| *** achanda_ has quit IRC | 00:04 | |
| *** hj-hp has joined #openstack-ironic | 00:04 | |
| *** achanda has joined #openstack-ironic | 00:05 | |
| *** spandhe has quit IRC | 00:07 | |
| TheJulia | JayF: bad config param/file names :) | 00:08 |
|---|---|---|
| *** hj-hp has quit IRC | 00:15 | |
| *** absubram has joined #openstack-ironic | 00:17 | |
| *** davide has quit IRC | 00:20 | |
| *** achanda has quit IRC | 00:20 | |
| *** zhidong has joined #openstack-ironic | 00:23 | |
| TheJulia | JayF: Left a comment in the review | 00:25 |
| *** penick has quit IRC | 00:25 | |
| *** penick has joined #openstack-ironic | 00:29 | |
| *** yuanying has joined #openstack-ironic | 00:30 | |
| *** mtanino has quit IRC | 00:33 | |
| *** penick has quit IRC | 00:34 | |
| TheJulia | jroll: I think the most recent test of 166903 just failed because of a potential race with another test. The mock call is on the specific default driver instead something in the middle in the driver layer. *headache* | 00:37 |
| *** jmccrory has quit IRC | 00:45 | |
| *** jcoufal has quit IRC | 00:46 | |
| *** Marga_ has quit IRC | 00:48 | |
| *** jerryz has joined #openstack-ironic | 01:04 | |
| *** Sukhdev has joined #openstack-ironic | 01:08 | |
| *** Marga_ has joined #openstack-ironic | 01:16 | |
| *** Marga_ has quit IRC | 01:19 | |
| *** zyu4_ has joined #openstack-ironic | 01:25 | |
| *** Marga_ has joined #openstack-ironic | 01:25 | |
| *** zhenguo has quit IRC | 01:37 | |
| *** korekhov_ has joined #openstack-ironic | 01:52 | |
| *** Haomeng|2 has quit IRC | 01:53 | |
| *** kan__ has joined #openstack-ironic | 01:55 | |
| *** korekhov has quit IRC | 01:56 | |
| *** kan__ has quit IRC | 01:56 | |
| *** korekhov_ has quit IRC | 01:56 | |
| *** chenglch has joined #openstack-ironic | 02:00 | |
| *** anderbubble has joined #openstack-ironic | 02:00 | |
| *** zhidong has quit IRC | 02:12 | |
| *** zyu4_ has quit IRC | 02:12 | |
| *** Marga_ has quit IRC | 02:13 | |
| *** ijw_ has quit IRC | 02:16 | |
| *** bradjones has quit IRC | 02:16 | |
| *** spandhe has joined #openstack-ironic | 02:16 | |
| *** harlowja is now known as harlowja_away | 02:17 | |
| *** spandhe_ has joined #openstack-ironic | 02:20 | |
| *** spandhe has quit IRC | 02:21 | |
| *** spandhe_ is now known as spandhe | 02:21 | |
| *** bradjones has joined #openstack-ironic | 02:21 | |
| *** bradjones has quit IRC | 02:21 | |
| *** bradjones has joined #openstack-ironic | 02:21 | |
| *** absubram has quit IRC | 02:21 | |
| *** kan__ has joined #openstack-ironic | 02:30 | |
| *** Sukhdev has quit IRC | 02:31 | |
| *** penick has joined #openstack-ironic | 02:34 | |
| *** Haomeng has joined #openstack-ironic | 02:38 | |
| *** ramineni has joined #openstack-ironic | 02:39 | |
| *** r-daneel has quit IRC | 02:41 | |
| *** achanda has joined #openstack-ironic | 02:45 | |
| Haomeng | morning | 02:46 |
| Haomeng | GheRivero: ping | 02:47 |
| Haomeng | GheRivero: I have idea about the i18n supporting for ironic exceptions, can we add code into - https://github.com/openstack/ironic/blob/master/ironic/common/exception.py#L50 base exception class to support exceptiom message translation, any comments? | 02:48 |
| Haomeng | GheRivero: call oslo_i18n.translate() method to translate the error message based on the current user locale | 02:50 |
| Haomeng | GheRivero: based on our current ironic code, we did not call oslo_i18n.translate(), so I dont think our Ironic support i18n for messages, but for log, maybe it works with log libs | 02:51 |
| *** kevinbenton has joined #openstack-ironic | 02:54 | |
| *** Marga_ has joined #openstack-ironic | 02:56 | |
| *** Marga_ has quit IRC | 02:59 | |
| *** Marga_ has joined #openstack-ironic | 02:59 | |
| *** ChuckC has quit IRC | 03:00 | |
| Haomeng | GheRivero: or we should add oslo_i18n.translate() call into this class - https://github.com/openstack/ironic/blob/master/ironic/api/middleware/parsable_error.py#L37 | 03:17 |
| *** anderbubble has quit IRC | 03:17 | |
| *** lintan_ has joined #openstack-ironic | 03:27 | |
| *** lintan has quit IRC | 03:28 | |
| *** morgabra has quit IRC | 03:28 | |
| *** absubram has joined #openstack-ironic | 03:29 | |
| *** morgabra has joined #openstack-ironic | 03:29 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Fallback to RIBCL https://review.openstack.org/167151 | 03:32 |
| *** purp_2 has joined #openstack-ironic | 03:36 | |
| *** purp has quit IRC | 03:37 | |
| *** achanda has quit IRC | 03:38 | |
| *** purp_2 is now known as purp | 03:41 | |
| *** anderbubble has joined #openstack-ironic | 03:52 | |
| *** Sukhdev has joined #openstack-ironic | 03:54 | |
| *** Nisha has joined #openstack-ironic | 04:03 | |
| *** krtaylor has quit IRC | 04:05 | |
| *** krtaylor has joined #openstack-ironic | 04:08 | |
| *** naohirot has joined #openstack-ironic | 04:17 | |
| *** bradjones has quit IRC | 04:24 | |
| *** achanda has joined #openstack-ironic | 04:27 | |
| *** bradjones has joined #openstack-ironic | 04:28 | |
| *** bradjones has quit IRC | 04:29 | |
| *** bradjones has joined #openstack-ironic | 04:29 | |
| *** saripurigopi has joined #openstack-ironic | 04:40 | |
| *** sandhya_ has joined #openstack-ironic | 04:41 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Refactor code in RIS https://review.openstack.org/167117 | 04:47 |
| *** Marga_ has quit IRC | 04:49 | |
| openstackgerrit | Merged stackforge/proliantutils: Fix exceptions https://review.openstack.org/167139 | 04:52 |
| *** rameshg87 has joined #openstack-ironic | 05:02 | |
| rameshg87 | good morning ironic | 05:02 |
| Haomeng | rameshg87: morning:) | 05:03 |
| rameshg87 | hello Haomeng, morning :) | 05:05 |
| Haomeng | rameshg87: :) | 05:05 |
| *** absubram has quit IRC | 05:19 | |
| *** Sukhdev has quit IRC | 05:22 | |
| *** spandhe has quit IRC | 05:23 | |
| mrda | hi rameshg87 and Haomeng | 05:27 |
| rameshg87 | hello mrda | 05:34 |
| Haomeng | mrda: hello, mrda, morning:) | 05:36 |
| *** anderbubble has quit IRC | 05:38 | |
| *** penick has quit IRC | 05:47 | |
| *** jamielennox is now known as jamielennox|away | 05:49 | |
| *** Marga_ has joined #openstack-ironic | 05:52 | |
| *** chenglch has quit IRC | 06:03 | |
| *** killer_prince has quit IRC | 06:14 | |
| *** killer_prince has joined #openstack-ironic | 06:17 | |
| *** killer_prince is now known as lazy_prince | 06:17 | |
| *** achanda has quit IRC | 06:18 | |
| *** Marga_ has quit IRC | 06:20 | |
| *** achanda has joined #openstack-ironic | 06:21 | |
| *** korekhov has joined #openstack-ironic | 06:25 | |
| *** TheJulia has quit IRC | 06:29 | |
| *** ukalifon has joined #openstack-ironic | 06:42 | |
| *** achanda has quit IRC | 06:44 | |
| *** achanda has joined #openstack-ironic | 06:45 | |
| *** achanda has quit IRC | 06:49 | |
| *** achanda has joined #openstack-ironic | 06:50 | |
| *** achanda has quit IRC | 06:51 | |
| *** kalpase has joined #openstack-ironic | 06:54 | |
| *** korekhov has quit IRC | 07:04 | |
| *** korekhov has joined #openstack-ironic | 07:05 | |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: supports i18n translation https://review.openstack.org/167507 | 07:05 |
| *** korekhov has quit IRC | 07:09 | |
| *** Nisha has quit IRC | 07:18 | |
| *** Nisha has joined #openstack-ironic | 07:35 | |
| Haomeng | GheRivero: I proposed a fix to enable i18n translation - https://review.openstack.org/#/c/167507/ , would you please help to take a quick review if you have time ? thanks! | 07:39 |
| Haomeng | GheRivero: :) | 07:45 |
| *** chenglch has joined #openstack-ironic | 07:46 | |
| *** chlong has quit IRC | 07:52 | |
| *** mrda is now known as mrda-away | 07:55 | |
| *** yuanying has quit IRC | 08:05 | |
| *** Nisha has quit IRC | 08:32 | |
| *** Haomeng has quit IRC | 08:33 | |
| *** Haomeng|2 has joined #openstack-ironic | 08:33 | |
| *** pas-ha has joined #openstack-ironic | 08:34 | |
| *** jistr has joined #openstack-ironic | 08:38 | |
| *** ndipanov has joined #openstack-ironic | 08:42 | |
| *** lazy_prince has quit IRC | 08:46 | |
| *** dtantsur|afk is now known as dtantsur | 08:48 | |
| dtantsur | Morning! | 08:48 |
| *** jgrimm is now known as zz_jgrimm | 08:51 | |
| *** andreykurilin_ has joined #openstack-ironic | 08:52 | |
| *** Nisha has joined #openstack-ironic | 08:53 | |
| *** pradipta has joined #openstack-ironic | 08:54 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Switch to oslo.config https://review.openstack.org/163281 | 08:56 |
| openstackgerrit | Merged stackforge/ironic-discoverd: Update from global requirements https://review.openstack.org/167266 | 08:57 |
| *** dlpartain has joined #openstack-ironic | 09:01 | |
| *** hj-hp has joined #openstack-ironic | 09:03 | |
| *** dlpartain has left #openstack-ironic | 09:03 | |
| *** mgoddard has joined #openstack-ironic | 09:06 | |
| *** hj-hp has quit IRC | 09:10 | |
| *** igordcard_ has joined #openstack-ironic | 09:12 | |
| *** hj-hp has joined #openstack-ironic | 09:12 | |
| *** lucasagomes has joined #openstack-ironic | 09:18 | |
| *** rbogorodskiy has joined #openstack-ironic | 09:19 | |
| rbogorodskiy | hi! is it possible to use chassis metadata (the extra field) for scheduling (e.g. to filter some hosts based on the values of the metadata)? | 09:20 |
| *** hj-hp has quit IRC | 09:20 | |
| *** hj-hp has joined #openstack-ironic | 09:21 | |
| rameshg87 | rbogorodskiy: right now, no | 09:23 |
| rameshg87 | rbogorodskiy: only node properties are used for scheduling | 09:23 |
| rbogorodskiy | rameshg87: I see, thanks for the information | 09:24 |
| *** andreykurilin_ has quit IRC | 09:25 | |
| *** andreykurilin__ has joined #openstack-ironic | 09:25 | |
| *** tiagogomes_ has joined #openstack-ironic | 09:29 | |
| *** hj-hp has quit IRC | 09:37 | |
| *** hj-hp has joined #openstack-ironic | 09:39 | |
| *** pradipta has quit IRC | 09:41 | |
| *** kan__ has quit IRC | 09:44 | |
| *** pradipta has joined #openstack-ironic | 09:53 | |
| *** naohirot has quit IRC | 09:55 | |
| *** hj-hp has quit IRC | 09:55 | |
| *** hj-hp has joined #openstack-ironic | 09:57 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Ensure configdrive isn't mounted for ipxe/elilo https://review.openstack.org/167449 | 09:58 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add option to delete ports after introspection https://review.openstack.org/166915 | 09:59 |
| *** kalpase1 has joined #openstack-ironic | 10:03 | |
| *** pelix has joined #openstack-ironic | 10:04 | |
| *** hj-hp has quit IRC | 10:05 | |
| *** pradipta has quit IRC | 10:05 | |
| *** romcheg has joined #openstack-ironic | 10:06 | |
| *** hj-hp has joined #openstack-ironic | 10:06 | |
| *** kalpase has quit IRC | 10:07 | |
| *** yog__ has joined #openstack-ironic | 10:10 | |
| *** hj-hp has quit IRC | 10:14 | |
| *** hj-hp has joined #openstack-ironic | 10:16 | |
| *** chenglch has quit IRC | 10:17 | |
| *** hj-hp has quit IRC | 10:23 | |
| *** hj-hp has joined #openstack-ironic | 10:24 | |
| *** yog__ has quit IRC | 10:28 | |
| *** enikanorov has joined #openstack-ironic | 10:30 | |
| *** Nisha has quit IRC | 10:31 | |
| *** derekh has joined #openstack-ironic | 10:33 | |
| *** hj-hp has quit IRC | 10:33 | |
| *** enikanorov__ has quit IRC | 10:33 | |
| *** yog__ has joined #openstack-ironic | 10:33 | |
| *** hj-hp has joined #openstack-ironic | 10:33 | |
| *** hj-hp has quit IRC | 10:41 | |
| *** hj-hp has joined #openstack-ironic | 10:42 | |
| *** dtantsur is now known as dtantsur|brb | 10:47 | |
| *** hj-hp has quit IRC | 10:50 | |
| *** hj-hp has joined #openstack-ironic | 10:51 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Fallback to RIBCL https://review.openstack.org/167151 | 10:52 |
| *** sandhya_ has quit IRC | 10:58 | |
| *** hj-hp has quit IRC | 10:59 | |
| *** hj-hp has joined #openstack-ironic | 11:00 | |
| *** ramineni has quit IRC | 11:05 | |
| *** yog__ has quit IRC | 11:06 | |
| *** hj-hp has quit IRC | 11:08 | |
| *** hj-hp has joined #openstack-ironic | 11:09 | |
| rameshg87 | lucasagomes: hi | 11:13 |
| *** pelix has quit IRC | 11:13 | |
| lucasagomes | rameshg87, hi there | 11:13 |
| rameshg87 | lucasagomes: with the configdrive thing - the cirros image boots fine for me, but the ubuntu cloud image doesn't boot and still looks for the metadata contents http://169.254.169.254 | 11:14 |
| rameshg87 | lucasagomes: any quick pointers ? i built ubuntu image from diskimage-builder | 11:15 |
| lucasagomes | rameshg87, hmm thinking | 11:15 |
| rameshg87 | lucasagomes: the ubuntu image just doesn't allow me to login ssh - i think its not picking up the configdrive contents written to the disk | 11:15 |
| lucasagomes | I don't have anything on the top off my head :/ I hardly use ubuntu on tests. Unless I need to use ubuntu | 11:16 |
| rameshg87 | lucasagomes: which image do you use ? fedora cloud image ? | 11:16 |
| lucasagomes | yeah | 11:16 |
| rameshg87 | built using diskimage-builder ? | 11:16 |
| lucasagomes | yes | 11:16 |
| rameshg87 | okay, will give a try .. | 11:16 |
| rameshg87 | diskimage-create fedora baremetal | 11:16 |
| rameshg87 | that's it ? | 11:16 |
| lucasagomes | yes | 11:17 |
| rameshg87 | okay, thanks .. will give a try .. | 11:17 |
| *** hj-hp has quit IRC | 11:17 | |
| lucasagomes | I'm looking if I can find some problem on the ubuntu image | 11:17 |
| lucasagomes | or image building itself | 11:17 |
| *** hj-hp has joined #openstack-ironic | 11:18 | |
| lucasagomes | rameshg87, not sure man, perhaps someone on TripleO might be able to help you better with it? | 11:20 |
| *** pelix has joined #openstack-ironic | 11:20 | |
| lucasagomes | I mean #tripleo | 11:20 |
| rameshg87 | lucasagomes: okay | 11:20 |
| rameshg87 | lucasagomes: will update you after i try fedora :) | 11:21 |
| lucasagomes | ok :) | 11:21 |
| * rameshg87 goes home | 11:22 | |
| *** rameshg87 has left #openstack-ironic | 11:22 | |
| *** dlpartain has joined #openstack-ironic | 11:22 | |
| *** vdrok_afk is now known as vdrok | 11:23 | |
| *** hj-hp has quit IRC | 11:26 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: grub2 bootloader support for uefi boot mode https://review.openstack.org/166192 | 11:27 |
| *** hj-hp has joined #openstack-ironic | 11:27 | |
| *** dlpartain has left #openstack-ironic | 11:32 | |
| *** yog__ has joined #openstack-ironic | 11:33 | |
| *** athomas has joined #openstack-ironic | 11:33 | |
| *** hj-hp has quit IRC | 11:35 | |
| devananda | g'morning, all | 11:35 |
| *** hj-hp has joined #openstack-ironic | 11:36 | |
| devananda | so while I'm sort of AFK this week, I'm actualy doing some really fun things | 11:36 |
| devananda | like I just got to play with a BMC that uses a REST API | 11:37 |
| lucasagomes | devananda, morning | 11:40 |
| lucasagomes | BMC with rest sounds nice! Does it do async calls? | 11:41 |
| devananda | good question, i'll find out | 11:43 |
| devananda | so far, i've only done basic things like GET to discover and walk the schema | 11:44 |
| *** hj-hp has quit IRC | 11:44 | |
| lucasagomes | I see, nice! | 11:45 |
| *** hj-hp has joined #openstack-ironic | 11:45 | |
| devananda | I learned one thing that i dont think we have done in any of OpenStack's RST APIs | 11:49 |
| devananda | to avoid race conditions where >1 client tries to update the same resource, every resource has an etag, which is returned as a header any time the resource is returned | 11:49 |
| devananda | the etag id changes on every write, and any write request must have the etag included in the request | 11:50 |
| devananda | the server then can reject a PUT if the etag is stale | 11:50 |
| lucasagomes | devananda, oh, that's a good thing because right now in Ironic we get a lock on the node to update it | 11:50 |
| lucasagomes | but our locks are based on the conductor that can manage that node | 11:50 |
| devananda | yup | 11:50 |
| lucasagomes | with the driver | 11:50 |
| *** chlong has joined #openstack-ironic | 11:50 | |
| lucasagomes | BUT | 11:50 |
| devananda | there are still things that should hold hte lock -- but there really are very few of those | 11:51 |
| lucasagomes | devananda, if there's no conductor up with a driver to control that node | 11:51 |
| lucasagomes | we can't even update it | 11:51 |
| devananda | all the rest of the places we (over)use the lock today, we should use etags | 11:51 |
| lucasagomes | say, changing the driver of that node | 11:51 |
| devananda | right | 11:51 |
| lucasagomes | that would solve it | 11:51 |
| * lucasagomes writes down to look at etags | 11:52 | |
| lucasagomes | devananda, yeah locks will continue for other operations, but maybe updating information on the node might not require a conductor with a specific driver to do it | 11:52 |
| *** hj-hp has quit IRC | 11:53 | |
| *** hj-hp has joined #openstack-ironic | 11:54 | |
| rbogorodskiy | one more thing I don't quite understand. if i want to use baremetal/ironic and virt/kvm on the same environment, do I have to use nova-cells or there are some other ways? | 11:56 |
| devananda | rbogorodskiy: you can use host aggregates | 11:58 |
| devananda | you don't need to use cells (but that would work, too) | 11:58 |
| *** dprince has joined #openstack-ironic | 11:59 | |
| *** trown|outttypeww is now known as trown | 11:59 | |
| rbogorodskiy | devananda: hm... how scheduler works in this case? as far as i understand, ironic needs its own scheduler with its own host manager | 12:00 |
| rbogorodskiy | and when running two schedulers (for example) with different configuration, the first who grabs the message wins | 12:01 |
| rbogorodskiy | or I have a wrong understanding of that? | 12:01 |
| devananda | rbogorodskiy: that's what host aggregates are for. you can have two schedulers, each taking different types of requests | 12:02 |
| *** hj-hp has quit IRC | 12:02 | |
| *** hj-hp has joined #openstack-ironic | 12:03 | |
| rbogorodskiy | hm, I've found only AggregateInstanceExtraSpecsFilter filter that doesn't help with schedulers. let me google some more :) | 12:04 |
| *** athomas has quit IRC | 12:04 | |
| *** openstackgerrit has quit IRC | 12:06 | |
| *** openstackgerrit has joined #openstack-ironic | 12:07 | |
| *** pelix has quit IRC | 12:07 | |
| *** saripurigopi has quit IRC | 12:08 | |
| rbogorodskiy | devananda: am I still right that in case of using aggregates, I'd have to run two instances of schedulers, one with ironic host manager, and one with the nova host manager? | 12:09 |
| *** andreykurilin__ has quit IRC | 12:09 | |
| *** pelix has joined #openstack-ironic | 12:14 | |
| *** bradjones has quit IRC | 12:21 | |
| *** hj-hp has quit IRC | 12:21 | |
| *** bradjones has joined #openstack-ironic | 12:21 | |
| *** bradjones has quit IRC | 12:21 | |
| *** bradjones has joined #openstack-ironic | 12:21 | |
| *** lucasagomes is now known as lucas-hungry | 12:21 | |
| *** jcoufal has joined #openstack-ironic | 12:22 | |
| *** hj-hp has joined #openstack-ironic | 12:22 | |
| *** athomas has joined #openstack-ironic | 12:22 | |
| *** hj-hp has quit IRC | 12:29 | |
| *** hj-hp has joined #openstack-ironic | 12:30 | |
| devananda | rbogorodskiy: yes - you'll need a different nova scheduler for VMs and for bare metal | 12:34 |
| rbogorodskiy | devananda: hm.. these two schedulers will sit on the same message queue and the first that grabes the message will do the processing. is that understanding correct as well? | 12:40 |
| devananda | rbogorodskiy: no. AIUI they will respond only to requests for the host aggregate they are configured to be a part of | 12:41 |
| *** dprince has quit IRC | 12:42 | |
| *** TheJulia has joined #openstack-ironic | 12:44 | |
| rbogorodskiy | hm | 12:46 |
| rbogorodskiy | how does scheduler know what host aggregate it's assigned to? | 12:48 |
| *** rameshg87 has joined #openstack-ironic | 12:48 | |
| devananda | flavor | 12:50 |
| rbogorodskiy | flavor tells what kind of aggregate user wants his VM (or baremetal instance) to put in. but how scheduler knows if it's responsible for this aggregate? | 12:52 |
| *** hj-hp has quit IRC | 12:56 | |
| *** rwsu has quit IRC | 12:56 | |
| *** hj-hp has joined #openstack-ironic | 12:57 | |
| *** dprince has joined #openstack-ironic | 13:01 | |
| *** hj-hp has quit IRC | 13:06 | |
| *** hj-hp has joined #openstack-ironic | 13:06 | |
| *** jmank has joined #openstack-ironic | 13:12 | |
| *** hj-hp has quit IRC | 13:14 | |
| *** hj-hp has joined #openstack-ironic | 13:15 | |
| *** foexle has joined #openstack-ironic | 13:17 | |
| *** foexle has quit IRC | 13:17 | |
| *** foexle has joined #openstack-ironic | 13:17 | |
| *** zz_jgrimm is now known as jgrimm | 13:17 | |
| *** hj-hp has quit IRC | 13:23 | |
| *** lucas-hungry is now known as lucasagomes | 13:24 | |
| *** mdbooth has quit IRC | 13:25 | |
| *** hj-hp has joined #openstack-ironic | 13:25 | |
| *** gabriel-bezerra has joined #openstack-ironic | 13:27 | |
| *** Marga_ has joined #openstack-ironic | 13:28 | |
| *** Marga_ has quit IRC | 13:30 | |
| *** Marga_ has joined #openstack-ironic | 13:31 | |
| *** bradjones has quit IRC | 13:31 | |
| *** bradjones has joined #openstack-ironic | 13:32 | |
| *** bradjones has quit IRC | 13:32 | |
| *** bradjones has joined #openstack-ironic | 13:32 | |
| *** hj-hp has quit IRC | 13:33 | |
| *** hj-hp has joined #openstack-ironic | 13:34 | |
| *** krtaylor has quit IRC | 13:35 | |
| *** openstackgerrit has quit IRC | 13:36 | |
| *** openstackgerrit has joined #openstack-ironic | 13:36 | |
| *** hj-hp has quit IRC | 13:41 | |
| *** rameshg87 is now known as rameshg87-away | 13:41 | |
| *** krtaylor has joined #openstack-ironic | 13:41 | |
| *** hj-hp has joined #openstack-ironic | 13:42 | |
| *** r-daneel has joined #openstack-ironic | 13:45 | |
| *** igordcard_ has quit IRC | 13:49 | |
| *** hj-hp has quit IRC | 13:51 | |
| *** mtanino has joined #openstack-ironic | 13:51 | |
| *** hj-hp has joined #openstack-ironic | 13:52 | |
| *** mdbooth has joined #openstack-ironic | 13:59 | |
| *** hj-hp has quit IRC | 13:59 | |
| *** romcheg has quit IRC | 14:00 | |
| *** mgoddard1 has joined #openstack-ironic | 14:00 | |
| *** hj-hp has joined #openstack-ironic | 14:00 | |
| *** ijw has joined #openstack-ironic | 14:01 | |
| NobodyCam | Morning Ironic | 14:01 |
| *** mgoddard has quit IRC | 14:01 | |
| lucasagomes | NobodyCam, morning :D | 14:05 |
| TheJulia | good morning | 14:05 |
| *** mjturek1 has joined #openstack-ironic | 14:05 | |
| NobodyCam | morning lucasagomes and TheJulia | 14:06 |
| NobodyCam | was just reading back scroll about etags | 14:06 |
| NobodyCam | that a cool idea | 14:07 |
| NobodyCam | :) | 14:07 |
| NobodyCam | oh coffee is ready | 14:07 |
| *** uschreiber_ has joined #openstack-ironic | 14:08 | |
| *** hj-hp has quit IRC | 14:09 | |
| *** ChuckC has joined #openstack-ironic | 14:09 | |
| *** hj-hp has joined #openstack-ironic | 14:10 | |
| *** jgrimm is now known as zz_jgrimm | 14:11 | |
| *** dtantsur|brb is now known as dtantsur | 14:17 | |
| *** hj-hp has quit IRC | 14:17 | |
| dtantsur | morning NobodyCam, TheJulia, devananda | 14:18 |
| *** hj-hp has joined #openstack-ironic | 14:18 | |
| *** zz_jgrimm is now known as jgrimm | 14:20 | |
| BadCub | morning folks | 14:21 |
| devananda | morning, all | 14:21 |
| devananda | (and afternoon) | 14:21 |
| dtantsur | BadCub, morning you too :) | 14:23 |
| *** anderbubble has joined #openstack-ironic | 14:23 | |
| BadCub | morning devananda , dtantsur :-) | 14:24 |
| pshige | BadCub: morning | 14:24 |
| pshige | devananda: morning | 14:24 |
| BadCub | morning pshige | 14:24 |
| BadCub | :-) | 14:24 |
| devananda | session over, gotta move rooms, bbiab | 14:25 |
| * BadCub needs to grab coffee and smoke. brb | 14:25 | |
| *** saripurigopi has joined #openstack-ironic | 14:25 | |
| NobodyCam | morning devananda pshige and BadCub :) | 14:26 |
| pshige | NobodyCam: morning | 14:27 |
| rameshg87-away | folks what do you think about change https://review.openstack.org/#/c/167208/1/ironic/conductor/manager.py | 14:28 |
| rameshg87-away | it's correct to return at that place, though we have statement right next which will make it return back with the same return value | 14:28 |
| rameshg87-away | i mean the proposed change is correct, but without much value | 14:29 |
| *** saripurigopi has quit IRC | 14:29 | |
| rameshg87-away | do we accept the change in these cases ? | 14:29 |
| jroll | morning all :) | 14:30 |
| jroll | devananda: lucasagomes: etags almost sound like object versions ;D | 14:31 |
| pshige | jroll: morning | 14:31 |
| jroll | rameshg87-away: we could save 5 cpu cycles! | 14:32 |
| *** rameshg87-away is now known as rameshg87 | 14:32 | |
| jroll | rameshg87-away: to be fair, that loop does need to run quickly, the slower it is the more conductors you need | 14:32 |
| jroll | I'm fine with it, it won't have a huge impact but can't hurt | 14:33 |
| jroll | morning pshige :) | 14:33 |
| rameshg87 | jroll: yeah may be | 14:33 |
| *** saripurigopi has joined #openstack-ironic | 14:33 | |
| rameshg87 | jroll: but anyway cpu cycles aren't our problem there :) | 14:33 |
| jroll | rameshg87: heh, true | 14:34 |
| rameshg87 | jroll: we have plenty of cpu cycles, we will be just waiting for i/o | 14:34 |
| jroll | yeah | 14:34 |
| rameshg87 | and interpretter might be smart enough to optimize that as well | 14:34 |
| rameshg87 | anyways :) | 14:34 |
| jroll | idk about that :P | 14:34 |
| jroll | idk, I don't want to turn someone away just because we don't care about their optimization | 14:34 |
| jroll | if we tell them it's not a useful patch they might never come back | 14:35 |
| * jroll hasn't looked at who submitted it yet | 14:35 | |
| *** romcheg has joined #openstack-ironic | 14:35 | |
| *** ChuckC has quit IRC | 14:35 | |
| * jroll +2 | 14:35 | |
| rameshg87 | jroll: i didn't say that implicitly | 14:36 |
| jroll | rameshg87: no, I know | 14:36 |
| jroll | rameshg87: just looking at it from their perspective :) | 14:36 |
| BadCub | g'mornin jroll | 14:36 |
| rameshg87 | jroll: yeah, +Aed :) | 14:36 |
| lucasagomes | jroll, morning | 14:36 |
| jroll | rameshg87: my first patch to ironic was to add "apt-get install git" to the developer setup docs. who would have thought a dev wouldn't have git installed? (I was using a fresh VM) | 14:37 |
| lucasagomes | BadCub, morning :) | 14:37 |
| BadCub | g'mornin lucasagomes | 14:37 |
| BadCub | :-) | 14:37 |
| rameshg87 | jroll: :) | 14:37 |
| jroll | rameshg87: if someone -2'd that patch I would have said ironic sucks and maybe not switched from our custom thing to ironic | 14:37 |
| jroll | morning BadCub and lucasagomes :) | 14:37 |
| rameshg87 | hehe | 14:37 |
| pshige | (^^) | 14:39 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Correct update_dhcp_opts methods https://review.openstack.org/166903 | 14:40 |
| NobodyCam | morning jroll and rameshg87 :) | 14:41 |
| rameshg87 | morning NobodyCam | 14:41 |
| dtantsur | jroll, morning | 14:41 |
| *** kkoski has joined #openstack-ironic | 14:41 | |
| NobodyCam | morning dtantsur | 14:42 |
| rameshg87 | lucasagomes: i think you meant to still keep create_cleaning_ports and delete_cleaning_ports, right ? ^^^^ | 14:42 |
| dtantsur | NobodyCam, morning! | 14:42 |
| rameshg87 | lucasagomes: https://review.openstack.org/166903 | 14:42 |
| dtantsur | oops, I actually said it already :D | 14:42 |
| pshige | dtantsur: morning | 14:43 |
| jroll | heya dtantsur, NobodyCam :) | 14:43 |
| * dtantsur needs a tool to track afternoon good mornings | 14:43 | |
| dtantsur | pshige, o/ | 14:43 |
| jroll | lol | 14:43 |
| lucasagomes | rameshg87, looks | 14:43 |
| jroll | dtantsur: we need a good morning bot | 14:43 |
| BadCub | 161474 got approved by Nova. 164313 has one +2 from Nova | 14:43 |
| NobodyCam | dtantsur: lol | 14:43 |
| rameshg87 | lucasagomes: i would think we should add create_cleaning_ports and delete_cleaning_ports | 14:43 |
| lucasagomes | rameshg87, right, that's fine to add if it's a common interface | 14:43 |
| lucasagomes | the docstring was mentioning neutron | 14:44 |
| lucasagomes | so if it's not neutron specific sure let's add it | 14:44 |
| rameshg87 | yeah | 14:44 |
| jroll | NobodyCam: can you bring back pixieboots and make it say good morning to everyone? | 14:44 |
| jroll | :P | 14:44 |
| rameshg87 | Julia Kreger is in irc ? | 14:44 |
| TheJulia | I'm here | 14:44 |
| lucasagomes | TheJulia, ^ | 14:44 |
| rameshg87 | TheJulia: oh :) | 14:44 |
| jroll | rameshg87: meet TheJulia | 14:44 |
| pshige | :-p | 14:44 |
| jroll | :P | 14:44 |
| TheJulia | hi | 14:44 |
| NobodyCam | jroll: not a bad idea | 14:44 |
| rameshg87 | TheJulia: hello | 14:44 |
| lucasagomes | rameshg87, we need to be careful on the abstractmethod decorator as well | 14:44 |
| lucasagomes | since that breaks downstream plugins | 14:44 |
| BadCub | mornin rameshg87 | 14:44 |
| TheJulia | I kind of feel like they are fairly neutron specific | 14:44 |
| rameshg87 | morning BadCub | 14:44 |
| lucasagomes | the best case when adding new methods IMO, is just to make then no-op | 14:45 |
| rameshg87 | TheJulia: do you mean create_cleaning_ports ? | 14:45 |
| TheJulia | even the vifs and optional token addition is neutron specific | 14:45 |
| BadCub | morning TheJulia :-) | 14:45 |
| rameshg87 | yeah i guess the whole of DHCPApi is inclined to neutron :) | 14:45 |
| TheJulia | rameshg87: I'm generally speaking in terms of the drift of the neutron driver from the base class | 14:45 |
| TheJulia | It is | 14:45 |
| rameshg87 | we have update_port_address that shouldn't be part of the DHCPApi | 14:46 |
| TheJulia | and none is purely meant to be a "oh, okay, pass" framework | 14:46 |
| NobodyCam | in general what are folks thoughts on patches like: https://review.openstack.org/#/c/167415 | 14:46 |
| rameshg87 | and it is | 14:46 |
| rameshg87 | update_port_address isn't called from anywhere outside | 14:46 |
| rameshg87 | but create_cleaning_ports which is used by ironic isn't part of the DHCPApi :( | 14:46 |
| jroll | rameshg87: TheJulia: to be clear, this API is meant such that someone could use another thing, like maybe a client to dynamically update isc-dhcp-server | 14:47 |
| jroll | rameshg87: TheJulia: it isn't limited to no-op or neutron | 14:47 |
| jroll | rameshg87: TheJulia: I tend to think everything should be part of the base class | 14:47 |
| *** Marga_ has quit IRC | 14:47 | |
| rameshg87 | yeah, jroll i agree | 14:48 |
| jroll | NobodyCam: heh, interesting. I think I'd rather a vagrantfile in the bifrost repo :P | 14:48 |
| TheJulia | agreed, and thats what I was thinking as well, I can put it back, but I think the doc strings should indicate why they exist, which means they will indicate their original use of neutron | 14:48 |
| jroll | rameshg87: yeah, presumably if you used a neutron competitor it would have a concept of auth tokens and ports | 14:48 |
| rameshg87 | TheJulia: yeah, unfortunately yes :( | 14:48 |
| rameshg87 | jroll: heh, but DHCPApi shouldn't be assuming we would always have a neutron competitor there :( | 14:49 |
| rameshg87 | it could be someone who was never interested in competing with neutron :) | 14:49 |
| jroll | rameshg87: well sure, competitor is a bad word | 14:49 |
| *** subscope has joined #openstack-ironic | 14:50 | |
| rameshg87 | jroll: yeah but i still feel when we created a none provider, we should have designed the api to be agonostic of neutron | 14:50 |
| TheJulia | the case could still be someone who roles their own network driver that still leverages neutron. The good thing is the cleaning code checks and only calls it if it is present | 14:50 |
| * rameshg87 feels somewhere it isn't | 14:50 | |
| *** kalpase1 has left #openstack-ironic | 14:51 | |
| jroll | rameshg87: indeed... idk. maybe just bad names. I think 'token' is fine. 'vifs' could probably be 'ports' | 14:51 |
| TheJulia | It seems like it was more "Oh, we're totally depenent upon neutron, lets make so people don't HAVE to use neutron" | 14:51 |
| jroll | rameshg87: I also tend to think this should be a more general NetworkAPI | 14:52 |
| rameshg87 | yeah | 14:52 |
| jroll | TheJulia: that was initially the goal... but I'd love to have a dhcp provider that isn't neutron | 14:52 |
| rameshg87 | TheJulia: yeah | 14:52 |
| TheJulia | I agree that it should be labeled ports, but intead of vifs. Just means I'll need to go update things elsewhere to make it consistent | 14:53 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Expose node id to users https://review.openstack.org/167623 | 14:53 |
| *** kalpase has joined #openstack-ironic | 14:53 | |
| *** kalpase has left #openstack-ironic | 14:53 | |
| TheJulia | jroll: too many things on my plate at the moment, although I suspect if bifrost gets any adoption, need and momentum for additional dhcp drivers will grow. | 14:53 |
| jroll | TheJulia: indeed. | 14:54 |
| lucasagomes | one idea on the vifs/token= is to have the base to pass kwargs | 14:54 |
| lucasagomes | so each driver can pass whatever kwarg it wants/expects | 14:54 |
| lucasagomes | or none | 14:54 |
| jroll | oh true | 14:54 |
| rameshg87 | lucasagomes: but is it good for an api to do that ? :) | 14:54 |
| jroll | that would let us keep compatibility if we want to s/vifs/ports/ too | 14:54 |
| lucasagomes | we don't need to expect a argument specifically | 14:54 |
| lucasagomes | rameshg87, well for those cases, I don't see why not | 14:54 |
| *** ChuckC has joined #openstack-ironic | 14:55 | |
| rameshg87 | :) | 14:55 |
| lucasagomes | rameshg87, the base class, has to be agnostic of implementations | 14:55 |
| rameshg87 | lucasagomes: yeah i agree .. | 14:55 |
| rameshg87 | lucasagomes: but just wondering since we are passing task, shouldn't the implementations have enough thing to figure out the rest ? i mean token could be task.context.auth_token | 14:56 |
| rameshg87 | lucasagomes: wondering if we needed to pass token there really | 14:56 |
| *** kalpase1 has joined #openstack-ironic | 14:56 | |
| rameshg87 | or was it admin token ? :) | 14:56 |
| lucasagomes | right, but and the vifs? | 14:56 |
| lucasagomes | it's being passed to neutron to work right? | 14:57 |
| lucasagomes | but I doubt it's needed for jroll's isc-dhcp-server | 14:57 |
| rameshg87 | may be i might be saying something stupid here | 14:57 |
| rameshg87 | lucasagomes: but i think it should have been other way round | 14:57 |
| rameshg87 | lucasagomes: we pass in the macs to be updated | 14:58 |
| lucasagomes | oh | 14:58 |
| rameshg87 | lucasagomes: and we figure out the vifs from the ironic ports | 14:58 |
| jroll | lucasagomes: to be clear, I don't have any plans to make that, but it would be nice :) | 14:58 |
| rameshg87 | vifs are specific to neutron | 14:58 |
| lucasagomes | rameshg87, that's also true | 14:58 |
| lucasagomes | rameshg87, yeah I agree with you here | 14:58 |
| rameshg87 | lucasagomes: because if we had another real dhcp provider, vifs might not make sense to them | 14:58 |
| lucasagomes | the vifs are on the ports/extra | 14:58 |
| lucasagomes | because we put it there after updating it to neutron | 14:58 |
| rameshg87 | yeah | 14:59 |
| lucasagomes | so neutron dhcp provider can look at that | 14:59 |
| openstackgerrit | Zhenguo Niu proposed openstack/python-ironicclient: Expose node id to users https://review.openstack.org/167627 | 14:59 |
| lucasagomes | jroll, TheJulia ^ I agree | 14:59 |
| lucasagomes | I have a call in 1 min, brb | 14:59 |
| rameshg87 | lucasagomes: but that said it's all released abstract methods | 14:59 |
| lucasagomes | it's quick I hope | 14:59 |
| dtantsur | devananda, could you reconsider -2 on https://review.openstack.org/#/c/165860/ ? | 14:59 |
| rameshg87 | lucasagomes: so can't do anything i hope | 14:59 |
| lucasagomes | rameshg87, right well if we figure out the vifs looking at the ports | 14:59 |
| jroll | something just occurred to me -- we can't use kwargs because the things calling the API have to know what to pass in | 14:59 |
| * BadCub has to run off to several conference calls this morning *groans* | 15:00 | |
| * lucasagomes jumps on the call | 15:00 | |
| * TheJulia was waiting for that realization :) | 15:01 | |
| * TheJulia gets more coffee | 15:03 | |
| lucasagomes | dtantsur, it's only me or nobody is talking on that call? | 15:04 |
| dtantsur | lucasagomes, nobody :) | 15:04 |
| lucasagomes | ack | 15:04 |
| * rameshg87 runs for dinner | 15:06 | |
| *** rameshg87 is now known as rameshg87-dinner | 15:06 | |
| NobodyCam | have a good dinner rameshg87-dinner | 15:06 |
| TheJulia | Dinner sounds really good... although it is still morning here. | 15:07 |
| NobodyCam | :-p | 15:07 |
| pshige | :-P | 15:07 |
| NobodyCam | I heard of breakfast for dinner but never the other way around | 15:07 |
| * devananda returns | 15:08 | |
| *** romcheg has quit IRC | 15:08 | |
| *** romcheg1 has joined #openstack-ironic | 15:08 | |
| *** openstackgerrit has quit IRC | 15:08 | |
| devananda | jroll: but only latest || !latest | 15:09 |
| *** uschreiber_ has quit IRC | 15:09 | |
| *** openstackgerrit has joined #openstack-ironic | 15:09 | |
| NobodyCam | wb devananda | 15:09 |
| jroll | devananda: it's still an identifier | 15:09 |
| devananda | jroll: yup | 15:09 |
| jroll | devananda: you can either generate a random string or bump an integer | 15:10 |
| jroll | the latter can be useful to humans :) | 15:10 |
| jroll | also be useful* | 15:10 |
| devananda | lucasagomes: @abstract is a tool for us to define the minimum required methods any interface must implement | 15:11 |
| devananda | lucasagomes: so, eg, I could see ManagemtnInterface defining set-boot-device as @abstract | 15:11 |
| lucasagomes | devananda, yup, yeah. But we need to be a bit careful with that | 15:11 |
| lucasagomes | by adding a new abstractmethod to an interface | 15:11 |
| devananda | lucasagomes: and yes, we need to be deliberate if/when we promote any interface.method to required | 15:11 |
| lucasagomes | breaks the interfaces that inherent from it and not have that method defined | 15:12 |
| lucasagomes | especially downstream interfaces | 15:12 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Deploys post whole disk image deploy fails https://review.openstack.org/167636 | 15:12 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add INSPECTFAIL as a valid state to start introspection https://review.openstack.org/167640 | 15:15 |
| *** kalpase1 has left #openstack-ironic | 15:17 | |
| *** romcheg has joined #openstack-ironic | 15:18 | |
| *** romcheg1 has quit IRC | 15:18 | |
| openstackgerrit | Merged openstack/ironic: Add a return after saving node power state https://review.openstack.org/167208 | 15:22 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add option to delete ports after introspection https://review.openstack.org/166915 | 15:22 |
| *** anderbubble has quit IRC | 15:26 | |
| devananda | JayF, lucasagomes: just checking, know if anyone on your teams is proposing a talk to linuxCon or cloudOpen? | 15:27 |
| devananda | russell_h: ^ | 15:27 |
| lucasagomes | devananda, not sure, linuxcon america? | 15:29 |
| devananda | lucasagomes: either one | 15:30 |
| devananda | I think linuxcon EU is in dublin in the fall? | 15:30 |
| lucasagomes | devananda, yes! | 15:30 |
| lucasagomes | the next linuxcon europe :) | 15:30 |
| * lucasagomes will def go! | 15:30 | |
| devananda | awesome | 15:30 |
| devananda | lucasagomes: you should totally talk about ironic there :) | 15:31 |
| devananda | save me the trip ;) | 15:31 |
| lucasagomes | hah hmm that's not a bad idea actually | 15:31 |
| lucasagomes | I could def take a look at it | 15:31 |
| * lucasagomes writes it down | 15:31 | |
| jlvillal | lucasagomes: And OSCON is in Amsterdam this year :) | 15:31 |
| * jlvillal likes visiting Amsterdam. Thinks it would be a nice place to live. | 15:32 | |
| lucasagomes | devananda, I leave literally 2 blocks away of where the conference is going to be held here in dublin | 15:32 |
| * NobodyCam thinks nothing will get done at OSCON this year | 15:32 | |
| lucasagomes | live* | 15:32 |
| lucasagomes | jlvillal, o/ | 15:33 |
| * jlvillal I'd like to go to PyCon, but don't think my boss will approve me for that. Montreal in about two weeks | 15:33 | |
| devananda | lucasagomes: hah! awesome | 15:35 |
| *** Sukhdev has joined #openstack-ironic | 15:38 | |
| jlvillal | lucasagomes: For some reason I was thinking you lived in Portugal | 15:42 |
| jlvillal | :) | 15:42 |
| * BadCub crosses out all the merged bug fixes on pad | 15:43 | |
| russell_h | devananda: we were just talking about that yesterday | 15:44 |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic: Added vagrant VM for developer use https://review.openstack.org/167415 | 15:44 |
| *** yog__ has quit IRC | 15:44 | |
| *** stendulker has joined #openstack-ironic | 15:44 | |
| russell_h | devananda: I think we'll propose something to at least one of LinuxCon or CloudOpen but can't promise it | 15:45 |
| russell_h | our travel situation is unusually screwed up | 15:45 |
| devananda | russell_h: :-/ | 15:46 |
| *** dlpartain has joined #openstack-ironic | 15:47 | |
| *** trown is now known as trown|lunch | 15:47 | |
| *** dlpartain has left #openstack-ironic | 15:47 | |
| lucasagomes | jlvillal, my mother language is portuguese :) | 15:48 |
| lucasagomes | I was born and raised in brazil | 15:49 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add option to delete ports after introspection https://review.openstack.org/166915 | 15:50 |
| lucasagomes | devananda, while you here, can I get ur opnion on where the config drive generation should live? | 15:50 |
| jlvillal | lucasagomes: Ah :) We have a few Brazilians working here. I was born in Argentina myself. But was only there until about 3 years old. No hablo espanol :( | 15:50 |
| lucasagomes | devananda, under tools/contrib in the Ironic tree, or it should it be part of the ironic client | 15:50 |
| devananda | lucasagomes: client would make it more easily reusable. ironic/contrib is fine with me, too | 15:51 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Deploys post whole disk image deploy fails https://review.openstack.org/167636 | 15:51 |
| lucasagomes | jlvillal, oh nice! My dad is spanish (and mom brazilian) but my spanish is pretty dodge :) | 15:51 |
| devananda | since the client can be expected to pass a config drive to the server, it seems reasonable to me that the clien tshould have a way to generate it | 15:51 |
| lucasagomes | devananda, right, fair enuff... yeah it sounds easier indeed | 15:52 |
| devananda | lucasagomes: isn't there code in nova already to generate configdrives? | 15:52 |
| lucasagomes | devananda, I was just wondering about the dependency on the makeisofs | 15:52 |
| lucasagomes | devananda, there is yes | 15:52 |
| devananda | lucasagomes: perhaps the right answer is to librarify it, then consume that in a CLI tool (or in our CLI) | 15:52 |
| jlvillal | lucasagomes: My dad is Mexican-American (grand-parents from Mexico) and my mother is from Birmingham, England. My father was a US diplomat in Argentina when I was born. Unfortunately I only know English and a few words of greeting in some other languages. | 15:53 |
| *** rwsu has joined #openstack-ironic | 15:53 | |
| lucasagomes | devananda, indeed it would be good to share some code | 15:53 |
| lucasagomes | oslo seems the right place | 15:53 |
| lucasagomes | jlvillal, gotcha, that's pretty cool! | 15:54 |
| JayF | devananda: we talked about it yesterday, idk for sure | 15:54 |
| JayF | russell_h: ^ deva was asking if we were going to submit something for cloudopen/linuxcon | 15:54 |
| lucasagomes | jlvillal, I've been to argentina a couple of times, mostly buenos aires. But been to the south too bariloche and so on | 15:54 |
| lucasagomes | in the patagonia | 15:54 |
| NobodyCam | morning JayF | 15:54 |
| jroll | JayF: read more scrollback :P | 15:55 |
| jlvillal | lucasagomes: I want to go someday :) | 15:55 |
| lucasagomes | jlvillal, it's beautiful down there | 15:55 |
| lucasagomes | jlvillal, yeah def do it | 15:55 |
| JayF | jroll: heh | 15:55 |
| lucasagomes | folks I was thinking, would a --json parameter for the client be useful? That would simple print the response body json instead of formatting the data as a table etc | 16:07 |
| lucasagomes | so we could consume the return as json (like using jq for bash) | 16:07 |
| *** jmccrory has joined #openstack-ironic | 16:07 | |
| lucasagomes | instead of having to do a bunch of grep, awk, cut etc... | 16:07 |
| lucasagomes | with --json the client would be like a curl with a nicer syntax :) | 16:07 |
| lucasagomes | makes sense? | 16:07 |
| jroll | see at that point I just use curl | 16:09 |
| jlvillal | lucasagomes: Makes sense to me. | 16:09 |
| jroll | (not that I don't see your point) | 16:10 |
| lucasagomes | jroll, heh yeah that's why I was asking | 16:10 |
| lucasagomes | the client kinda facilitate on the syntax but doesn't add any benefit that curl can't give you | 16:10 |
| *** romcheg has quit IRC | 16:11 | |
| jlvillal | lucasagomes: What would be cool on the client is a bash-completion script :) | 16:11 |
| * lucasagomes thought we had it | 16:12 | |
| *** romcheg has joined #openstack-ironic | 16:12 | |
| jroll | I thought we did too | 16:12 |
| NobodyCam | there was a patch for it | 16:12 |
| *** Marga_ has joined #openstack-ironic | 16:12 | |
| jlvillal | Oh, cool :) | 16:13 |
| jlvillal | Then it makes it easier than curl. <tab><tab> | 16:13 |
| lucasagomes | https://review.openstack.org/#/c/101765/ | 16:13 |
| lucasagomes | tho I'm trying here and it seems to not be working anymore | 16:13 |
| lucasagomes | hmmmmm | 16:13 |
| jlvillal | I see it in the tools directory. Now that I look. | 16:14 |
| *** romcheg has quit IRC | 16:14 | |
| lucasagomes | oh it does but you gotta source a file before | 16:16 |
| lucasagomes | source tools/ironic.bash_completion | 16:16 |
| tiagogomes_ | Hi, on pxe_ssh, shouldn't I tell the name of the VM somewhere? | 16:17 |
| *** mgoddard1 has quit IRC | 16:17 | |
| NobodyCam | tiagogomes_: what back end | 16:18 |
| tiagogomes_ | the driver is pxe_ssh, the virt_type is virsh | 16:19 |
| NobodyCam | with virsh you don't need to set the vm name, thats for windows and virtual box | 16:19 |
| NobodyCam | the driver should find the node based on the mac addres | 16:20 |
| tiagogomes_ | ah ok | 16:20 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Add check for usable tftp directory to PXE driver https://review.openstack.org/167451 | 16:21 |
| *** chlong has quit IRC | 16:22 | |
| *** rameshg87-dinner is now known as rameshg87 | 16:23 | |
| NobodyCam | wb rameshg87 | 16:23 |
| rameshg87 | thanks NobodyCam | 16:23 |
| rameshg87 | :) | 16:23 |
| NobodyCam | :) | 16:23 |
| *** ukalifon has quit IRC | 16:26 | |
| * NobodyCam git's moar koffi | 16:27 | |
| dtantsur | lucasagomes, I've been dreaming about --json for a long time... | 16:28 |
| *** davide has joined #openstack-ironic | 16:31 | |
| *** ijw has quit IRC | 16:31 | |
| *** anderbubble has joined #openstack-ironic | 16:32 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Add check for usable tftp directory to PXE driver https://review.openstack.org/167451 | 16:32 |
| *** chlong has joined #openstack-ironic | 16:33 | |
| devananda | lucasagomes: --json ++ | 16:34 |
| devananda | lucasagomes: I use curl for this all the time | 16:34 |
| lucasagomes | hah ok | 16:35 |
| lucasagomes | I will open a wishlist bug about it | 16:35 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Ensure configdrive isn't mounted for ipxe/elilo https://review.openstack.org/167449 | 16:36 |
| rameshg87 | JayF: ^^^ | 16:36 |
| rameshg87 | just fixed the elilo syntax thing | 16:36 |
| JayF | thanks rameshg87, what was the deal with the other comment you made? | 16:36 |
| rameshg87 | JayF: the double quotes wasn't moved | 16:37 |
| rameshg87 | JayF: otherwise it would lead to syntax error while elilo tries to parse it | 16:37 |
| JayF | "We should do the same for iscsi_ilo and agent_ilo driver as well. For now, they blindly pass all the options provided by iscsi_deploy.build_deploy_ramdisk_options()." | 16:37 |
| JayF | ah, you mean like the params they send over | 16:37 |
| JayF | with the vmedia | 16:37 |
| rameshg87 | JayF: yeah, the iscsi_ilo and agent_ilo | 16:37 |
| rameshg87 | yeah exactly | 16:37 |
| JayF | That's a really good call | 16:37 |
| JayF | You wanna add it to that patch or do a followp? | 16:38 |
| rameshg87 | i will fix up in next patch | 16:38 |
| JayF | okay cool | 16:38 |
| rameshg87 | i will do a followup | 16:38 |
| rameshg87 | just fixed the elilo thing on your patch | 16:38 |
| JayF | https://review.openstack.org/#/c/167449/ <-- this has been through a few rounds of back and forth, ready to land if someone core who hasn't pushed a patchset to it wants to look | 16:38 |
| JayF | jroll: ^ | 16:38 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Invalidate cached node information after hooks run https://review.openstack.org/167685 | 16:39 |
| NobodyCam | brb | 16:39 |
| *** chlong has quit IRC | 16:39 | |
| lucasagomes | devananda, dtantsur another cool idea for our api | 16:40 |
| *** romcheg has joined #openstack-ironic | 16:40 | |
| lucasagomes | is if it was possible to specify what fields I want it to return | 16:40 |
| lucasagomes | like GET /v1/nodes/<uuid>?fields=uuid,driver_info | 16:41 |
| lucasagomes | something like that | 16:41 |
| dtantsur | lucasagomes, oh yeah, python-openstackclient is capable of that (on CLI level) | 16:41 |
| dtantsur | (though maybe only one field at once, not sure) | 16:41 |
| lucasagomes | dtantsur, right, but we could do right on the API and not waste badwidth returning something we don't care about | 16:41 |
| lucasagomes | we have /detail right now in the resources | 16:42 |
| lucasagomes | to retuning everything, but being able to specify which field seems powerful | 16:42 |
| lucasagomes | anyhoo, food for thought :) | 16:42 |
| stendulker | NobodyCam: Hi | 16:44 |
| devananda | lucasagomes: one more thing -- the API schema should be discoverable | 16:44 |
| devananda | lucasagomes: eg, ishould be able to GET /v1/schema | 16:44 |
| lucasagomes | devananda, oh yeah that's a good one | 16:45 |
| devananda | and have it return the prototypical structures and all the resource links | 16:45 |
| * devananda learned a lot today, playing with hardware that has a REST API in the BMC | 16:45 | |
| lucasagomes | :D | 16:45 |
| *** saripurigopi has quit IRC | 16:46 | |
| stendulker | devananda: Hi | 16:46 |
| lucasagomes | next big thing would be to support webhooks | 16:46 |
| *** mgoddard has joined #openstack-ironic | 16:46 | |
| stendulker | devananda: This is regarding defect https://bugs.launchpad.net/ironic/+bug/1432498 "Whole disk image feature is broken in UEFI environment for pxe drivers" | 16:47 |
| openstack | Launchpad bug 1432498 in Ironic "Whole disk image feature is broken in UEFI environment for pxe drivers" [High,Confirmed] - Assigned to Shivanand Tendulker (shivanand-tendulker) | 16:47 |
| devananda | stendulker: hi | 16:47 |
| stendulker | devananda: syslinux do not support chainloading https://wiki.archlinux.org/index.php/syslinux#Limitations_of_UEFI_Syslinux | 16:47 |
| devananda | stendulker: i have a question for you, too, or for anusha. When will you enale the RIS/REST interface in proliantutils? | 16:48 |
| stendulker | devananda: elilo bootloader do not support chain loading https://wiki.archlinux.org/index.php/Boot_loaders_%28%E7%AE%80%E4%BD%93%E4%B8%AD%E6%96%87%29 | 16:48 |
| stendulker | devananda: we are enabling in kilo | 16:48 |
| stendulker | stendulker: Gen9 servers some of the calls would be using RIS. e.g. setting secure boot | 16:49 |
| devananda | stendulker: so in the next two weeks? | 16:49 |
| stendulker | devananda: yes, its in final stages of test | 16:49 |
| * lucasagomes remembers a big workaround to boot windows from linux | 16:49 | |
| stendulker | devananda: so how do we handle whole disk image support in uefi for pxe drivers? | 16:50 |
| stendulker | devananda: On proliant if we fix the syntax erros it would fall to disk (next in the boot order) and deployed image would get booted | 16:51 |
| stendulker | devananda: But that would not be a real pxe boot. | 16:51 |
| devananda | ah | 16:51 |
| rameshg87 | jlvillal: thanks for that - https://review.openstack.org/#/c/167317/2//COMMIT_MSG | 16:51 |
| stendulker | devananda: I think it would be better to make disk as the boot device for whole disk image for pxe drivers | 16:51 |
| devananda | stendulker: so if I understand, you can deploy a whole disk image and then boot it in UEFI mode | 16:51 |
| jroll | can whole disk image be done for pxe? | 16:52 |
| rameshg87 | jlvillal: but i think tralining space on commit msg should be fine | 16:52 |
| devananda | stendulker: but you can not deploy a whole disk image and then net-boot it in UEFI mode? | 16:52 |
| rameshg87 | jlvillal: because i think we never get a chance to edit the commit msg again :) | 16:52 |
| rameshg87 | jlvillal: but i am not sure | 16:52 |
| jlvillal | rameshg87: Okay :) | 16:52 |
| stendulker | devaanda: yes | 16:52 |
| devananda | jroll: you mean "for iscsi-based deployment" - and i think they added this | 16:52 |
| rameshg87 | jlvillal: i edited it from the gerrit review | 16:52 |
| jlvillal | rameshg87: I just said nit and gave a +1. | 16:52 |
| stendulker | devananda: it would be disk boot | 16:52 |
| rameshg87 | jlvillal: i think that's how it came :) | 16:52 |
| devananda | stendulker: in that case, what is functionally different between pxe_ilo and iscsi_ilo driver? | 16:53 |
| lucasagomes | jroll, yeah it was added | 16:53 |
| jroll | devananda: let me rephrase, how does pxe booting a whole disk image work :| | 16:53 |
| stendulker | jroll: whole disk image should be possible if bootloader supports chainloading | 16:53 |
| jroll | oh, right | 16:53 |
| jroll | that's silly | 16:53 |
| jroll | (imo) | 16:53 |
| stendulker | devananda: if we make disk boot, both would be same | 16:54 |
| stendulker | except for deploy being done through cdrom in case of iscsi_ilo and by pxe in pxe_ilo | 16:54 |
| devananda | if you deploy a whole disk image tht has no boot sector ... then you can essentially prevent boot from disk | 16:54 |
| NobodyCam | mornign jlvillal :) | 16:54 |
| stendulker | devnanda: then its partition image right? | 16:55 |
| jlvillal | NobodyCam: Morning! :) | 16:55 |
| devananda | stendulker: oh, good point. pxe_ilo works with proliant hardware that doesn't have an advanced license then, whereas iscsi_ilo does not? | 16:55 |
| stendulker | devananda: yes | 16:55 |
| devananda | stendulker: image structure and image content are not the same thing. but we digress... | 16:55 |
| stendulker | devananda: What do you suggest? | 16:57 |
| devananda | besides fixing it? do you know why it's not working? | 16:57 |
| *** jerryz has quit IRC | 16:58 | |
| *** jistr has quit IRC | 16:58 | |
| stendulker | primarily there is syntax errors in the config file that gets generated. | 16:59 |
| stendulker | then it tries to use chain.c32 which is valid only for BIOS | 16:59 |
| devananda | stendulker: oh, right. UEFI support for chain loading is handled differently | 16:59 |
| stendulker | if we were to use syslinux for UEFI, it does not support chainloding | 17:00 |
| stendulker | thirdly elilo do not support chainloading itself | 17:00 |
| stendulker | grub2 does have support for chainloading, but that is a different investigation | 17:00 |
| devananda | stendulker: grub2 is also local and has to be invoked by UEFI, so really, we'd be unnecessarily layering a net boot on top of grub2 bootloader on top of UEFI. | 17:02 |
| devananda | *has to be invoked by UEFI, in the case where you're booting via UEFI | 17:02 |
| tiagogomes_ | on the ssh_agent driver, when I run `validate-node` I get this error: "failed to validate deploy image info". Any clue about what is missing? | 17:02 |
| devananda | stendulker: so if I understand correctly, UEFI actually has some support for chainloading natively, but we dont have a good mechanism to pass that in remotely right now | 17:03 |
| stendulker | devananda: yes | 17:03 |
| devananda | stendulker: you can provide support for UEFI boot of whole disk images (eg, for windows deployment) via iscsi_ilo driver. is there a need for the pxe_ilo driver to support this? | 17:03 |
| stendulker | devananda: i think this feature never came from ilo driver team. It was implemeneted as a independent feature. Also this bug would apply to all pxe drivers. | 17:05 |
| devananda | stendulker: it seems like a poor interaction between UEFI support, whole-disk-imaging, and PXE drivers (ie, requiring the instance to net-boot after provisioning) | 17:06 |
| *** ijw has joined #openstack-ironic | 17:06 | |
| devananda | *and using a legacy network boot loader (eg, not UEFI's own network boot program) | 17:07 |
| stendulker | devananda: i feel we should disable whole disk image support for pxe drivers in uefi boot mode | 17:07 |
| devananda | stendulker: I'd suggest we add some error checking to the PXE driver and error early if this combination was requested | 17:07 |
| devananda | stendulker: :) | 17:07 |
| stendulker | devananda: ok. I will investigate and fix on these lines. | 17:08 |
| devananda | stendulker: thanks much | 17:08 |
| stendulker | devananda: thank you :) | 17:08 |
| * devananda copies this chat to the bug | 17:08 | |
| devananda | gotta move, bbiab | 17:11 |
| * BadCub finishes round of morning calls | 17:13 | |
| rameshg87 | tiagogomes_: have you specified an image_source ? | 17:14 |
| *** Nisha has joined #openstack-ironic | 17:16 | |
| tiagogomes_ | rameshg87, yes | 17:16 |
| rameshg87 | tiagogomes_: oh it's deploy image info | 17:17 |
| rameshg87 | tiagogomes_: have you mentioned deploy_kernel deploy_ramdisk ? | 17:17 |
| rameshg87 | tiagogomes_: may be copy-pasting full error on paste.openstack.org might help | 17:17 |
| rameshg87 | i mean help me to understand :) | 17:17 |
| *** harlowja_away is now known as harlowja | 17:18 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Ensure configdrive isn't mounted for ilo drivers https://review.openstack.org/167700 | 17:18 |
| rameshg87 | JayF: ^^^ | 17:18 |
| devananda | stendulker: for the RIS library in proliantutils, is this just using the redfish api - or is it something else? | 17:20 |
| tiagogomes_ | rameshg87, http://paste.openstack.org/show/196701/ | 17:20 |
| stendulker | devananda: it is based on redfish | 17:20 |
| *** Nisha has quit IRC | 17:21 | |
| *** Nisha_away has joined #openstack-ironic | 17:21 | |
| rameshg87 | tiagogomes_: oh some part of the error message is actually missing | 17:22 |
| *** ndipanov has quit IRC | 17:23 | |
| rameshg87 | tiagogomes_: it should have been "Node <> failed to validate deploy image info. Some parameters were missing: ......." | 17:23 |
| rameshg87 | tiagogomes_: do you not see the full error in node-validate ? | 17:23 |
| *** Sukhdev has quit IRC | 17:23 | |
| tiagogomes_ | rameshg87, no | 17:23 |
| rameshg87 | tiagogomes_: the issue is pxe_deploy_ramdisk should be just deploy_ramdisk | 17:24 |
| rameshg87 | tiagogomes_: ditto for pxe_deploy_kernel. it should be deploy_kernel | 17:24 |
| *** absubram has joined #openstack-ironic | 17:24 | |
| rameshg87 | tiagogomes_: if you don't see full msg, can you pls consider raising a vug | 17:25 |
| rameshg87 | *bug | 17:25 |
| rameshg87 | tiagogomes_: i guess it doesn't throw anything valuable enough to operator :( | 17:25 |
| tiagogomes_ | rameshg87, that worked, thanks | 17:25 |
| rameshg87 | tiagogomes_: pls consider raising a bug if you don't see full msg | 17:26 |
| rameshg87 | tiagogomes_: i think it should be fixed | 17:26 |
| tiagogomes_ | rameshg87, ok | 17:27 |
| rameshg87 | devananda: can you please decide the fate of https://review.openstack.org/#/c/165788/ for k3 | 17:29 |
| rameshg87 | devananda: even if they address rloo's comment, can it go in (it has a new string) | 17:29 |
| *** trown|lunch is now known as trown | 17:31 | |
| *** Marga_ has quit IRC | 17:32 | |
| *** achanda has joined #openstack-ironic | 17:32 | |
| * rameshg87 goes to sleep | 17:34 | |
| rameshg87 | good night ironic | 17:34 |
| *** rameshg87 has quit IRC | 17:34 | |
| NobodyCam | night ramesh87 | 17:36 |
| *** achanda has quit IRC | 17:43 | |
| *** achanda has joined #openstack-ironic | 17:51 | |
| NobodyCam | anyine gots a minute to look at https://review.openstack.org/#/c/166903 | 17:54 |
| NobodyCam | anyone even :-p | 17:54 |
| NobodyCam | TheJulia: great catch with that btw | 17:54 |
| *** tiagogomes_ has quit IRC | 17:55 | |
| *** romcheg has quit IRC | 17:55 | |
| *** derekh has quit IRC | 17:56 | |
| *** korekhov has joined #openstack-ironic | 17:56 | |
| stendulker | NobodyCam: Hi | 17:57 |
| NobodyCam | hi stendulker :) | 17:58 |
| TheJulia | NobodyCam: ty, although I don't think we reached consensus on if vifs should get renamed :) | 17:58 |
| clif_h | is checking for the agent_url in driver_info/driver_internal_info enough to determine if there's an agent running on a node? | 17:58 |
| stendulker | NobodyCam: Something to confess. I had given you incorrect information about changing the CDROM to persistent device in iscsi_ilo drivers' prepare method during secure boot review | 17:59 |
| NobodyCam | TheJulia: but thats just "a" name could easliy be changed with a follow up | 17:59 |
| TheJulia | true :) | 17:59 |
| *** Marga_ has joined #openstack-ironic | 17:59 | |
| *** vdrok is now known as vdrok_afk | 17:59 | |
| NobodyCam | if I understand we is broke now | 17:59 |
| jroll | clif_h: idk if we clear that out | 17:59 |
| stendulker | NobodyCam: secure boot does retain one time boot settings even when it reboots twice inetrnally | 17:59 |
| NobodyCam | ahh | 18:00 |
| *** korekhov_ has joined #openstack-ironic | 18:00 | |
| jroll | clif_h: look at what the cleaning stuff is, I'd just keep it consistent with that | 18:00 |
| TheJulia | NobodyCam: correct, standalone w/o neutron will not deploy right now without that patch | 18:00 |
| stendulker | NobodyCam: The virtual media has a peculiar behavior in UEFI boot mode that is breaking iscsi_ilo driver if its set as a persistent boot device in prepare() | 18:01 |
| NobodyCam | stendulker: I saw chatter about bumpping that support to L | 18:01 |
| JoshNang | clif_h: agent_url isn't enough, you could have a node that was on the agent driver, switches to pxe driver, and i don't think agent_url would get removed. last heartbeat *might* be enough? | 18:01 |
| stendulker | NobodyCam: Not grub2 thing... | 18:01 |
| NobodyCam | TheJulia: so I'd rather land as vif and rename later after we have a better name :-p | 18:01 |
| clif_h | jroll: I'm looking at that patch in ironic, but I don't see anything trying to determine whether the agent is there or not | 18:01 |
| stendulker | NobodyCam: Please have a look at this defect fix https://review.openstack.org/#/c/167636/ | 18:02 |
| NobodyCam | stendulker: oh sorry to many things moving atm :-p | 18:02 |
| TheJulia | I was about to say perhaps we should start a queue for NobodyCam's attention | 18:02 |
| jroll | clif_h: I think it assumes it is not there and boots it, JoshNang ? | 18:02 |
| clif_h | so image caching should assume that as well? | 18:02 |
| JoshNang | clif_h: its pretty much working on the agent driver, as i couldn't figure out anyway to determine iscsi on agent and we can revist with bootinterface | 18:02 |
| NobodyCam | stendulker: will do, but going pop out for a quick smoke ... will look after that | 18:02 |
| stendulker | NobodyCam: np :) Have a look at this review when you get time. | 18:02 |
| jroll | clif_h: I'm not sure :| | 18:03 |
| stendulker | NobodyCam: Thank you :) | 18:03 |
| *** korekhov has quit IRC | 18:03 | |
| NobodyCam | :) | 18:03 |
| NobodyCam | TheJulia: I need the help | 18:03 |
| NobodyCam | lol, the collender I call a mind is not working like it use to | 18:03 |
| NobodyCam | lol | 18:03 |
| clif_h | ok, I'll just assume the agent isn't there I guess | 18:03 |
| TheJulia | NobodyCam: you need one of those "Now serving #" signs | 18:04 |
| NobodyCam | lol | 18:04 |
| clif_h | I just thought there'd be some relatively easy way to check and avoid putting the agent there if its already present | 18:04 |
| BadCub | lol | 18:04 |
| NobodyCam | oh that would be cool | 18:04 |
| NobodyCam | I could reply with "take a number" | 18:04 |
| NobodyCam | lol | 18:04 |
| jroll | clif_h: ironic upstream basically doesn't have a way to keep the agent running... so it's always assumed to be off when nothing is in progress | 18:04 |
| clif_h | I see | 18:04 |
| * BadCub goes to look on eBay for a "Now Serving" sign to hang above NobodyCam 's desk | 18:05 | |
| NobodyCam | then modify pixie boots to tell me the next number | 18:05 |
| JoshNang | clif_h: yeah, this was the major stumbling block for cleaning | 18:05 |
| clif_h | I guess I thought that the agent stayed put or was preloaded onto nodes | 18:05 |
| jroll | clif_h: the weirder part is going to be that the agent will forget it has an image cached when rebooting... might need to change the code in prepare() | 18:05 |
| jroll | clif_h: it is at rackspace :| | 18:05 |
| clif_h | haha | 18:05 |
| jroll | clif_h: see also my super-old-not-yet-merged spec for long-running ramdisks | 18:05 |
| JayF | jroll: we should readd long-running ramdisks to the list for L | 18:06 |
| jroll | yeah | 18:06 |
| *** Marga_ has quit IRC | 18:06 | |
| * JayF added | 18:06 | |
| jroll | thanks | 18:07 |
| *** Marga_ has joined #openstack-ironic | 18:07 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/python-ironicclient: Add support for generating a config drive https://review.openstack.org/167725 | 18:08 |
| *** ChuckC has quit IRC | 18:08 | |
| *** ijw_ has joined #openstack-ironic | 18:08 | |
| lucasagomes | first stab at it... lemme know what you guys think | 18:08 |
| lucasagomes | I will call it a day folks | 18:08 |
| lucasagomes | have a good night everyone! | 18:08 |
| JayF | whoa, that sounds cool | 18:08 |
| *** ChuckC has joined #openstack-ironic | 18:09 | |
| *** ijw has quit IRC | 18:09 | |
| lucasagomes | (: | 18:10 |
| *** lucasagomes is now known as lucas-dinner | 18:10 | |
| *** ijw has joined #openstack-ironic | 18:10 | |
| openstackgerrit | Merged openstack/ironic: Add root device hints documentation https://review.openstack.org/166801 | 18:11 |
| *** ijw_ has quit IRC | 18:12 | |
| *** Marga_ has quit IRC | 18:13 | |
| *** Marga_ has joined #openstack-ironic | 18:13 | |
| *** stendulker has quit IRC | 18:14 | |
| NobodyCam | night lucas-dinner | 18:14 |
| BadCub | g'night lucas-dinner | 18:15 |
| *** pas-ha has quit IRC | 18:22 | |
| * BadCub updates med/low bugs on pad | 18:22 | |
| BadCub | how we doin on all of our documentation stuffs? | 18:26 |
| *** pelix has quit IRC | 18:27 | |
| *** krtaylor has quit IRC | 18:29 | |
| dtantsur | g'night! | 18:29 |
| *** dtantsur is now known as dtantsur|afk | 18:29 | |
| NobodyCam | night dtantsur|afk | 18:29 |
| BadCub | g'night dtantsur|afk | 18:29 |
| BadCub | are any non-HP cores available to do a review on: Deploys post whole disk image deploy fails - https://review.openstack.org/#/c/167636/ ? | 18:32 |
| *** Marga_ has quit IRC | 18:38 | |
| *** anderbubble has quit IRC | 18:39 | |
| *** Marga_ has joined #openstack-ironic | 18:39 | |
| NobodyCam | JoshNang: some how I had in my mind we were going to wait until L to enable cleaning, was that only if the nova patches didn't land? | 18:42 |
| JoshNang | NobodyCam: that was the original discussion. one +2 away from both nova patches landing | 18:42 |
| NobodyCam | ya | 18:42 |
| JoshNang | though, if we want to change that, we need to get a string freeze exception to change the config | 18:43 |
| NobodyCam | ugg | 18:44 |
| JoshNang | and as i'm still tracking down why tempest doesn't like the config change, i'm ok letting it default to false until we get it more battle tested. deva still had some question on whether it would scale | 18:44 |
| NobodyCam | JoshNang: for /me I would feel better waiting, but can be convinced to "jump" then gun :-p | 18:45 |
| BadCub | New bugs are added to pad for rc-1 | 18:45 |
| NobodyCam | its a big change for the user experence | 18:45 |
| BadCub | the only ting we are't tracking at this point are docs and critical patches for rc-1 | 18:46 |
| NobodyCam | (and a easy to filp conf setting) :) | 18:46 |
| *** Marga_ has quit IRC | 18:46 | |
| *** Marga_ has joined #openstack-ironic | 18:48 | |
| JoshNang | NobodyCam: agreed :) | 18:50 |
| *** Sukhdev has joined #openstack-ironic | 18:51 | |
| *** krtaylor has joined #openstack-ironic | 18:52 | |
| *** ijw_ has joined #openstack-ironic | 18:57 | |
| *** ijw has quit IRC | 19:00 | |
| *** Nisha_away has quit IRC | 19:00 | |
| *** harlowja is now known as harlowja_away | 19:02 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Attach ilo_boot_iso only if node is active https://review.openstack.org/167317 | 19:04 |
| *** anderbubble has joined #openstack-ironic | 19:06 | |
| *** achanda has quit IRC | 19:10 | |
| NobodyCam | brb | 19:11 |
| *** Marga_ has quit IRC | 19:14 | |
| *** Marga_ has joined #openstack-ironic | 19:14 | |
| *** Marga_ has quit IRC | 19:14 | |
| *** Marga_ has joined #openstack-ironic | 19:15 | |
| * BadCub has to run out for a bit | 19:21 | |
| *** BadCub is now known as BadCub_Away | 19:21 | |
| *** Marga_ has quit IRC | 19:25 | |
| *** harlowja_away is now known as harlowja | 19:38 | |
| NobodyCam | adam_g: happen to be around? | 19:41 |
| NobodyCam | never mind :-p | 19:46 |
| *** lucas-dinner has quit IRC | 19:46 | |
| *** mtanino has quit IRC | 19:51 | |
| *** igordcard_ has joined #openstack-ironic | 19:53 | |
| *** pas-ha has joined #openstack-ironic | 19:54 | |
| *** mtanino has joined #openstack-ironic | 19:55 | |
| *** davide has quit IRC | 20:02 | |
| adam_g | NobodyCam, hey | 20:18 |
| NobodyCam | hey hey adam_g | 20:18 |
| NobodyCam | I was looking at https://review.openstack.org/#/c/167038 | 20:18 |
| NobodyCam | and https://review.openstack.org/164369 | 20:19 |
| NobodyCam | and was thinking should we use 1.0 in 167038 | 20:20 |
| adam_g | NobodyCam, sure. i made 167038 against what is current. if 164369 lands ya--it should be 1.0 | 20:21 |
| adam_g | ill WIP it | 20:21 |
| NobodyCam | :) | 20:21 |
| *** Marga_ has joined #openstack-ironic | 20:26 | |
| *** Marga_ has quit IRC | 20:31 | |
| *** anderbubble has quit IRC | 20:32 | |
| *** anderbubble has joined #openstack-ironic | 20:38 | |
| *** BadCub_Away is now known as BadCub | 20:38 | |
| *** ijw has joined #openstack-ironic | 20:39 | |
| *** achanda has joined #openstack-ironic | 20:39 | |
| *** ijw_ has quit IRC | 20:43 | |
| *** pas-ha has quit IRC | 20:43 | |
| *** ijw has quit IRC | 20:44 | |
| *** jiangfei|2 has quit IRC | 20:44 | |
| jroll | is it generally accepted that we shouldn't do anything to a node in maintenance? | 20:45 |
| jroll | e.g. would anybody be opposed to a patch that doesn't do cleaning on a node in maintenance | 20:45 |
| jroll | or at least waits for it | 20:46 |
| jroll | to come out of maintenance | 20:46 |
| * jroll is explaining this poorly | 20:46 | |
| *** ijw_ has joined #openstack-ironic | 20:46 | |
| *** jiangfei has joined #openstack-ironic | 20:47 | |
| JayF | Cleaning not happening for maint'd nodes++ | 20:49 |
| JoshNang | i'm +1 to it | 20:50 |
| TheJulia | jroll: That makes sense to me | 20:51 |
| *** mjturek1 has quit IRC | 20:51 | |
| jroll | whee. | 20:51 |
| jroll | thanks | 20:51 |
| *** Sukhdev has quit IRC | 20:51 | |
| JayF | jroll: tbh, worst case it's a config option | 20:51 |
| NobodyCam | jroll: ya, because if the node was put there because of a failur then the op may need to look at it | 20:52 |
| NobodyCam | and cleaning it would be bad at that point | 20:52 |
| jroll | right | 20:53 |
| jroll | I'm thinking just noop AgentBaseVendor.heartbeat if in maintenance | 20:53 |
| jroll | (handles deploys and such too) | 20:53 |
| jroll | but I'm not entirely sure | 20:53 |
| jroll | at any rate | 20:53 |
| * jroll writes tests | 20:53 | |
| *** Marga_ has joined #openstack-ironic | 20:54 | |
| *** alexpilotti has joined #openstack-ironic | 20:54 | |
| *** jgrimm is now known as zz_jgrimm | 20:56 | |
| *** mrda-away is now known as mrda | 20:57 | |
| NobodyCam | mornign mrda | 20:57 |
| NobodyCam | :-p | 20:57 |
| mrda | Morning Ironic | 20:58 |
| mrda | ...and NobodyCam :) | 20:58 |
| NobodyCam | :-p | 20:58 |
| *** trown is now known as trown|outttypeww | 21:00 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Support the equivalent of the Juno API https://review.openstack.org/164369 | 21:02 |
| NobodyCam | rloo: are you around? | 21:03 |
| *** dprince has quit IRC | 21:12 | |
| *** davide has joined #openstack-ironic | 21:13 | |
| *** achanda has quit IRC | 21:22 | |
| *** mjturek1 has joined #openstack-ironic | 21:22 | |
| *** Marga_ has quit IRC | 21:23 | |
| *** kkoski has quit IRC | 21:23 | |
| *** achanda has joined #openstack-ironic | 21:25 | |
| *** Marga_ has joined #openstack-ironic | 21:26 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Agent driver: no-op heartbeat for maintenanced node https://review.openstack.org/167801 | 21:28 |
| jroll | JoshNang: JayF: ^^ | 21:29 |
| JoshNang | jroll: "self.assertEqual(0, ncrc_mock)" is that a shortcut for ncrc_mock.call_count? | 21:31 |
| jroll | oh snap | 21:31 |
| jroll | thanks | 21:31 |
| jroll | lol | 21:31 |
| jroll | why doesn't python just know what I mean | 21:31 |
| JoshNang | lol | 21:31 |
| jroll | I'm also wondering if we should just fail these and leave it to the operator to clean up | 21:32 |
| JoshNang | np. i was kinda hoping it was a real thing | 21:32 |
| jroll | but that's more work for me | 21:32 |
| JoshNang | meh, i think this is fine | 21:33 |
| jroll | fixeded | 21:33 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Agent driver: no-op heartbeat for maintenanced node https://review.openstack.org/167801 | 21:33 |
| JoshNang | jroll: +1'ed. much more elegant than the downstream patch | 21:34 |
| jroll | heh | 21:34 |
| jroll | I was fixing that and was like... this is simpler | 21:34 |
| JoshNang | heh definitely | 21:35 |
| *** penick has joined #openstack-ironic | 21:36 | |
| *** achanda has quit IRC | 21:41 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Allow node.instance_uuid to be removed during cleaning https://review.openstack.org/167033 | 21:44 |
| JoshNang | ^ jroll we were talking about this bug a couple days ago. | 21:44 |
| jroll | mmm | 21:45 |
| jroll | will look | 21:45 |
| jroll | is that what's failing tests elsewhere? | 21:45 |
| JoshNang | part of it. there's still another bug somewhere that's causing tempest to fail | 21:45 |
| jroll | JoshNang: what if you just added CLEANING to UPDATE_ALLOWED_STATES? | 21:46 |
| JoshNang | heh that was patch 1 | 21:46 |
| jroll | oh you only want to allow instance id to change | 21:46 |
| jroll | got it. | 21:46 |
| JoshNang | right | 21:46 |
| *** achanda has joined #openstack-ironic | 21:52 | |
| *** mgoddard has quit IRC | 21:55 | |
| jroll | +2 btw | 21:57 |
| JoshNang | thanks! | 21:58 |
| JoshNang | agreed though, i don't love the hack | 21:59 |
| *** penick has quit IRC | 21:59 | |
| jlvillal | JoshNang: Thanks for the hack :) | 21:59 |
| JoshNang | jlvillal: np. thanks for the review! | 21:59 |
| jlvillal | JoshNang: I'm hoping to get better at reviews. I'm still working on figuring what is and isn't important. | 22:00 |
| JoshNang | jlvillal: no worries, this is how we learn! | 22:00 |
| jlvillal | JoshNang: Thanks | 22:00 |
| NobodyCam | Non-work: w00 h00 NobodyCam's seti@home team has passed 1,000,000 work units / credits :-p | 22:02 |
| *** penick has joined #openstack-ironic | 22:03 | |
| *** jamielennox|away is now known as jamielennox | 22:08 | |
| JoshNang | \o/ | 22:08 |
| NobodyCam | :-p and its only taken sense 1999 | 22:10 |
| NobodyCam | lol | 22:10 |
| JoshNang | heh i used to run folding@home on every idle computer when my apartment had flat-rate utilities | 22:13 |
| NobodyCam | nice! I had lots of sysytems at one time. now its just a screen saver on one... I need to change that sometime :-p | 22:15 |
| *** jcoufal has quit IRC | 22:16 | |
| *** achanda has quit IRC | 22:17 | |
| *** Marga_ has quit IRC | 22:19 | |
| *** Marga_ has joined #openstack-ironic | 22:20 | |
| *** davide has quit IRC | 22:23 | |
| *** achanda has joined #openstack-ironic | 22:23 | |
| *** athomas has quit IRC | 22:27 | |
| *** anderbubble has quit IRC | 22:30 | |
| *** jcoufal has joined #openstack-ironic | 22:31 | |
| *** andreykurilin_ has joined #openstack-ironic | 22:37 | |
| *** jcoufal has quit IRC | 22:40 | |
| NobodyCam | j* would would you guys think if I put up a patch like: http://paste.openstack.org/show/MEVNf1S6XFJLXeffsl1j/ | 22:43 |
| NobodyCam | swould would/what would/ | 22:44 |
| jroll | Ran 2140 tests in 46.352s <- we should improve this | 22:49 |
| JoshNang | NobodyCam: great start. | 22:49 |
| jroll | NobodyCam: I hate docs, throw it away | 22:49 |
| jroll | ;D | 22:49 |
| JoshNang | for the building bit: https://github.com/openstack/ironic-python-agent/tree/master/imagebuild/coreos | 22:49 |
| NobodyCam | i'll continue if you like ? | 22:50 |
| JayF | NobodyCam: that's awesome, builds are doc'd like JoshNang ^ linked | 22:50 |
| NobodyCam | ack :) | 22:50 |
| *** Marga_ has quit IRC | 22:51 | |
| *** Marga_ has joined #openstack-ironic | 22:52 | |
| *** andreykurilin_ has quit IRC | 22:56 | |
| *** korekhov_ has quit IRC | 22:57 | |
| *** achanda has quit IRC | 22:57 | |
| *** korekhov has joined #openstack-ironic | 22:57 | |
| *** foexle has quit IRC | 23:00 | |
| *** penick has quit IRC | 23:02 | |
| *** korekhov has quit IRC | 23:02 | |
| *** achanda has joined #openstack-ironic | 23:03 | |
| *** penick has joined #openstack-ironic | 23:04 | |
| *** Marga_ has quit IRC | 23:06 | |
| *** Marga_ has joined #openstack-ironic | 23:07 | |
| *** ijw_ has quit IRC | 23:09 | |
| *** chlong has joined #openstack-ironic | 23:10 | |
| *** ChuckC has quit IRC | 23:18 | |
| *** krtaylor has quit IRC | 23:21 | |
| *** jmccrory has quit IRC | 23:21 | |
| jroll | who wants to see something super interesting :D | 23:23 |
| BadCub | just how super interesting? | 23:23 |
| jroll | https://gist.github.com/jimrollenhagen/f422fce1d4d1d215d266 | 23:23 |
| BadCub | hehehe | 23:23 |
| jroll | lol | 23:23 |
| jroll | test breakage interesting :| | 23:23 |
| jroll | devananda: ^^ | 23:24 |
| jroll | oh you're in europe, bah | 23:24 |
| jroll | NobodyCam: ^^ | 23:24 |
| jroll | :P | 23:24 |
| BadCub | it is so polite with it's little "congratulations :-)" message lol | 23:24 |
| jroll | ikr | 23:24 |
| NobodyCam | huh | 23:24 |
| jroll | but yeah... something racy with either inspect code or rpcapi tests | 23:25 |
| * BadCub always finds the most irrelevant part that makes him giggle | 23:25 | |
| jroll | this is in our downstream repo, so it could be something I did... but kinda doubt it | 23:25 |
| jroll | dunno. | 23:25 |
| jroll | I added some rpcapi tests | 23:25 |
| *** achanda has quit IRC | 23:25 | |
| * jroll will try master too | 23:25 | |
| NobodyCam | have you tried that agenst the upstram code? | 23:25 |
| NobodyCam | jroll: +++ | 23:25 |
| BadCub | I say you broke it. It is all your fault. World Famine will ensue because of this. | 23:26 |
| BadCub | hehehe | 23:26 |
| jroll | :) | 23:26 |
| NobodyCam | lol | 23:26 |
| *** achanda has joined #openstack-ironic | 23:27 | |
| jroll | NobodyCam: it works :( | 23:30 |
| jroll | I can't imagine how I broke this | 23:30 |
| jroll | I still conjecture that inspection code broke our downstream :P | 23:31 |
| *** ChuckC has joined #openstack-ironic | 23:32 | |
| *** chlong has quit IRC | 23:37 | |
| *** chlong has joined #openstack-ironic | 23:37 | |
| jroll | oh my | 23:41 |
| jroll | I think it's racing this https://github.com/openstack/ironic/blob/master/ironic/tests/api/v1/test_nodes.py#L1509-1529 | 23:42 |
| NobodyCam | jroll: :( | 23:42 |
| jroll | the commit where it breaks doesn't touch rpcapi or test_rpcapi, but it does touch this | 23:42 |
| jroll | I'm going to fix that tonight | 23:42 |
| jroll | why don't we just patch the rpcapi client | 23:43 |
| * jroll stops grumbling and fixes | 23:43 | |
| jroll | OH | 23:44 |
| jroll | hahaha | 23:44 |
| jroll | I managed to miss a addCleanup() there | 23:44 |
| openstackgerrit | Chris Krelle proposed openstack/ironic-python-agent: Update README file. https://review.openstack.org/167841 | 23:44 |
| jroll | dang, y'all were right | 23:45 |
| NobodyCam | jroll: feel free to -2 ^^^^^ | 23:45 |
| jroll | NobodyCam: why would I -2 that | 23:45 |
| NobodyCam | because you don't like docs | 23:45 |
| NobodyCam | :-p | 23:45 |
| NobodyCam | lol | 23:45 |
| jroll | lol | 23:45 |
| jroll | whoa, http://www.eteknix.com/samsung-allegedly-acquire-amd/ | 23:46 |
| *** ijw has joined #openstack-ironic | 23:47 | |
| NobodyCam | oh does that mean M$ is going to try and buy intel | 23:48 |
| * NobodyCam thinks about buyingintel stock | 23:48 | |
| JoshNang | ooo fun | 23:49 |
| *** Haomeng has joined #openstack-ironic | 23:49 | |
| * BadCub slaps a -1 on it | 23:50 | |
| NobodyCam | lol market down aalmost 300 points and our biggest loser (overall) stock managed to gain 2.8 % today... lol | 23:50 |
| * NobodyCam looks | 23:51 | |
| BadCub | eek | 23:51 |
| *** Haomeng|2 has quit IRC | 23:52 | |
| *** r-daneel has quit IRC | 23:52 | |
| *** alexpilotti has quit IRC | 23:53 | |
| *** ijw has quit IRC | 23:58 | |
| BadCub | hmmm... Intel may be good one to have | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!