova, the count of memory_mb will increase by the amount that ironic node has | 09:20 |
| sambetts | the same for the other stats, free disk and cpus | 09:21 |
| sambetts | mkovacik: hehe ;) | 09:22 |
| *** mgould has joined #openstack-ironic | 09:26 | |
| irf | ohh ok ..so is there any way where i can say that nova has undrstood my BM node and it has now taken it under considertation for compute .. actually i dont have any stats befor adding BM node so dont know wheather it has added in nova compute or not | 09:32 |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add portgroups to support LAG interfaces - API https://review.openstack.org/206244 | 09:33 |
| mgould | morning all | 09:33 |
| mkovacik | mgould, morning! | 09:33 |
| irf | morning mgould | 09:34 |
| mgould | morning mkovacik irf | 09:36 |
| irf | ohh ok ..so is there any way where i can say that nova has undrstood my BM node and it has now taken it under considertation for compute .. actually i dont have any stats befor adding BM node so dont know wheather it has added in nova compute or not... | 09:37 |
| dtantsur | Morning Ironic | 09:40 |
| irf | here is the output of command : http://pastebin.com/XwruKUQa | 09:41 |
| mgould | morning dtantsur | 09:49 |
| mkovacik | dtantsur, morning | 09:49 |
| *** fragatina has quit IRC | 09:52 | |
| irf | please suggest ... | 10:02 |
| sivaramakrishna | Hi irf, | 10:04 |
| sivaramakrishna | how many nodes do you have? | 10:04 |
| sivaramakrishna | if you do a nova hypervisor-list | 10:04 |
| sivaramakrishna | the ironic node uuid show as 'up' | 10:04 |
| irf | i ahve just single machine | 10:06 |
| sivaramakrishna | can you run nova hypervisor-list | 10:07 |
| irf | nova hypervisor-list command shoes just one node | 10:07 |
| irf | whihc is my compute node | 10:07 |
| irf | let me show the output in pastebin | 10:07 |
| *** ekarlso has quit IRC | 10:08 | |
| irf | http://pastebin.com/dfs6x0Uh | 10:08 |
| *** hoangcx has quit IRC | 10:08 | |
| sivaramakrishna | hm..your ironic node uuid isn't appearing there | 10:09 |
| sivaramakrishna | check if the power sync up happened successfully | 10:09 |
| mkovacik | dtantsur, since I've been annoying with it aarefiev and sambetts already I have to do the same to you :P How about the HA spec? ;) | 10:10 |
| dtantsur | mkovacik, slowly getting to it :) | 10:10 |
| dtantsur | ETOOMANYEMAILS | 10:10 |
| irf | i guess not | 10:11 |
| mkovacik | yes | mail > /dev/null & 2&1 | 10:11 |
| irf | the reason being .. i am seeing some errors in the nova-schedulr logs | 10:12 |
| sivaramakrishna | check ironic-conductor logs as well | 10:12 |
| irf | ok | 10:13 |
| irf | irconic conductor logs does not show any logs related to the issue i m facing | 10:13 |
| irf | let me paste the output | 10:14 |
| irf | this is nova scheduler logs : http://pastebin.com/5zXaTgZU | 10:14 |
| mgould | mkovacik, is the HA inspector spec now basically stable? | 10:15 |
| sivaramakrishna | is your node in available state.. and what's the power state? | 10:15 |
| sivaramakrishna | output ot ironic node-list | 10:15 |
| irf | node is in available state and is power on | 10:18 |
| irf | ironic node-list shows my bare metal node and it is in power on mode | 10:18 |
| sambetts | if you run "ironic node validate", whats the output? | 10:19 |
| sambetts | ironic node validate <uuid> | 10:20 |
| *** electrofelix has quit IRC | 10:20 | |
| *** dims has quit IRC | 10:20 | |
| mkovacik | mgould, yup, the spec should be safe to read now ;) | 10:21 |
| mgould | mkovacik, :-) | 10:23 |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add network drivers https://review.openstack.org/285852 | 10:24 |
| irf | here is the output of the comamnd : http://pastebin.com/LECAKayX | 10:24 |
| irf | ironic node-validate | 10:24 |
| sambetts | irf: can you paste an ironic node-list and ironic node-show <uuid> for that node | 10:25 |
| irf | ok | 10:25 |
| sambetts | btw we have our own paste service, paste.openstack.org if you want to use it | 10:26 |
| irf | here it is : http://pastebin.com/Yvx2PJyW | 10:28 |
| irf | next time sure | 10:28 |
| *** electrofelix has joined #openstack-ironic | 10:32 | |
| sambetts | irf: are you trying to hook Ironic into a hybrid BM and VM env? | 10:32 |
| * dtantsur just reviewed a huge portgroups patch | 10:34 | |
| sambetts | dtantsur: the API patch? | 10:35 |
| dtantsur | yep | 10:35 |
| * dtantsur switches to the network drivers one | 10:35 | |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add network drivers https://review.openstack.org/285852 | 10:35 |
| *** alaski is now known as lascii | 10:36 | |
| irf | sorry dod not get | 10:36 |
| irf | what is that ? | 10:36 |
| sambetts | are you trying to make nova control BM via Ironic and VMs? | 10:37 |
| irf | my opesntack rest of the services are deployed on VMs along with Ironic service and BM node is differemt ... | 10:38 |
| sivaramakrishna | dtansur: thanks for the review; had rebased it to include get_patch_values fix.. | 10:38 |
| sivaramakrishna | made required changes in the port/portgroups | 10:39 |
| mgould | irf: if you succeed, will users be able to start VMs using "nova boot"? | 10:39 |
| *** MattMan has quit IRC | 10:40 | |
| *** MattMan has joined #openstack-ironic | 10:40 | |
| dtantsur | jroll, left a bulk of questions on the network drivers patch re interaction with the driver composition | 10:41 |
| sambetts | dtantsur: you make some good points there ;) | 10:43 |
| irf | i never tried that | 10:43 |
| dtantsur | ok, now to the HA spec | 10:43 |
| sambetts | irf: I can see in your nova hypervisor stats output that you have a nova compute serving VMs | 10:44 |
| sambetts | irf: with 5 of them running | 10:44 |
| mgould | irf, sure, but is that part of what you're trying to achieve? I think not, but it seems you've accidentally made that happen... | 10:44 |
| irf | yes | 10:44 |
| sambetts | irf: do you have 2 nova computes running, one for vms and one for Ironic? | 10:45 |
| irf | nop | 10:45 |
| irf | i have just sigle compute node | 10:45 |
| irf | i have just single compute node and one controller node ... | 10:46 |
| irf | on controller node all services are deployed | 10:46 |
| irf | along with Ironic | 10:46 |
| irf | nova scheduler is running on controller node and nova compute service is running on compute node | 10:47 |
| sambetts | from novas perspective, Ironic is a hypervisor, so you need to have a nova compute running with the ironic virt driver running to interact with Ironic | 10:47 |
| sambetts | http://docs.openstack.org/developer/ironic/deploy/install-guide.html#configure-compute-to-use-the-bare-metal-service | 10:47 |
| irf | ok .. | 10:48 |
| irf | let me chk this doc | 10:48 |
| irf | but i have already configured the nova conf | 10:48 |
| irf | fle | 10:48 |
| irf | file | 10:48 |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Update the deploy drivers with network flipping logic https://review.openstack.org/213262 | 10:48 |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Added operator documentation for ironic portgroups https://review.openstack.org/228496 | 10:49 |
| sivaramakrishna | vsaienko:can you look at https://review.openstack.org/#/c/256364/61 .. is it half done or something? | 10:51 |
| *** Goneri has quit IRC | 10:52 | |
| vdrok | sivaramakrishna: he'll be back on tuesday, it's holidays in Ukraine. That patch is done, tho some people requested to move it out of chain. I'd leave it as is for now | 10:53 |
| sivaramakrishna | thanks vdrok.. | 10:56 |
| sivaramakrishna | leaving it as is for now, will try to resolve merge conflict after a while. | 10:57 |
| *** dmk0202 has quit IRC | 10:57 | |
| *** dmk0202 has joined #openstack-ironic | 10:57 | |
| *** sivaramakrishna has quit IRC | 10:58 | |
| *** jtomasek has quit IRC | 10:59 | |
| sambetts | I personally think it should be broken out of the networking patch chain, and it'll need reworking anyway, because the way we handle hwinfo in devstack has changed recently too | 10:59 |
| lucasagomes | vmud213, awesome fix! | 11:02 |
| lucasagomes | vmud213, left a comment/suggestion regarding the release note, I find that a bit misleading (since it does not mention local boot, where GPT + BIOS w/o it is fine) | 11:02 |
| lucasagomes | other than that it LGTM | 11:02 |
| *** dims has joined #openstack-ironic | 11:02 | |
| vmud213 | lucasagomes: thanks for the review. Will look into the comments and get back to you soon :) | 11:03 |
| sambetts | mgould: https://review.openstack.org/#/c/313103/, tinyipa -> http://logs.openstack.org/03/313103/1/check/gate-tempest-dsvm-ironic-pxe_ipa-tinyipa-src-nv/dda65de/console.html, coreos -> http://logs.openstack.org/03/313103/1/check/gate-tempest-dsvm-ironic-pxe_ipa-src/24af878/console.html | 11:06 |
| sambetts | mgould: check out the difference in test run times | 11:06 |
| mgould | sambetts, 337s versus 675s - awesome! | 11:08 |
| sambetts | mgould: yeah :D need more test results because of potential differences in the speed of the test slave its landed on, but looks promising | 11:09 |
| mgould | hurrah for N=1 samples :-) | 11:10 |
| sambetts | :-P | 11:11 |
| mgould | sambetts, that's with precompiled Python libs on tinyipa? | 11:14 |
| sambetts | mgould: yup, so the gate is slower than my local machine which run in 220-230 | 11:15 |
| sambetts | total job time was 35 minutes, which is 21 minutes of devstack + 5 minutes of test run + something | 11:16 |
| * mgould wonders what's taking up the other 9 minutes | 11:17 | |
| * sambetts guesses its the time to setup the slave, and time to clean up / post results | 11:18 | |
| mgould | can we reduce devstack time by eg installing fewer services? I guess that means we won't discover harmful interactions with other services | 11:18 |
| mgould | but OTOH, we'd have our gate broken by other projects less often... | 11:18 |
| sambetts | mgould: we already reduce the number services quite a bit | 11:19 |
| sambetts | I'm not sure what the slow part of devstack is, because we're already caching the repos etc so we don't have to git clone everytime | 11:21 |
| sambetts | devananda was saying that we shouldn't need to be running swift for either agent or iscsi deployment :/ | 11:22 |
| sambetts | so we may be able to looking killing that service | 11:22 |
| jroll | 20 minutes sounds about right for devstack | 11:23 |
| jroll | also, morning | 11:23 |
| vdrok | morning jroll | 11:24 |
| vdrok | sambetts: I think it's better to run swift in devstack, people are using it | 11:24 |
| jroll | +1, for the agent driver | 11:25 |
| jroll | for iscsi, don't think it's needed | 11:25 |
| vdrok | yep | 11:25 |
| sambetts | jroll: devananda seemed to think we didn't even need it for agent deploy | 11:26 |
| sambetts | :/ not sure though | 11:26 |
| mgould | morning jroll | 11:26 |
| vdrok | sambetts: we can use direct http links, not swift temp urls, but still, lots of people will be using swift | 11:26 |
| jroll | sambetts: technically we could use apache | 11:26 |
| jroll | but yeah that, I'd rather not break tempurls | 11:26 |
| sambetts | ah ok | 11:26 |
| jroll | majority of our timeouts are test/build timeouts anyway, not job timeouts | 11:27 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: Remove check on carrier from the DHCP wait loop during inspection https://review.openstack.org/313511 | 11:29 |
| dtantsur | lucasagomes, jroll, mind checking ^^ please? gives me a lot of headache downstream :( | 11:29 |
| dtantsur | (this is a one-line patch) | 11:30 |
| lucasagomes | dtantsur, will do | 11:30 |
| jroll | Firefox can't find the server at file.rdu.redhat.com | 11:33 |
| jroll | :| | 11:33 |
| * jroll trusts, +2 | 11:33 | |
| dtantsur | ouch, this is an internal server | 11:33 |
| * dtantsur complains about people using too many internal things for no reason | 11:34 | |
| sambetts | wow... nice to see some nic drivers not playing the game with the carrier field... | 11:34 |
| dtantsur | jroll, the thing there is very simple: has_carrier=False for all NIC's in the posted back data. which is obviously false. | 11:34 |
| dtantsur | or it would not be able to post back any data | 11:34 |
| jroll | dtantsur: right, I gathered that :) | 11:35 |
| jroll | just was curious to see | 11:35 |
| *** alex_xu has quit IRC | 11:42 | |
| *** baoli has joined #openstack-ironic | 11:42 | |
| mgould | sambetts, carrier field? | 11:43 |
| *** yolanda has quit IRC | 11:43 | |
| lucasagomes | jroll, dtantsur just approved that | 11:44 |
| dtantsur | thnx | 11:44 |
| jroll | ty lucas | 11:44 |
| lucasagomes | mgould, yeah, that's a flag for ethernet devices to say if the plug is connected to it or not | 11:44 |
| sambetts | /sys/class/net/eth0/carrier | 11:44 |
| sambetts | mgould: ^ | 11:44 |
| sambetts | etc | 11:44 |
| dtantsur | yeah... and it does not seem to play well with something on our ramdisks... maybe DIB magic again, dunno (and don't want to dive into it tbh) | 11:45 |
| mgould | lucasagomes, sambetts thanks! | 11:46 |
| sambetts | dtantsur: have you got the IPA logs from the run that was in the bugzilla, I'd like to see if carrier == 0 or is carrier == The device not up, because both lead to has_carrier=False | 11:46 |
| dtantsur | sambetts, we don't log that unfortunately | 11:46 |
| dtantsur | but IIRC when device is not up, it throws an error | 11:46 |
| sambetts | it drops a log.debug | 11:46 |
| *** alex_xu has joined #openstack-ironic | 11:47 | |
| sambetts | https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/hardware.py#L418 | 11:47 |
| *** baoli_ has joined #openstack-ironic | 11:47 | |
| lucasagomes | sambetts, device is not up is operstate | 11:48 |
| lucasagomes | I think carrier should be independent of it no? | 11:48 |
| * lucasagomes googles | 11:48 | |
| sambetts | if the device isn't up you can't read the carrier | 11:48 |
| dtantsur | sambetts, no such debug logs | 11:48 |
| *** yolanda has joined #openstack-ironic | 11:48 | |
| sambetts | https://linuxconfig.org/how-to-detect-whether-a-physical-cable-is-connected-to-network-card-slot-on-linux | 11:48 |
| sambetts | hmm :/ | 11:48 |
| dtantsur | so it must be just "0" | 11:49 |
| mgould | sambetts, clearly the answer is "rear-facing webcam + OpenCV" | 11:49 |
| sambetts | dtantsur: or any other string thats not 1 :-P | 11:49 |
| lucasagomes | interesting, I always thought it would be independent of whether the iface is up or down | 11:49 |
| dtantsur | well, if it's possible, then yeah | 11:49 |
| dtantsur | according to https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-class-net it can only be 1 or 0 | 11:50 |
| sambetts | dtantsur: no idea :-P it appears that some drivers aren't playing by the rules, so it could be anything | 11:50 |
| *** baoli has quit IRC | 11:50 | |
| sambetts | mgould: heh | 11:51 |
| dtantsur | that's why I'm just dropping the check: I don't want to guess in which way a vendor may screw it >_< | 11:51 |
| dtantsur | mgould++ | 11:51 |
| dtantsur | oh wonderful, and if interface is not up yet, carrier is not available. so yeah, it's a good idea to drop this check, it seems | 11:53 |
| lucasagomes | folks if you have some time, mind taking a look at https://review.openstack.org/#/c/312575/ (to enable pyghmi to be installed from source on devstack) | 11:54 |
| lucasagomes | we have a patch in project config with 2+2 waiting for this change to enable a job in pyghmi's gate | 11:54 |
| lucasagomes | thanks | 11:54 |
| mat128 | dtantsur, sambetts, mgould, lucasagomes: (good morning) about dropping the carrier check | 11:55 |
| mat128 | does that mean an imposed wait time of X (60s I think?) when I have only one NIC active but 2 NICs in a machine? | 11:55 |
| mat128 | That's usually how our machines end up during that phase | 11:56 |
| dtantsur | mat128, morning. unfortunately, it does. we found no reliable way to avoid that. | 11:56 |
| *** thrash|g0ne is now known as thrash | 11:57 | |
| mgould | morning mat128 | 11:57 |
| mat128 | what do you do in a multi-NIC scenario, put them all on the same network that has a DHCP? I found this to be unreliable because you were assigned 2 IPs from the same subnet and routing would be broken half the time | 11:57 |
| dtantsur | mat128, you can tweak the wait time, if you're sure your ramdisk does not have the problem in question | 11:57 |
| mgould | mat128, why were you using that setup? It doesn't sound very useful... | 11:58 |
| * mgould <- networking n00b | 11:58 | |
| mat128 | we have multiple NICs and do bonding on the tenant network, but during deploying/cleaning/inspecting/any time we need IPA, we only have one NIC active | 11:58 |
| mgould | ah, OK | 11:58 |
| mat128 | mainly because you can't PXE boot with a bond | 11:58 |
| mgould | gotcha | 11:58 |
| mat128 | because you need both ends to cooperate on the modes/etc. | 11:59 |
| mat128 | hence the "bonding" kernel module :) | 11:59 |
| * mgould nods | 11:59 | |
| mat128 | I would have hoped the carrier thing was reliable :( | 11:59 |
| dtantsur | it's not :( | 11:59 |
| dtantsur | you can play with https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/cmd/agent.py#L118 if you know your hardware and ramdisk | 11:59 |
| mgould | let me think through the operstate thing out loud... | 12:01 |
| mgould | operstate = 1 if the NIC is powered up, and 0 if it's off, right? | 12:01 |
| mat128 | dtantsur: but if we merge your change as-is, and I know my cards reliably report carrier state, I am forced to wait for the callback, right? | 12:01 |
| mat128 | s/callback/timeout/ | 12:01 |
| mgould | so in mat128's case, operstate=0 for one NIC and 1 for the other, and it will stay that way throughout introspection | 12:01 |
| dtantsur | mat128, even if your card are fine with carrier, DIB does some interesting magic in its ramdisks | 12:02 |
| mat128 | I guess I can always check on a real machine | 12:02 |
| dtantsur | mgould, operstate is not 1/0, it has a lot of values | 12:02 |
| mgould | but we might have a situation where a NIC gets powered on half-way through introspection, so we can't in general ignore NICs with operstate=0 | 12:02 |
| mgould | oh, right | 12:02 |
| mgould | bah | 12:02 |
| dtantsur | mgould, check https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-class-net | 12:02 |
| mat128 | only on DIB ones or even CoreOS and TinyCore ? | 12:02 |
| dtantsur | mat128, we use DIB in our products, so I can't say for sure. I doubt anyone uses tinyipa in production right now (sambetts?) | 12:03 |
| lucasagomes | mat128, dtantsur not sure if this element was/is used in the bash ramdisk | 12:03 |
| lucasagomes | https://github.com/openstack/diskimage-builder/blob/master/elements/dhcp-all-interfaces/install.d/dhcp-all-interfaces.sh | 12:03 |
| lucasagomes | but it may be the one | 12:03 |
| mat128 | we are still on CoreOS right now, can,t move to tinyipa because of the lack of a shell on the screen :( lol | 12:03 |
| dtantsur | lucasagomes, it is and was | 12:03 |
| sambetts | mat128: lack of shell ? | 12:04 |
| dtantsur | mat128, it might happen you can set this option to e.g. 5 (seconds), and live happily. or even drop it to 0, because I think coreos forces NIC/DHCP initialization before running anything | 12:04 |
| lucasagomes | right, cause it does have a loop to wait for the link there | 12:04 |
| mat128 | with coreos you can ssh to it, not sure (havent really checked) tinyipa does it | 12:04 |
| mgould | dtantsur, OK, but the point is that operstate can change during the course of introspection? | 12:04 |
| dtantsur | mgould, looks like everything can change during introspection :( | 12:05 |
| mgould | dammit | 12:05 |
| dtantsur | mgould, mat128, here is how it looked for us: http://paste.openstack.org/show/496321/ | 12:06 |
| sambetts | mat128: it doesn't have ssh yet, but it has a shell if you serial console/KVM to the node running it | 12:06 |
| dtantsur | notice network data collection before DHCP | 12:06 |
| sambetts | ok, so I still think we're being screwed over by the dhcp-all-interfaces script here... | 12:06 |
| dtantsur | yep. so maybe for coreos it's safe to set inspection_dhcp_wait_timeout=0 | 12:07 |
| sambetts | I think IPA is starting in parrallel with that script, and breaking all the things | 12:07 |
| mat128 | sambetts: thought IPA was running in inittab or something, I'll check on my end, but if it does we might simply switch | 12:07 |
| dtantsur | sambetts, most likely you're right | 12:07 |
| * dtantsur hates DIB | 12:07 | |
| mat128 | lol | 12:07 |
| sambetts | dtantsur: maybe we can modify DIB ironic agent to make it run after that script? | 12:08 |
| sambetts | dtantsur: that script ifups all the interfaces etc | 12:08 |
| dtantsur | sambetts, I've no idea how to do that tbh. the dhcp thing is pretty weird there | 12:09 |
| *** rbudden has joined #openstack-ironic | 12:09 | |
| sambetts | dtantsur: they use a .service or .init script to activate it based on which init system your building into | 12:10 |
| sambetts | dtantsur: https://github.com/openstack/diskimage-builder/blob/master/elements/dhcp-all-interfaces/install.d/dhcp-interface%40.service | 12:10 |
| sambetts | dtantsur: maybe we can add an After= to ironic-agent to make sure it happens after that? | 12:11 |
| * lucasagomes go grab some lunch | 12:11 | |
| *** lucasagomes is now known as lucas-hungry | 12:11 | |
| *** dprince has joined #openstack-ironic | 12:11 | |
| sambetts | in here -> https://github.com/openstack/diskimage-builder/blob/master/elements/ironic-agent/install.d/ironic-agent-source-install/ironic-python-agent.service | 12:12 |
| * sambetts isn't super familiar with systemd units so :/ | 12:12 | |
| dtantsur | sambetts, after will work only if this script reports it's finishing.. which is unlikely IMO | 12:14 |
| sambetts | this script actually only brings up links where carrier = 1 -> https://github.com/openstack/diskimage-builder/blob/master/elements/dhcp-all-interfaces/install.d/dhcp-all-interfaces.sh | 12:15 |
| mgould | sambetts, me neither, but AIUI providing that kind of dependency graph info is most of the point of systemd | 12:16 |
| mgould | so it should definitely be possible | 12:16 |
| dtantsur | there is no one dhcp-all-interfaces unit btw. it's some kind of a special systemd unit | 12:16 |
| dtantsur | so I've no idea. we already depend on network-online, and it does not help much | 12:16 |
| sambetts | dhcp-all-interfaces' unit also depends on that which is why we're starting at the same time | 12:17 |
| sambetts | this is how it gets installed -> https://github.com/openstack/diskimage-builder/blob/master/elements/dhcp-all-interfaces/install.d/50-dhcp-all-interfaces | 12:17 |
| dtantsur | no, it depends on network. these are different | 12:17 |
| *** trown|outtypewww is now known as trown | 12:17 | |
| mgould | dtantsur, https://github.com/openstack/diskimage-builder/blob/master/elements/ironic-agent/install.d/ironic-agent-source-install/ironic-python-agent.service doesn't mention network-online | 12:18 |
| dtantsur | mgould, this one is not used in production fwiw | 12:18 |
| dtantsur | as we don't use source install | 12:18 |
| dtantsur | (but worth fixing, yeah) | 12:18 |
| * dtantsur -> lunch, brb | 12:18 | |
| mgould | ah, OK | 12:18 |
| * mgould -> also lunch | 12:19 | |
| *** ekarlso has joined #openstack-ironic | 12:19 | |
| *** alex_xu has quit IRC | 12:22 | |
| *** alex_xu has joined #openstack-ironic | 12:25 | |
| sambetts | dtantsur: is that a custom package you ship then? | 12:32 |
| mat128 | fwiw, carrier detection is reliable on CoreOS, and operstate is up/down as expected | 12:36 |
| *** ankkumar5 has quit IRC | 12:36 | |
| mat128 | http://imgur.com/9TU2LV6 | 12:37 |
| *** ankkumar5 has joined #openstack-ironic | 12:37 | |
| sambetts | mat128: I'm pretty sure the unreliable nature comes from the DIB dhcp-all-interfaces script being used to bring up every interface on the node, and that running in parrallel with IPA so IPA gets weird results | 12:38 |
| dtantsur | mat128, it works for me too, but not for some of our folks | 12:38 |
| dtantsur | sambetts, well, the RDO package, yeah | 12:38 |
| sambetts | its a race condition | 12:38 |
| *** vmud213 has quit IRC | 12:39 | |
| sambetts | or I think thats part of it | 12:39 |
| mat128 | maybe we can have a "trust sysfs" config option until it is resolved? | 12:39 |
| mat128 | instead of dropping the checks | 12:39 |
| sambetts | dropping the checks does seem wrong to me :/ | 12:40 |
| *** mat128 is now known as mat128|afk | 12:42 | |
| *** ankkumar5 has quit IRC | 12:42 | |
| *** irf has quit IRC | 12:42 | |
| *** caiobo has joined #openstack-ironic | 12:42 | |
| *** hshiina has joined #openstack-ironic | 12:44 | |
| *** Goneri has joined #openstack-ironic | 12:44 | |
| *** ayoung has quit IRC | 12:48 | |
| *** athomas has quit IRC | 12:49 | |
| *** alejandrito has joined #openstack-ironic | 12:52 | |
| dtantsur | sambetts, carrier check is broken, that's it. | 12:52 |
| dtantsur | we might think of better checks, but this one proved to not work | 12:52 |
| dtantsur | or rather: one carrier=1 is reliable. carrier=0 means nothing specific. | 12:53 |
| dtantsur | s/one/only/ | 12:53 |
| caiobo | good morning everybody :) how are you all doing? | 12:54 |
| dtantsur | morning caiobo! we're arguing about networking on Friday, what can be better? | 12:54 |
| jroll | dtantsur: wanna talk about api versions instead? :D | 12:54 |
| *** mat128|afk is now known as mat128 | 12:55 | |
| dtantsur | jroll, of course \o/ what do we version today? :) | 12:55 |
| sambetts | dtantsur: maybe we should make it if operstate == up and carrier == 1 wait else don't wait for it? | 12:55 |
| jroll | oh idk, just wanted to make your friday more fun :) | 12:55 |
| caiobo | morning dtantsur, nothing can be better haha | 12:55 |
| *** sacharya has joined #openstack-ironic | 12:55 | |
| TheJulia | jroll: that is evil, i like it :) | 12:55 |
| sambetts | hehehe | 12:55 |
| caiobo | jroll: +++ | 12:56 |
| dtantsur | sambetts, and we get people broken again, because at this moment carrier=0... | 12:56 |
| dtantsur | but it does not mean that there is no carrier, just.. because | 12:56 |
| dtantsur | jroll, lets version drivers, wdyt? | 12:56 |
| jroll | dtantsur: don't give me any ideas :) | 12:56 |
| mat128 | dtantsur: the drivers API or drivers themselves? | 12:57 |
| dtantsur | sambetts, I mean, I agree that's unfortunate. I just don't want to guess when this carrier check will break the next time.. | 12:57 |
| sambetts | I'd be super surprised if it really is being reliable ... :/ | 12:57 |
| TheJulia | I actually had some people ask why we were not versioning drivers | 12:57 |
| sambetts | unreliable | 12:57 |
| dtantsur | mat128, both? | 12:57 |
| jroll | versioning drivers is easy, move them all out and rely on pypi versioning | 12:57 |
| mat128 | dtantsur: interesting | 12:57 |
| *** thiagop-afk has joined #openstack-ironic | 12:58 | |
| dtantsur | sambetts, well, I have a guy with carrier=0 on all interfaces... and whatever dirty magic DIB does (and it does!), it should not be the case in my mind | 12:58 |
| dtantsur | I mean, I'm pretty sure this DHCP script does not unplug the cable :) | 12:58 |
| *** thiagop-afk is now known as thiagop | 12:58 | |
| dtantsur | jroll, awesome, lets do it!! | 12:58 |
| sambetts | dtantsur: carrier = 0 before or after dhcp-all-interfaces has ifuped the interface? | 12:58 |
| mat128 | dtantsur: do we know what magic it does so that carrier ends up 0 but still plugged in | 12:58 |
| thiagop | TGIF! Morning, Ironicers! | 12:58 |
| mat128 | o/ thiagop | 12:58 |
| sambetts | and actually carrier 0 or has_carrier=False? | 12:59 |
| dtantsur | sambetts, 1. during the process; 2. I can only say that carrier!=1 :) | 12:59 |
| dtantsur | morning thiagop, happy Friday | 12:59 |
| sambetts | dtantsur: during which process? inspection? | 13:00 |
| mat128 | dtantsur: do you know which kernel module was used for these interfaces? | 13:00 |
| mat128 | or network card model | 13:00 |
| dtantsur | sambetts, as you said, we have dhcp-all-interfaces racing with IPA, and during that carrier!=1, which is a lie to me | 13:00 |
| *** sacharya has quit IRC | 13:00 | |
| dtantsur | mat128, mmm. no, but I can ask, if it matters | 13:01 |
| *** ifarkas has quit IRC | 13:01 | |
| dtantsur | but I suspect we're just missing something in how linux works with network interfaces, tbh | 13:01 |
| mat128 | well, either DIB does something that we can just reproduce / workaround, or it's something in the modules themselves | 13:01 |
| *** jaypipes has joined #openstack-ironic | 13:02 | |
| *** yolanda has quit IRC | 13:02 | |
| mat128 | https://github.com/torvalds/linux/blob/master/net/core/net-sysfs.c#L180 | 13:02 |
| openstackgerrit | Aparna proposed openstack/proliantutils: Using python 'subprocess' to execute hpssacli commands https://review.openstack.org/313254 | 13:03 |
| *** athomas has joined #openstack-ironic | 13:04 | |
| sambetts | IMO I think we should also be checking IF_OPER_UP on the interfaces in that loop too, otherwise we'll be waiting for timeout on nics that'll never get an IP because they are down | 13:05 |
| dtantsur | sambetts, mat128, I think our problem here is that "carrier" in /sys is not a *hardware* property. it's a property of the interface, which is a software entity, that can be down or can be doing something | 13:05 |
| dtantsur | sambetts, I'm open to any property that is derived from underlying hardware, not from a high-level object | 13:06 |
| sambetts | its as much a hw property as /sys/class/net/<nic>/address is | 13:06 |
| *** piet has joined #openstack-ironic | 13:06 | |
| mat128 | https://www.kernel.org/doc/Documentation/networking/operstates.txt | 13:06 |
| mat128 | IFF_LOWER_UP maybe? | 13:07 |
| *** yolanda has joined #openstack-ironic | 13:07 | |
| dtantsur | sambetts, as we see, it's not. it depends on state of software (i.e. interface implementation) | 13:08 |
| mat128 | interesting find: If I ifconfig XX down, then carrier returns "Invalid argument", after ifup, kernel says "registered PHC device on XX", then carrier is 0, once kernel reports "link becomes ready" then carrier becomes 1 | 13:08 |
| dtantsur | so, my point with this patch is to get rid of a check that does not work, making sure we can also backport it | 13:09 |
| mat128 | so in that case I guess I could have carrier=0 on all interfaces, if it's in the middle if getting brought back up | 13:09 |
| dtantsur | then we can iterate on how to improve that. sambetts, mat128, are you fine with that? | 13:09 |
| mat128 | dtantsur: merging your patch as it means an imposed wait time for everyone, which bothers me a little bit | 13:10 |
| dtantsur | mat128, more than breaking people? | 13:10 |
| mat128 | can we identify in which case they are broken? so far we only have "DIB does magic and it breaks" | 13:10 |
| mat128 | I think there is more to it | 13:10 |
| dtantsur | mat128, I welcome any ideas :) it's a bit hard to debug a random failure on a ramdisk, y'know.. | 13:12 |
| mat128 | dtantsur: I perfectly understand that, don't worry :) | 13:12 |
| dtantsur | :) I'm totally fine with reverting it as soon as we have anything to replace. the last thing I want to do, though, is to dig into DIB | 13:13 |
| openstackgerrit | Akira YOSHIYAMA proposed openstack/ironic: WIP: SSH: add SSHIcsConsole class https://review.openstack.org/307181 | 13:14 |
| openstackgerrit | Akira YOSHIYAMA proposed openstack/ironic: WIP: IPMItool: add IPMIIcsConsole class https://review.openstack.org/301715 | 13:14 |
| openstackgerrit | Akira YOSHIYAMA proposed openstack/ironic: WIP: console utils: add methods for ironic console server https://review.openstack.org/301714 | 13:14 |
| openstackgerrit | Akira YOSHIYAMA proposed openstack/ironic: WIP: conductor manager: apply console-log related methods https://review.openstack.org/301717 | 13:14 |
| openstackgerrit | Akira YOSHIYAMA proposed openstack/ironic: WIP: REST API: add an API controller for console log https://review.openstack.org/301718 | 13:14 |
| *** jjohnson2 has joined #openstack-ironic | 13:15 | |
| *** cdearborn has joined #openstack-ironic | 13:17 | |
| *** mtanino has joined #openstack-ironic | 13:17 | |
| *** lucas-hungry is now known as lucasagomes | 13:18 | |
| *** piet has quit IRC | 13:19 | |
| *** piet has joined #openstack-ironic | 13:20 | |
| jroll | fyi, gate is going to be clogged up for a bit | 13:21 |
| jroll | there is suddenly tens of thousands of jobs in the queue :) | 13:21 |
| lucasagomes | :-( | 13:21 |
| caiobo | hey guys, does all patches need release notes? | 13:23 |
| vdrok | heh, much monasca-analytics, wow :) | 13:23 |
| jroll | caiobo: not all, but if it fixes a significant bug, adds a feature, has an upgrade impact, etc, it needs one | 13:24 |
| dtantsur | Friday gate \o/ | 13:24 |
| * dtantsur goes to the beach | 13:24 | |
| jroll | ++ | 13:24 |
| *** pcaruana has quit IRC | 13:25 | |
| * lucasagomes is thinking about removing the python-daemon depenency from VirtualBMC if the g-r ppl pushes back on it | 13:25 | |
| caiobo | jroll: oh I see, so even bugfix can have release notes. Thanks :) | 13:26 |
| jroll | np | 13:26 |
| *** jcoufal has joined #openstack-ironic | 13:26 | |
| *** mtanino has quit IRC | 13:27 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Modified IronicNode mock to use single source of data https://review.openstack.org/307927 | 13:30 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Changed action from 'deploy' to 'active' https://review.openstack.org/307926 | 13:30 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: IronicNodeProvision resource https://review.openstack.org/282457 | 13:30 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Added styling for driver list https://review.openstack.org/301735 | 13:30 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Added provision state manipulation https://review.openstack.org/307928 | 13:30 |
| *** joprovost has joined #openstack-ironic | 13:30 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Implement Python3 compatibility https://review.openstack.org/313012 | 13:32 |
| *** jjohnson2_ has joined #openstack-ironic | 13:33 | |
| *** [4]cdearborn has joined #openstack-ironic | 13:33 | |
| *** jjohnson2 has quit IRC | 13:35 | |
| mkovacik | dtantsur, mgould, sambetts thx for the reviews! :) | 13:43 |
| *** ifarkas has joined #openstack-ironic | 13:43 | |
| *** wajdi has joined #openstack-ironic | 13:44 | |
| mgould | mkovacik, np | 13:44 |
| wajdi | good morning all | 13:46 |
| *** absubram has joined #openstack-ironic | 13:46 | |
| *** cdearborn has quit IRC | 13:46 | |
| jlvillal | lucasagomes, Glad you hit the resource creation error too | 13:46 |
| wajdi | jlvillal: hi! I missed a chunk of messages yesterday near the end of the day. How did your test run go for grenade? | 13:47 |
| jlvillal | I moved to using tinyipa | 13:47 |
| *** absubram_ has joined #openstack-ironic | 13:47 | |
| jlvillal | No change in the failure. And did a little yak-shaving over 'short_source' | 13:47 |
| lucasagomes | jlvillal, :-( yeah I'm partially looking into it | 13:48 |
| jlvillal | sambetts, https://review.openstack.org/313132 | 13:48 |
| NobodyCam | Good Morning Iroinc'ers and OFC ... | 13:49 |
| NobodyCam | TGIF :) | 13:49 |
| wajdi | good morning! | 13:49 |
| mkovacik | :D | 13:49 |
| *** jaypipes has quit IRC | 13:50 | |
| NobodyCam | morning wajdi mkovacik | 13:50 |
| mkovacik | NobodyCam, o/ | 13:50 |
| mgould | morning NobodyCam | 13:50 |
| mkovacik | morning! | 13:50 |
| *** trown is now known as trown|outtypewww | 13:50 | |
| NobodyCam | morning mgould | 13:50 |
| NobodyCam | :) | 13:50 |
| thiagop | morning NobodyCam | 13:50 |
| *** absubram has quit IRC | 13:50 | |
| *** absubram_ is now known as absubram | 13:50 | |
| NobodyCam | hey hey thiagop morning | 13:51 |
| jroll | does anyone want to paste the resource creation error message anywhere? | 13:51 |
| dtantsur | morning jlvillal, wajdi, NobodyCam | 13:53 |
| wajdi | morning dtantsur | 13:53 |
| jlvillal | Good morning dtantsur wajdi lucasagomes jroll NobodyCam thiagop mgould sambetts mkovacik and everyone else | 13:54 |
| jlvillal | mgould, Any luck with the grenade stuff | 13:54 |
| mkovacik | jlvillal, morning! :) | 13:54 |
| jlvillal | mgould, We should try to keep the etherpad updated each day with the current status | 13:54 |
| jlvillal | mgould, I saw "we" since you were kind enough to volunteer to be a cat herder for Grenade work too :) | 13:55 |
| jlvillal | s/I saw/I say/ | 13:55 |
| mgould | jlvillal, yep | 13:55 |
| * mgould is feeling extra-braindead today, one moment... | 13:56 | |
| *** dansmith is now known as snotdan | 13:56 | |
| mgould | jlvillal, I updated the Etherpad with the work I did with sambetts yesterday | 13:57 |
| NobodyCam | morning dtantsur jroll jlvillal | 13:57 |
| jlvillal | mgould, Great and thanks :) | 13:57 |
| sambetts | jlvillal: https://review.openstack.org/#/c/313103/ | 13:57 |
| jlvillal | jroll, On the error. Line 18 & 19 of https://etherpad.openstack.org/p/ironic-newton-grenade-whiteboard | 13:58 |
| NobodyCam | morning sambetts | 13:58 |
| sambetts | o/ NobodyCam TGIF | 13:58 |
| jlvillal | jroll, Is where I am at | 13:58 |
| NobodyCam | Oh yes | 13:58 |
| *** snotdan is now known as superdan | 13:58 | |
| jlvillal | sambetts, Hopefully you saw the patch to use the qemu https://github.com repo instead of git://qeum-???? | 13:59 |
| jlvillal | It got merged yesterday | 13:59 |
| *** jtomasek has joined #openstack-ironic | 13:59 | |
| sambetts | jlvillal: no i didn't but thats awesome :D | 13:59 |
| jlvillal | Hmmm. snotdan doesn't have the ring of superdan | 13:59 |
| superdan | jlvillal: indeed.. I was trying to be accurate, but.. it was a little gross :) | 14:00 |
| jlvillal | superdan, I had an idea of what was happening :) | 14:00 |
| jlvillal | sambetts, https://review.openstack.org/313115 | 14:00 |
| jroll | jlvillal: ok cool | 14:00 |
| *** ametts has joined #openstack-ironic | 14:00 | |
| *** Goneri has quit IRC | 14:01 | |
| superdan | jlvillal: yeah, the "post summit sickness" | 14:01 |
| jlvillal | superdan, A lot of people at our San Antonio site seemed to have been hit as well :( | 14:01 |
| superdan | yeah, sucks | 14:01 |
| * sambetts also got hit with summit sickness | 14:02 | |
| caiobo | morning NobodyCam | 14:02 |
| jlvillal | jroll, I started a run right before I left work last night. Will be checking it when I get back in. I will update etherpad with any new info. | 14:02 |
| jlvillal | sorry sambetts :( | 14:02 |
| jroll | 2016-05-03 17:41:55.650 | + /opt/stack/new/grenade/projects/60_nova/resources.sh:create:94 : openstack ip floating add 172.24.5.6 nova_server1 | 14:02 |
| jroll | 2016-05-03 17:41:57.241 | No nw_info cache associated with instance (HTTP 400) (Request-ID: req-08647e54-1512-4285-adfe-6e1e1b1089a1) | 14:02 |
| jroll | hmm, seems odd | 14:02 |
| jroll | I feel like we just need a delay there? | 14:03 |
| jlvillal | jroll, Maybe but I'm not sure. Previously I did some work to re-run that after the test pass failed. Just the resources.sh script. And it kept failing. | 14:03 |
| NobodyCam | morning caiobo | 14:03 |
| jlvillal | I will try again today. | 14:03 |
| jroll | jlvillal: I mean between the server create and the flip create | 14:04 |
| jlvillal | jroll, As an FYI, the 'vm-setup.sh' script works well on RackSpace VMs. https://github.com/JohnVillalovos/devstack-gate-test | 14:04 |
| * jroll will play around | 14:04 | |
| jroll | ew, VMs. | 14:04 |
| jlvillal | I tried it a couple days ago. | 14:04 |
| jroll | :P | 14:04 |
| *** alejandrito has quit IRC | 14:05 | |
| *** catintheroof has joined #openstack-ironic | 14:05 | |
| jlvillal | jroll, Ah okay on where the delay should be. I will check that out. Thanks | 14:05 |
| jroll | jlvillal: blah, it has a --wait so maybe not | 14:05 |
| jroll | idk, I'll play around | 14:06 |
| *** ayoung has joined #openstack-ironic | 14:07 | |
| jjohnson2_ | lucasagomes, fyi, I did another patch. I stopped doing range = xrange, so I'm no longer overwriting a builtin | 14:07 |
| lucasagomes | jjohnson2_, cool, will take a look soon-ish | 14:08 |
| lucasagomes | thanks for that | 14:08 |
| * mgould -> relocate | 14:08 | |
| jjohnson2_ | lucasagomes, I left the other stuff as-is, since I'm also supporting some other projects where pulling in six would represent some silly work | 14:08 |
| *** mgould has quit IRC | 14:08 | |
| *** mtanino has joined #openstack-ironic | 14:09 | |
| jjohnson2_ | long story, in no way justifiable, but easier for me to roll with it over these relatively little things | 14:09 |
| *** mgould has joined #openstack-ironic | 14:11 | |
| lucasagomes | jjohnson2_, yeah for the other stuff it's fine I guess, just the thing about overwriting the builtin wasn't very nice | 14:13 |
| lucasagomes | jjohnson2_, in case we have to deal with more compat problems we can rethink about using six | 14:14 |
| *** Goneri has joined #openstack-ironic | 14:14 | |
| *** catintheroof has quit IRC | 14:15 | |
| *** vishwanathj has joined #openstack-ironic | 14:15 | |
| -openstackstatus- NOTICE: Zuul is currently recovering from a large number of changes, it will take a few hours until your job is processed. Please have patience and enjoy a great weekend! | 14:15 | |
| jjohnson2_ | lucasagomes, yeah, I have done very basic testing | 14:15 |
| jjohnson2_ | lucasagomes, and I know that the server side only works in python2, and I have more to do | 14:16 |
| jjohnson2_ | but I might have to park it for a few days while I handle some urgent requests... | 14:16 |
| *** links has quit IRC | 14:17 | |
| lucasagomes | jjohnson2_, yeah, it's all good | 14:18 |
| lucasagomes | thanks for looking into it btw | 14:19 |
| * jroll git clone git://git.openstack.org/openstack-dev/devstack just to spite jlvillal :D | 14:19 | |
| jlvillal | jroll, Good luck with that ;) | 14:19 |
| *** jaypipes has joined #openstack-ironic | 14:20 | |
| lucasagomes | mat128, fyi I've removed the python-daemon dependency from VirtualBMC as part of the refactor, since it's not present in the g-r and I only need a simple double-fork to detach the process from its parent/session | 14:21 |
| * lucasagomes expect it to pass Jenkins now | 14:21 | |
| *** ChrisAusten has joined #openstack-ironic | 14:22 | |
| jjohnson2_ | is there a hard timeline for everything needing Python 3 support I should conspicuously be paying attention to? | 14:22 |
| jroll | jjohnson2_: last week | 14:24 |
| * jroll is joking | 14:24 | |
| lucasagomes | jjohnson2_, not that I know of, but pretty much all ironic projects do support python3 so | 14:24 |
| jjohnson2_ | jroll, let me get my delorean | 14:24 |
| jroll | \o/ | 14:24 |
| lucasagomes | except for pyghmi | 14:24 |
| jroll | lucasagomes: this is waiting on you jfyi https://bugs.launchpad.net/ironic/+bug/1561137 | 14:24 |
| openstack | Launchpad bug 1561137 in Ironic "[RFE] Add support for operators on root device hints" [Wishlist,Confirmed] - Assigned to Lucas Alvares Gomes (lucasagomes) | 14:24 |
| jroll | when you have time, no rush | 14:24 |
| lucasagomes | jroll, no spec needed right? | 14:25 |
| lucasagomes | I think the plumbing patch for IPA already merged (/me checks) | 14:25 |
| lucasagomes | oh it didn't, | 14:26 |
| lucasagomes | https://review.openstack.org/#/c/296539/ | 14:26 |
| jroll | lucasagomes: well, the last comment is "can you update the description to match ___" :P | 14:26 |
| lucasagomes | jroll, fair enough, will do it today | 14:26 |
| *** r-daneel has joined #openstack-ironic | 14:26 | |
| jroll | cool, thanks | 14:26 |
| *** catintheroof has joined #openstack-ironic | 14:27 | |
| *** dprince has quit IRC | 14:37 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Created IronicNodePower resource https://review.openstack.org/313578 | 14:37 |
| *** hshiina has quit IRC | 14:40 | |
| *** absubram has quit IRC | 14:41 | |
| *** e0ne has quit IRC | 14:43 | |
| *** e0ne has joined #openstack-ironic | 14:44 | |
| -openstackstatus- NOTICE: Zuul has been restarted. As a results, we only preserved patches in the gate queue. Be sure to recheck your patches in gerrit if needed. | 14:46 | |
| sambetts | jroll: Does IPA really need i18n? babel is adding 21mb of bloat before compression, and ~10mb after | 14:48 |
| jroll | sambetts: no, who put that in there | 14:48 |
| jroll | heh, it isn't used | 14:48 |
| sambetts | its a dependecy of oslo.i18n | 14:49 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic-python-agent: Remove oslo.i18n https://review.openstack.org/313586 | 14:49 |
| sambetts | hahaha | 14:49 |
| sambetts | awesome! more shrinkage! | 14:49 |
| jroll | let's make sure something else doesn't install it before landing | 14:50 |
| krotscheck | I'm trying to test power state management with the fake driver. Is the "unknown" power state a legitimate state? In a real deployment, when/how would that change? | 14:50 |
| jroll | krotscheck: unknown is the state at enroll time, and when the bmc cannot be reached | 14:51 |
| jroll | is null in DB, None in api response | 14:51 |
| krotscheck | jroll: Right - so the fake driver has allowed me to advance a node to the active state, but the power state never changed. | 14:51 |
| * dtantsur will remove https://github.com/rdo-packages/ironic-python-agent-distgit/blob/rpm-master/openstack-ironic-python-agent.spec#L64 then | 14:51 | |
| jroll | krotscheck: interesting | 14:52 |
| krotscheck | jroll: I'm guessing that's not intentional. | 14:52 |
| sambetts | I guess it makes sense though, if its fake then is it on or is it off, is the cat dead? | 14:53 |
| jroll | krotscheck: here's the code https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/fake.py#L34 | 14:53 |
| jroll | krotscheck: get_power_state is what ensures the bmc can be reached | 14:53 |
| krotscheck | jroll: hrm.... ok, so it just does what I ask it to. | 14:54 |
| jroll | krotscheck: right, and then it should return what you did as current power state | 14:54 |
| krotscheck | Not super useful, but I think i can handle that. | 14:54 |
| *** jaypipes has quit IRC | 14:55 | |
| jroll | well, what would you rather it returns :) | 14:56 |
| *** caiobo is now known as caiobo-afk | 14:57 | |
| krotscheck | jroll: well, it'd be better if it represented a real deployment more. So, say, if the usual place for power state to be resolved is when it moves into the maange state, then that's where that value should be set. | 14:57 |
| krotscheck | But for now I can special case it ;_ | 14:57 |
| krotscheck | :) | 14:57 |
| *** caiobo-afk has quit IRC | 14:57 | |
| jroll | krotscheck: yeah I've thought about that | 14:57 |
| *** rloo has joined #openstack-ironic | 14:59 | |
| * sambetts wonders if we should add a validation into that driver so that it prevents it moving to managable without being in either power on or power offf | 14:59 | |
| sambetts | forcing you to set it | 15:00 |
| krotscheck | sambetts: Is that something which the API client must do, or is that something done by the driver? | 15:01 |
| sambetts | I'm suggesting that we make it so that you need to do a set-power-state on that node while in enroll in order to move it to managable | 15:04 |
| sambetts | ironic node-set-power-state ... etc | 15:04 |
| krotscheck | sambetts: So the user would, after enrolling, have to set power state, and then set manageable? This implies that a managed node cannot be powered off. | 15:05 |
| *** Guest42930 has quit IRC | 15:05 | |
| *** david-lyle has joined #openstack-ironic | 15:06 | |
| krotscheck | Different question - I'm getting 400 invalid when I gtry to issue PUT /state/power reboot - I'm guessing there's no API driven way to manually perform a reboot? | 15:06 |
| sambetts | there is the fake driver just doesn't support it | 15:07 |
| vdrok | sambetts: hrm, I thought that a node in enroll == can not do power control in general case | 15:08 |
| sambetts | hmmm :/ | 15:09 |
| sambetts | the fake driver is a hard one because its decided for testing, and basically just fudges everything | 15:10 |
| *** jaypipes has joined #openstack-ironic | 15:11 | |
| *** mjturek1 has quit IRC | 15:12 | |
| openstackgerrit | Merged openstack/ironic: DevStack: Support to install pyghmi from source https://review.openstack.org/312575 | 15:12 |
| krotscheck | jroll: I see a 'reboot' power state in http://docs.openstack.org/developer/ironic/_modules/ironic/common/states.html, but no way to access it via the API. Is that correct? | 15:14 |
| krotscheck | (As in- ironic handles all the rebooting, you can't do it as a user) | 15:14 |
| sambetts | it should work, I'm pretty sure I've done it | 15:15 |
| jroll | krotscheck: no, PUT /v1/nodes/uuid/states/power {target: reboot} | 15:15 |
| jroll | krotscheck: fake driver might not support it, like sam said | 15:15 |
| krotscheck | oh, right | 15:16 |
| krotscheck | drat | 15:16 |
| krotscheck | Hrm. I don't suppose there's a way for me to ask whether a node's driver supports that. | 15:16 |
| * krotscheck ponders | 15:17 | |
| *** ChrisAusten has quit IRC | 15:18 | |
| *** ChrisAusten has joined #openstack-ironic | 15:20 | |
| jroll | jlvillal: do you have a grenade patchset going for "things to make this work"? | 15:20 |
| jroll | found the create resource bug, I think | 15:20 |
| jlvillal | jroll: I have a bunch of patches in the 'update-projects.sh' file | 15:21 |
| jlvillal | And awesome :) | 15:21 |
| wajdi | ^^ yeah that's what I did to add patches yesterday | 15:21 |
| jroll | jlvillal: well, there's a bug in projects/60_nova/resources.sh | 15:21 |
| jroll | with neutron enabled | 15:21 |
| jlvillal | jroll: https://github.com/JohnVillalovos/devstack-gate-test/blob/master/ansible/roles/devstack-gate/files/update-projects.sh | 15:21 |
| jroll | oh my | 15:22 |
| jroll | okay, I'll spin a new patch for this specifically and you can put it there :) | 15:22 |
| jlvillal | Lions and Tigers and Bears... | 15:22 |
| *** ijw has joined #openstack-ironic | 15:22 | |
| jlvillal | jroll: Thanks | 15:22 |
| JayF | devananda: was out sick yesterday afternoon; re: manual cleaning -- I've never used it and have no access to anything that makes calls to it | 15:22 |
| jlvillal | JayF: Did you get the post-summit illness? :( | 15:23 |
| JayF | No, just sorta a chronic thing that triggers from time to time | 15:23 |
| jlvillal | I've heard a lot of reports from other people too | 15:23 |
| jlvillal | JayF: I hope you feel better today :) | 15:23 |
| NobodyCam | morning JayF hope your feeling better this morning | 15:24 |
| JayF | me too :P | 15:24 |
| JayF | thanks | 15:24 |
| jroll | bbiab/b 170 | 15:28 |
| jroll | ... | 15:28 |
| jroll | bbiab. | 15:28 |
| *** sacharya has joined #openstack-ironic | 15:31 | |
| *** ChrisAusten has quit IRC | 15:32 | |
| *** ChrisAusten has joined #openstack-ironic | 15:32 | |
| *** dmk0202 has quit IRC | 15:33 | |
| *** jtomasek has quit IRC | 15:33 | |
| dtantsur | have a great weekend | 15:33 |
| NobodyCam | you out dtantsur | 15:34 |
| *** dmk0202 has joined #openstack-ironic | 15:34 | |
| *** keedya has joined #openstack-ironic | 15:35 | |
| *** dtantsur is now known as dtantsur|afk | 15:35 | |
| *** jaypipes is now known as leakypipes | 15:36 | |
| *** catintheroof has quit IRC | 15:36 | |
| mgould | o/ dtantsur|afk | 15:36 |
| *** absubram has joined #openstack-ironic | 15:36 | |
| lucasagomes | jjohnson2_, FYI, the patch adding the ironic job in the pyghmi's gate have been merged | 15:38 |
| *** jcoufal has quit IRC | 15:38 | |
| jjohnson2_ | lucasagomes, ok, guess I'll give it a try with recheck this python3 thing | 15:38 |
| keedya | good morning o/ | 15:38 |
| lucasagomes | jjohnson2_, I just did :-) | 15:38 |
| jjohnson2_ | oh | 15:38 |
| keedya | dtantsur: thanks for the review :) | 15:39 |
| *** dmk0202 has quit IRC | 15:39 | |
| lucasagomes | it's non-voting from now, we can make it voting after we gather some data about it (to make sure it's stable) | 15:40 |
| jjohnson2_ | well, it should fail that patch | 15:40 |
| jjohnson2_ | that patch as-is breaks python2 execution of the server at the moment | 15:41 |
| *** fragatina has joined #openstack-ironic | 15:41 | |
| *** rloo has quit IRC | 15:42 | |
| *** rloo has joined #openstack-ironic | 15:42 | |
| *** jcoufal has joined #openstack-ironic | 15:44 | |
| JayF | I'm going to put up two patches to project-config today: 1 to make tinyipa the default ramdisk, and 1 to use ipxe for the majority of our jobs | 15:44 |
| lucasagomes | JayF, cool, I thought the ipxe one was a thing already | 15:46 |
| devananda | morning, all | 15:47 |
| JayF | I have a patch changing the ironic default | 15:48 |
| JayF | in devstack | 15:48 |
| JayF | to enable ipxe | 15:48 |
| JayF | but I think given how the gate jobs are factored we should do those separately | 15:48 |
| *** Nisha has joined #openstack-ironic | 15:49 | |
| *** Nisha has quit IRC | 15:54 | |
| lucasagomes | this change https://review.openstack.org/#/c/308965/ has 2+2 already, someone wants to give it a last review (or not) ? | 15:54 |
| *** rloo has quit IRC | 15:55 | |
| *** rloo has joined #openstack-ironic | 15:56 | |
| *** davideagnello has joined #openstack-ironic | 15:57 | |
| *** ChrisAusten has quit IRC | 15:57 | |
| vdrok | lucasagomes: hey, have you seen something like http://paste.openstack.org/show/496350/ ? | 16:00 |
| vdrok | vbmc command takes ~250 seconds to fail | 16:01 |
| * lucasagomes checks | 16:01 | |
| vdrok | I'm trying to run smoke on 3 tinyipa vms with vbmc | 16:01 |
| *** mjturek1 has joined #openstack-ironic | 16:02 | |
| lucasagomes | vdrok, oh that sucks... is it vbmc or ipmitool itself? | 16:02 |
| * lucasagomes do some tests | 16:02 | |
| vdrok | that I don't know :) | 16:02 |
| vdrok | does vbmc have a separate log? | 16:03 |
| *** yolanda has quit IRC | 16:03 | |
| sambetts | gahhhh why does oslo.concurrency require Babel.... | 16:03 |
| lucasagomes | vdrok, you can configure it yes, if you create a file ~/.vbmc/virtualbmc.conf with a [log] debug=True logfile=<path to your log file> | 16:04 |
| lucasagomes | vdrok, it will log all the actions running on the vbmc daemons | 16:04 |
| vdrok | lucasagomes: thanks, will do | 16:04 |
| lucasagomes | vdrok, so apparently it's ipmitool, I just ran it against a port that is not being listened | 16:04 |
| vdrok | lucasagomes: ipmitool itself seems to be ok when running from commandline at least | 16:04 |
| lucasagomes | and it's taking a lot of time | 16:04 |
| lucasagomes | vdrok, try it out, ipmitool -I lanplus -U admin -P password -H 127.0.0.1 -p 50000 -R 12 -N 5 power status | 16:04 |
| lucasagomes | I wonder if the vbmc daemon stopped listening for some reason (we need the logs for that) | 16:05 |
| lucasagomes | to know that* | 16:05 |
| vdrok | lucasagomes: yep, that port is reported to be listened | 16:05 |
| JayF | I played with vbmc a bit in a highly scaled out onmetal devstack setup | 16:05 |
| JayF | and found a bit of slowness compared to ssh driver | 16:05 |
| JayF | I can reproduce that environment if you all wanna take a look at perf with 32/64 nodes | 16:06 |
| *** derekh has quit IRC | 16:06 | |
| *** mjturek1 has quit IRC | 16:06 | |
| *** aginwala has joined #openstack-ironic | 16:06 | |
| lucasagomes | JayF, oh that's cool to have some data like that | 16:06 |
| *** mjturek1 has joined #openstack-ironic | 16:07 | |
| JayF | I will start spinning one up once I'm done with these meetings and I can drop your key on | 16:08 |
| vdrok | lucasagomes: aha, logging is enabled already :) I see 2016-05-06 18:51:08,184.184 24508 DEBUG VirtualBMC [-] Set boot device called for node-1 with boot device "network" there, no errors | 16:08 |
| vdrok | timing is close, so seems to be it | 16:08 |
| *** johnthetubaguy has quit IRC | 16:08 | |
| lucasagomes | vdrok, oh | 16:08 |
| *** johnthetubaguy has joined #openstack-ironic | 16:08 | |
| lucasagomes | vdrok, there's also this hardcoded timeout on pyghmi's session https://github.com/openstack/pyghmi/blob/65219f2353a62fe38eb4405d9ab70f13cc4f4af9/pyghmi/ipmi/bmc.py#L147 | 16:09 |
| vdrok | lucasagomes: you suggest to try pyghmi? | 16:10 |
| lucasagomes | vdrok, no, I mean vbmc uses pyghmi's on the background | 16:10 |
| vdrok | ah | 16:10 |
| lucasagomes | I wonder if that 30s timeout (because it's a bit high) implicates in something | 16:10 |
| lucasagomes | perhaps making it configurable would be a good thing to have | 16:11 |
| * lucasagomes sends a patch out to pyghmi | 16:11 | |
| *** piet has quit IRC | 16:11 | |
| *** stendulker has joined #openstack-ironic | 16:11 | |
| vdrok | lucasagomes: btw, I think pas-ha was having something similar, https://bugs.launchpad.net/ironic/+bug/1548652, but it was marked invalid by someone | 16:13 |
| openstack | Launchpad bug 1548652 in Ironic "tmp file for ipmi password might get reused which leads to failure" [Undecided,Invalid] - Assigned to Ramamani Yeleswarapu (ramamani-yeleswarapu) | 16:13 |
| vdrok | just recalled he said smth about it | 16:13 |
| lucasagomes | jlvillal, what errors you are seem on grenade? I'm now at this one here http://paste.openstack.org/show/496359/ | 16:14 |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add notification base classes and docs https://review.openstack.org/298461 | 16:14 |
| jlvillal | lucasagomes: I think same. jroll has proposed a patch | 16:14 |
| lucasagomes | jlvillal, before it was complaining about openvswitch-switch not being started, so before I run ./ironic-grenade.sh I've installed openvswitch-switch and started it manually | 16:14 |
| lucasagomes | jlvillal, oh right | 16:14 |
| jlvillal | Jim Rollenhagen proposed openstack-dev/grenade: Allow creating floating ip address with Neutron enabled https://review.openstack.org/313600 | 16:15 |
| lucasagomes | cool | 16:15 |
| * lucasagomes looks | 16:15 | |
| jroll | yeah, likely it | 16:15 |
| jlvillal | lucasagomes: On a different note, I confused a Brazilian guy at work when I said 'Oy' at the coffee machine. | 16:15 |
| jroll | lucasagomes: check grenade log, if it's failing at floating ip create then that should do it | 16:15 |
| jlvillal | He wondered if I was Brazilian for a moment :) | 16:15 |
| lucasagomes | vdrok, a-ha ok so it's ipmitool that is taking a long time to return (4minutes like it says in the bug), see: http://paste.openstack.org/show/496361/ | 16:16 |
| lucasagomes | jlvillal, lol cool | 16:16 |
| lucasagomes | jroll, yeah, that's exactly the failure I'm seeing now, thanks for the patch | 16:17 |
| wajdi | isn't "No nw_info cache associated with instance" resolved by setting the IRONIC_RAMDISK_TYPE to tinyipa? When I ran my test yesterday after setting that, I did not receive that error anymore | 16:17 |
| jroll | lucasagomes: np, haven't tested with the full stack but that should do it | 16:18 |
| jroll | I think/hope :) | 16:18 |
| lucasagomes | yea, I will give it a run here | 16:18 |
| jroll | nw info cache thing is unrelated to my patch and if that's the issue then it won't help | 16:18 |
| lucasagomes | problem is that it takes time :-) | 16:18 |
| lucasagomes | I usually run it and go do something else heh | 16:18 |
| jlvillal | wajdi: I don't know... | 16:18 |
| jlvillal | lucasagomes: Same here. A lot of time... | 16:19 |
| jlvillal | :( | 16:19 |
| lucasagomes | acutally we should look at how to improve that, we need to get faster results | 16:19 |
| jroll | right, that's why I'm not doing that and rather standing up devstack and running through it manually | 16:19 |
| jlvillal | Agreed! I'm just not sure how. | 16:19 |
| lucasagomes | I understand the scripts doesn't care becuase it runs on gate and it should just retunr the failure w/o retrying | 16:19 |
| jroll | well, not doing it for every change | 16:19 |
| lucasagomes | but damn, it takes an hour | 16:19 |
| *** piet has joined #openstack-ironic | 16:19 | |
| lucasagomes | jroll, yeah that's better for smoking tests but we are not there yet | 16:19 |
| *** piet has quit IRC | 16:19 | |
| lucasagomes | it fails before that | 16:19 |
| jroll | lucasagomes: well, that's what I did here, I found the failure and kinda went through the script manually | 16:20 |
| *** frontrunner has joined #openstack-ironic | 16:21 | |
| * jroll takes his dog to the groomers, bbl | 16:22 | |
| *** e0ne has quit IRC | 16:23 | |
| *** klindgren has quit IRC | 16:24 | |
| *** klindgren has joined #openstack-ironic | 16:25 | |
| *** dprince has joined #openstack-ironic | 16:25 | |
| *** catintheroof has joined #openstack-ironic | 16:27 | |
| *** devananda has quit IRC | 16:28 | |
| *** mariojv has quit IRC | 16:28 | |
| *** elfosardo has quit IRC | 16:28 | |
| *** mtreinish has quit IRC | 16:28 | |
| *** stevebaker has quit IRC | 16:29 | |
| *** devananda has joined #openstack-ironic | 16:30 | |
| sambetts | lucasagomes: I really want to work out why devstack is slow as treacle | 16:30 |
| sambetts | lucasagomes: 20+ minutes to install and start 10 or so services is crazy | 16:31 |
| wajdi | I'm sharing my latest run and the output I got w.r.t grenade: http://paste.openstack.org/show/496366/ | 16:31 |
| devananda | JayF: ok, np. I've just put a TODO in the docs, since I don't have any example code to use either for that | 16:31 |
| jlvillal | Anyone know if there is a slide deck on Ironic and what it does on a high level? | 16:33 |
| jlvillal | My program manager is asking me... | 16:33 |
| *** jaybeale has joined #openstack-ironic | 16:33 | |
| *** stevebaker has joined #openstack-ironic | 16:33 | |
| *** jaybeale has quit IRC | 16:34 | |
| sambetts | \0/ latest run using my even further optimised tinyipa took 56 seconds from power on to lookup | 16:34 |
| wajdi | ^^ \o/ | 16:34 |
| mgould | sambetts, woo! | 16:34 |
| sambetts | down to 50mb now | 16:34 |
| lucasagomes | sambetts, :-/ yeah it's slow | 16:35 |
| *** mtreinish has joined #openstack-ironic | 16:35 | |
| lucasagomes | it's procedural as well | 16:35 |
| sambetts | lucasagomes: yeah I think thats what kills it | 16:35 |
| mgould | sambetts, I have updated the etherpad accordingly :_) | 16:35 |
| mgould | :-) | 16:35 |
| mgould | sambetts, did you manage to eliminate Babel? | 16:36 |
| sambetts | mgould: we can work out the time to transfer 1 mb now because we know how long it took to boot 59 mb and how long it took to boot 50mb and we can work out the difference I think right? | 16:36 |
| *** mari0jv has joined #openstack-ironic | 16:36 | |
| *** chopmann has quit IRC | 16:36 | |
| sambetts | mgould: No :( because its imported by other oslo stuff, but I've manually killed a bunch of the .dat files that are shipped as part of Babel leaving only en_US which has eliminated like 20mb | 16:37 |
| sambetts | pre-compress | 16:37 |
| *** mariojv has joined #openstack-ironic | 16:38 | |
| devananda | sambetts: be mindful of making timing assumptions in the upstream test envs - there's a fair amount of variability between node providers and between test runs in general | 16:38 |
| devananda | sambetts: also, sub-minute boot is great progress! | 16:39 |
| wajdi | mgould: are you still receiving: "Unable to locate package liberasurecode-dev"? | 16:39 |
| wajdi | or is that old.. | 16:39 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/pyghmi: DO NOT REVIEW, TESTING GATE https://review.openstack.org/313638 | 16:40 |
| *** dmk0202 has joined #openstack-ironic | 16:40 | |
| *** elfosardo has joined #openstack-ironic | 16:40 | |
| *** mgould has quit IRC | 16:40 | |
| *** athomas has quit IRC | 16:45 | |
| *** aginwala has quit IRC | 16:46 | |
| klindgren | hello - I am trying to figure out what the status is of the nova -> ironic interaction re: scheduling. Specifically re: availability-zones and the interaction between ironic and nova-compute. | 16:46 |
| klindgren | http://lists.openstack.org/pipermail/openstack-dev/2015-December/082011.html | 16:46 |
| sambetts | devananda: very true, even though I'm running a similar size VM for running my devstack mine tends complete a full BareMetalBasicOps about 20-30seconds faster than a couple of gate runs I've seen | 16:47 |
| klindgren | is ^^ still the state of the things. I was in the newton nova -> Ironic cross project talk where you were talking about raid, which from an operator view was painful. | 16:47 |
| JayF | klindgren: http://specs.openstack.org/openstack/ironic-specs/priorities/newton-priorities.html | 16:48 |
| JayF | klindgren: we have two of the needed APIs for that as priorities this cycle | 16:48 |
| devananda | klindgren: not much has changed as far as nova/ironic scheduling *yet*. A lot of things are being designed / planned for this cycle. | 16:48 |
| JayF | klindgren: and the multiple compute hosts piece | 16:48 |
| lucasagomes | sambetts, dtantsur|afk mat128 sorry on the carrier stuff I wasn't paying much attention on IRC, I see there's no consensus whether we should or should not rely on that | 16:49 |
| lucasagomes | so I've revoked the approval | 16:49 |
| *** piet has joined #openstack-ironic | 16:49 | |
| lucasagomes | approval of that patch* | 16:49 |
| klindgren | jayf, devananda thanks | 16:49 |
| *** electrofelix has quit IRC | 16:50 | |
| *** garthb has joined #openstack-ironic | 16:50 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Power state may now be managed https://review.openstack.org/313641 | 16:50 |
| *** Nisha has joined #openstack-ironic | 16:50 | |
| *** [4]cdearborn has quit IRC | 16:53 | |
| *** ChrisAusten has joined #openstack-ironic | 16:55 | |
| *** dmk0202 has quit IRC | 16:55 | |
| *** dmk0202 has joined #openstack-ironic | 16:57 | |
| *** mari0jv has left #openstack-ironic | 16:57 | |
| lucasagomes | jroll, fwiw, your patch took me further in the grenade thing http://paste.openstack.org/show/496372/ | 16:57 |
| * lucasagomes will create some loop devices manually | 16:58 | |
| *** wajdi has quit IRC | 16:58 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic-specs: Place priorities list in reverse chronological order https://review.openstack.org/312794 | 17:00 |
| jlvillal | lucasagomes: Sweet. Progress :) | 17:00 |
| jlvillal | I'm still running stuff... | 17:01 |
| lucasagomes | yeah | 17:01 |
| *** dmk0202 has quit IRC | 17:01 | |
| * lucasagomes created 200 loop devices | 17:02 | |
| *** mjturek1 has quit IRC | 17:02 | |
| *** ijw_ has joined #openstack-ironic | 17:03 | |
| *** ijw has quit IRC | 17:06 | |
| *** jistr has quit IRC | 17:07 | |
| *** stendulker_ has joined #openstack-ironic | 17:12 | |
| JayF | lucasagomes: around? still interested in poking at the big devstack box | 17:14 |
| lucasagomes | JayF, still around yeah, not for long tho :-/ | 17:14 |
| lucasagomes | JayF, you have it setup there? | 17:14 |
| JayF | got a key? | 17:14 |
| *** wajdi has joined #openstack-ironic | 17:14 | |
| JayF | yeah, I never tore down my old ones | 17:14 |
| vdrok | lucasagomes: btw, I think depends-on on project-config patches does not work | 17:15 |
| JayF | vdrok: that's correct | 17:15 |
| lucasagomes | vdrok, :-( yeah it failed with the same error | 17:15 |
| *** stendulker has quit IRC | 17:15 | |
| lucasagomes | JayF, vdrok thanks... not sure how I test it | 17:15 |
| JayF | can you link me that change? | 17:15 |
| vdrok | lucasagomes: just merge it :) | 17:15 |
| lucasagomes | JayF, pm'ed | 17:16 |
| lucasagomes | JayF, https://review.openstack.org/313638 | 17:16 |
| wajdi | when running grenade tests a second time, do you simply remove the workspace and run again? Or do you bring up a new stack o na clean vm? | 17:17 |
| JayF | lucasagomes: +1 to just merging that | 17:17 |
| wajdi | want to make sure that my testing is valid. I have been removing the workspace and just running again. | 17:17 |
| JayF | lucasagomes: the p-c one you dep on | 17:17 |
| *** jjohnson2 has joined #openstack-ironic | 17:18 | |
| lucasagomes | thanks, I've set workflow to +0 and added a comment | 17:18 |
| lucasagomes | will wait someone from -infra | 17:19 |
| JayF | you link it to anyone there? | 17:19 |
| * JayF will | 17:19 | |
| devananda | anyone tried calling the send_raw passthru method of the ipmitool drivers? | 17:19 |
| *** jtomasek has joined #openstack-ironic | 17:21 | |
| *** jjohnson2_ has quit IRC | 17:22 | |
| jlvillal | As an FYI, I have pushed a new version of https://github.com/JohnVillalovos/devstack-gate-test with jroll's fix for an issue with resource phase create | 17:24 |
| jlvillal | NobodyCam: TheJulia cinerama lucasagomes JayF sambetts ^^^ | 17:25 |
| *** fragatina has quit IRC | 17:26 | |
| cinerama | thanks jlvillal | 17:26 |
| *** fragatina has joined #openstack-ironic | 17:26 | |
| *** aginwala has joined #openstack-ironic | 17:28 | |
| *** fragatina has quit IRC | 17:34 | |
| *** [4]cdearborn has joined #openstack-ironic | 17:34 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Add Hardware inventory from Lenovo Agentless https://review.openstack.org/312568 | 17:35 |
| *** aginwala has quit IRC | 17:37 | |
| *** klindgren has quit IRC | 17:40 | |
| *** irf has joined #openstack-ironic | 17:40 | |
| *** klindgren has joined #openstack-ironic | 17:41 | |
| mgagne | has anyone had an issue where nova-compute no longer report heartbeat for all hypervisors? compute_nodes.updated_at field in database is stalled until I restart the nova-compute service and nova-scheduler complains about stalled hypervisors. | 17:41 |
| mgagne | "has not been heard from in a while" | 17:42 |
| *** lintan has quit IRC | 17:44 | |
| *** mjturek1 has joined #openstack-ironic | 17:44 | |
| *** lintan has joined #openstack-ironic | 17:44 | |
| vdrok | lucasagomes: I reopened that bug (1548652), with incomplete status, will try to investigate a bit on the weekend | 17:46 |
| *** yolanda has joined #openstack-ironic | 17:46 | |
| JayF | win 32 | 17:46 |
| JayF | whoops | 17:46 |
| vdrok | good night everyone! have a good weekend | 17:46 |
| lucasagomes | vdrok, thanks a lot | 17:47 |
| lucasagomes | night | 17:47 |
| *** aginwala has joined #openstack-ironic | 17:47 | |
| *** xhku has joined #openstack-ironic | 17:47 | |
| jlvillal | vdrok: Dos ve dos | 17:48 |
| * sambetts is heading off | 17:53 | |
| *** sambetts is now known as sambetts|afk | 17:53 | |
| *** yolanda has quit IRC | 17:54 | |
| *** fragatina has joined #openstack-ironic | 17:55 | |
| *** stendulker_ has quit IRC | 17:55 | |
| *** fragatin_ has joined #openstack-ironic | 17:55 | |
| *** ohamada has quit IRC | 17:56 | |
| jlvillal | sambetts|afk: Ciao | 17:57 |
| *** Guest60954 has quit IRC | 17:57 | |
| *** yarkot1 has quit IRC | 17:57 | |
| *** akrish9 has joined #openstack-ironic | 17:58 | |
| *** fragatina has quit IRC | 17:59 | |
| *** Nisha has quit IRC | 18:01 | |
| *** joprovost has quit IRC | 18:05 | |
| *** joprovost has joined #openstack-ironic | 18:06 | |
| *** yarkot1 has joined #openstack-ironic | 18:07 | |
| *** lucasagomes is now known as lucas-afk | 18:07 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Update the api-ref documentation for Ports https://review.openstack.org/313187 | 18:13 |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Update the api-ref documentation for Root and Nodes https://review.openstack.org/312795 | 18:13 |
| devananda | moar docs | 18:14 |
| *** jjohnson2 has quit IRC | 18:18 | |
| *** jjohnson2_ has joined #openstack-ironic | 18:18 | |
| JayF | TIL about less -R -- it properly renders the colors in logfiles generated by devstack | 18:20 |
| *** martiby has joined #openstack-ironic | 18:26 | |
| *** wajdi has quit IRC | 18:26 | |
| *** rloo has quit IRC | 18:30 | |
| *** irf has quit IRC | 18:36 | |
| *** piet has quit IRC | 18:38 | |
| *** e0ne has joined #openstack-ironic | 18:39 | |
| *** keedya has quit IRC | 18:40 | |
| *** catintheroof has quit IRC | 18:41 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Add Hardware inventory from Lenovo Agentless https://review.openstack.org/312568 | 18:42 |
| mgagne | so I'm still trying to investigate that issue I have with compute_nodes entries in nova database getting stalled with Ironic | 18:56 |
| mgagne | in our cloud/virtual cell, everything is fine, compute_nodes entries are updated (updated_at field) | 18:57 |
| mgagne | but with baremetal, it's stalled | 18:57 |
| *** aginwala has quit IRC | 19:03 | |
| *** sivaramakrishna has joined #openstack-ironic | 19:04 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Add Hardware inventory from Lenovo Agentless https://review.openstack.org/312568 | 19:06 |
| *** wajdi has joined #openstack-ironic | 19:08 | |
| *** ifarkas has quit IRC | 19:08 | |
| *** aginwala has joined #openstack-ironic | 19:09 | |
| *** e0ne has quit IRC | 19:11 | |
| *** keedya has joined #openstack-ironic | 19:20 | |
| *** aginwala has quit IRC | 19:20 | |
| *** wajdi has quit IRC | 19:23 | |
| *** rcernin has joined #openstack-ironic | 19:26 | |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Add network drivers https://review.openstack.org/285852 | 19:26 |
| *** chenke_ has quit IRC | 19:30 | |
| openstackgerrit | Sivaramakrishna Garimella proposed openstack/ironic: Update the deploy drivers with network flipping logic https://review.openstack.org/213262 | 19:31 |
| *** keedya_ has joined #openstack-ironic | 19:31 | |
| mgagne | ok found the issue | 19:32 |
| mgagne | this line makes it so scheduler is not updated with latest compute node info if resource usage didn't change: https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L632 | 19:33 |
| *** yonglihe has quit IRC | 19:33 | |
| mgagne | this however means that ComputeFilter will complain since updated_at field will go outdated rapidly | 19:33 |
| mgagne | jroll: ^ | 19:34 |
| jroll | mgagne: that's fun, file a nova bug? | 19:34 |
| mgagne | has anyone encounter the bug yet? | 19:34 |
| jroll | is that the heartbeat though? | 19:34 |
| *** absubram has quit IRC | 19:35 | |
| *** keedya has quit IRC | 19:35 | |
| jroll | I haven't seen that afaik | 19:35 |
| mgagne | I'm not sure what is calling that thing | 19:35 |
| mgagne | maybe there is 2 code path to update the hearbeat | 19:35 |
| jroll | could be | 19:35 |
| openstackgerrit | Sam Betts proposed openstack/ironic-python-agent: TinyIPA: Shave off some file size from tinyipa ramdisk https://review.openstack.org/313675 | 19:35 |
| sambetts|afk | down to ~49mb ^ | 19:36 |
| sambetts|afk | night all | 19:36 |
| jroll | mgagne: I honestly have no idea how this works, I'll have to look later | 19:38 |
| mgagne | sure | 19:38 |
| *** sivaramakrishna has quit IRC | 19:38 | |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Add Hardware inventory from Lenovo Agentless https://review.openstack.org/312568 | 19:46 |
| *** piet has joined #openstack-ironic | 19:49 | |
| *** rloo has joined #openstack-ironic | 19:53 | |
| *** dprince has quit IRC | 19:59 | |
| *** mkovacik_ has joined #openstack-ironic | 20:05 | |
| *** dmk0202 has joined #openstack-ironic | 20:05 | |
| mgagne | jroll: bug #1579213 | 20:05 |
| openstack | bug 1579213 in OpenStack Compute (nova) "ComputeFilter fails because compute node has not been heard from in a while" [Undecided,New] https://launchpad.net/bugs/1579213 | 20:05 |
| *** joprovost has quit IRC | 20:05 | |
| *** wajdi has joined #openstack-ironic | 20:06 | |
| *** joprovost has joined #openstack-ironic | 20:06 | |
| *** aginwala has joined #openstack-ironic | 20:07 | |
| *** mkovacik has quit IRC | 20:07 | |
| *** joprovost has quit IRC | 20:11 | |
| *** mlavalle has joined #openstack-ironic | 20:14 | |
| *** xhku has quit IRC | 20:15 | |
| *** aginwala has quit IRC | 20:17 | |
| *** wajdi_ has joined #openstack-ironic | 20:19 | |
| *** wajdi has quit IRC | 20:19 | |
| *** keedya_ has quit IRC | 20:19 | |
| *** joprovost has joined #openstack-ironic | 20:20 | |
| *** keedya has joined #openstack-ironic | 20:24 | |
| *** garthb_ has joined #openstack-ironic | 20:27 | |
| *** sinval has quit IRC | 20:28 | |
| *** wajdi_ has quit IRC | 20:28 | |
| *** Sukhdev has joined #openstack-ironic | 20:29 | |
| *** garthb has quit IRC | 20:29 | |
| *** e0ne has joined #openstack-ironic | 20:29 | |
| *** openstackgerrit has quit IRC | 20:32 | |
| *** joprovost has quit IRC | 20:33 | |
| *** joprovost has joined #openstack-ironic | 20:33 | |
| *** openstackgerrit has joined #openstack-ironic | 20:33 | |
| *** xhku has joined #openstack-ironic | 20:33 | |
| *** joprovost has quit IRC | 20:38 | |
| *** xhku has quit IRC | 20:38 | |
| *** e0ne has quit IRC | 20:39 | |
| *** e0ne has joined #openstack-ironic | 20:40 | |
| *** e0ne has quit IRC | 20:40 | |
| *** Sukhdev has quit IRC | 20:40 | |
| *** rloo has quit IRC | 20:41 | |
| *** Sukhdev has joined #openstack-ironic | 20:41 | |
| devananda | jroll: the adopt-node-state patch has 3 +2's -- unless you have a reason to hold it back, I'm happy with landing it | 20:47 |
| openstackgerrit | Jarrod Johnson proposed openstack/pyghmi: Add Hardware inventory from Lenovo Agentless https://review.openstack.org/312568 | 20:47 |
| jroll | devananda: I haven't looked at it, I trust y'all | 20:47 |
| *** baoli_ has quit IRC | 20:50 | |
| *** fragatin_ has quit IRC | 20:53 | |
| *** joprovost has joined #openstack-ironic | 20:54 | |
| *** dmk0202 has quit IRC | 20:58 | |
| *** Sukhdev has quit IRC | 21:02 | |
| *** baoli has joined #openstack-ironic | 21:06 | |
| *** jjohnson2_ has quit IRC | 21:11 | |
| *** joprovost has quit IRC | 21:11 | |
| openstackgerrit | Merged openstack/ironic: API: Check for reserved words when naming a node https://review.openstack.org/308965 | 21:14 |
| *** baoli has quit IRC | 21:16 | |
| *** ayoung has quit IRC | 21:20 | |
| *** Sukhdev has joined #openstack-ironic | 21:23 | |
| *** ametts has quit IRC | 21:24 | |
| JayF | is there a reason we run both a tinyipa and a tinyipa-src job on Ironic proper? | 21:26 |
| JayF | About to push a proposed change to make tinyipa the default, and it seems to me the -src is a job that should only be running on IPA, in the same way we did with CoreOS | 21:26 |
| *** leakypipes has quit IRC | 21:35 | |
| *** jcoufal has quit IRC | 21:35 | |
| *** aginwala has joined #openstack-ironic | 21:38 | |
| *** aginwala has quit IRC | 21:40 | |
| *** fragatina has joined #openstack-ironic | 21:42 | |
| *** fragatin_ has joined #openstack-ironic | 21:43 | |
| *** aginwala has joined #openstack-ironic | 21:44 | |
| *** fragatina has quit IRC | 21:47 | |
| *** thiagop has quit IRC | 21:47 | |
| *** penick has joined #openstack-ironic | 21:47 | |
| *** davideagnello has quit IRC | 21:53 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Update the api-ref documentation for Drivers https://review.openstack.org/313708 | 21:54 |
| *** thrash is now known as thrash|wknd | 21:54 | |
| JayF | jroll: around? | 21:54 |
| JayF | or anyone, really, who has interest and possibly insight into how we want to adjust jobs with tinyipa being the default | 21:55 |
| openstackgerrit | Clif Houck proposed openstack/ironic-specs: Add spec for image caching to nodes via agent driver https://review.openstack.org/310594 | 21:56 |
| *** piet has quit IRC | 21:59 | |
| *** rbudden has quit IRC | 22:08 | |
| *** Sukhdev has quit IRC | 22:10 | |
| bodepd | bifrost question. Considering patching the csv code in bifrost_inventory.py to support netmask and gateway, not sure if it's worth doing (I'd either add it there or just create my own custom templates to generate the json) | 22:10 |
| JayF | TheJulia: ^ | 22:12 |
| bodepd | The question is if a patch would be considered since the csv stuff is already a bit of a mess (and any patch would add columns making it messier) | 22:12 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/bifrost: Updated from global requirements https://review.openstack.org/313712 | 22:15 |
| JayF | bodepd: TheJulia would be the best person to answer that, but I know generally they are accepting of new features and such | 22:16 |
| bodepd | All in all, not a bad experience with bifrost. The main thing that bit me was switching to virbr0 to a real interface | 22:16 |
| *** ayoung has joined #openstack-ironic | 22:16 | |
| JayF | RFR Jay Faulkner proposed openstack-infra/project-config: Use TinyIPA and iPXE for most Ironic jobs https://review.openstack.org/313717 | 22:16 |
| bodepd | JayF: sure. thanks. They've already accepted a patch from me, this is more of a question if that code is a dead-end or something worth contributing to. | 22:17 |
| JayF | My statement would be that bifrost is young enough that I doubt any code is considered a dead end at this point | 22:17 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/313729 | 22:17 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/313730 | 22:17 |
| JayF | but I don't use it ans as always can be wrong :) | 22:17 |
| *** [4]cdearborn has quit IRC | 22:21 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/313756 | 22:22 |
| *** keedya has quit IRC | 22:24 | |
| jroll | JayF: adjust jobs how | 22:29 |
| *** ChrisAusten has quit IRC | 22:30 | |
| *** ChrisAusten has joined #openstack-ironic | 22:31 | |
| JayF | krotscheck: https://review.openstack.org/313717 | 22:32 |
| JayF | erm | 22:32 |
| JayF | jroll: https://review.openstack.org/313717 | 22:32 |
| JayF | that's what I get for touch typing with tab completion, sorry krots | 22:32 |
| *** david-lyle has quit IRC | 22:35 | |
| *** david-lyle has joined #openstack-ironic | 22:37 | |
| jroll | JayF: at a glance that seems fine to me | 22:38 |
| JayF | it's not passing tests, there's a line here: | 22:38 |
| JayF | - name: ^gate-tempest-dsvm-ironic-pxe_ipa-tinyipa.*$ | 22:38 |
| JayF | branch: ^(?!stable/(kilo|liberty)).*$ | 22:38 |
| JayF | referencing a job I removed that I think is failing tests | 22:38 |
| JayF | and it makes me wonder if these changes would break stable jobs, since clearly they're marked to not run the tinyipa job | 22:38 |
| JayF | so wondering if that means I have to leave the tinyipa job named tinyipa, and let the old job stay the exact same to not break stable jobs | 22:39 |
| * JayF barely understands how this works | 22:39 | |
| jroll | ooo. | 22:39 |
| jroll | yeah, liberty/kilo ironic/devstack doesn't support tinyipa | 22:39 |
| jroll | I gotta bounce, but that's certainly something to think about | 22:40 |
| JayF | so do it that way, then? | 22:40 |
| jroll | you could do something in the job def | 22:40 |
| JayF | like if branch=, do it diff? | 22:40 |
| JayF | I'll lookup for an example of that | 22:40 |
| jroll | with like... if ZUUL_REF == kilo or liberty: force coreos | 22:40 |
| jroll | if [ "$ZUUL_BRANCH" != "stable/kilo" -a "$ZUUL_BRANCH" != "stable/liberty" ] ; then | 22:40 |
| jroll | there's already one in that section | 22:40 |
| JayF | as a note I'm out on Monday, so if anyone wants to poke this when I'm gone feel free | 22:40 |
| JayF | but i'll fix that up | 22:40 |
| jroll | cool | 22:41 |
| jroll | also I'd like something that's still !ipxe | 22:41 |
| jroll | (there might be already and I missed it) | 22:41 |
| jroll | but really gotta go, see ya | 22:41 |
| *** mkovacik__ has joined #openstack-ironic | 22:42 | |
| JayF | I left pxe_ssh as !ipxe | 22:43 |
| *** wajdi has joined #openstack-ironic | 22:44 | |
| *** mkovacik_ has quit IRC | 22:45 | |
| *** ChrisAusten has quit IRC | 22:47 | |
| *** wajdi_ has joined #openstack-ironic | 22:49 | |
| *** wajdi has quit IRC | 22:52 | |
| *** r-daneel has quit IRC | 22:52 | |
| *** wajdi_ is now known as wajdi | 22:55 | |
| openstackgerrit | Clint 'SpamapS' Byrum proposed openstack/bifrost: Convert ironic.conf to template generated https://review.openstack.org/305634 | 22:57 |
| *** aginwala has quit IRC | 23:12 | |
| *** daemontool has quit IRC | 23:12 | |
| *** ppiela has quit IRC | 23:20 | |
| *** garthb_ has quit IRC | 23:38 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Make iPXE + TinyIPA the defaults for devstack https://review.openstack.org/313035 | 23:39 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Make iPXE + TinyIPA the defaults for devstack https://review.openstack.org/313035 | 23:40 |
| *** xhku has joined #openstack-ironic | 23:45 | |
| *** absubram has joined #openstack-ironic | 23:48 | |
| *** absubram_ has joined #openstack-ironic | 23:49 | |
| *** xhku has quit IRC | 23:49 | |
| *** absubram has quit IRC | 23:53 | |
| *** absubram_ is now known as absubram | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!