*** chas_ has quit IRC | 00:00 | |
*** chas_ has joined #openstack-ironic | 00:02 | |
*** chas_ has quit IRC | 00:07 | |
*** aNuposic has quit IRC | 00:08 | |
*** aNuposic has joined #openstack-ironic | 00:09 | |
*** rloo has quit IRC | 00:16 | |
*** aNuposic has quit IRC | 00:22 | |
*** chlong has joined #openstack-ironic | 00:23 | |
*** chas_ has joined #openstack-ironic | 00:36 | |
*** chas_ has quit IRC | 00:40 | |
*** priteau has joined #openstack-ironic | 00:40 | |
*** priteau has quit IRC | 00:45 | |
*** siva has quit IRC | 00:47 | |
*** jerrygb has quit IRC | 00:48 | |
*** siva has joined #openstack-ironic | 00:48 | |
*** siva has quit IRC | 00:49 | |
*** jerrygb has joined #openstack-ironic | 00:49 | |
*** jerrygb has quit IRC | 00:49 | |
*** jerrygb has joined #openstack-ironic | 00:49 | |
*** aarefiev has quit IRC | 00:50 | |
*** rpioso has quit IRC | 00:59 | |
*** hoangcx has joined #openstack-ironic | 01:01 | |
*** hoangcx_ has joined #openstack-ironic | 01:02 | |
*** siva has joined #openstack-ironic | 01:02 | |
openstackgerrit | Sukhdev Kapur proposed openstack/ironic: Add support for Security Groups for baremetal servers https://review.openstack.org/361451 | 01:12 |
---|---|---|
*** Sukhdev has quit IRC | 01:14 | |
*** zhangjl has joined #openstack-ironic | 01:15 | |
*** rajinir has quit IRC | 01:16 | |
*** ijw has quit IRC | 01:18 | |
*** ijw has joined #openstack-ironic | 01:24 | |
*** tuanluong has joined #openstack-ironic | 01:30 | |
*** syed__ has quit IRC | 01:35 | |
*** chas_ has joined #openstack-ironic | 01:36 | |
*** vsaienko has joined #openstack-ironic | 01:39 | |
*** siva has quit IRC | 01:40 | |
*** vsaienko has quit IRC | 01:41 | |
*** chas_ has quit IRC | 01:41 | |
*** zhangyufei has joined #openstack-ironic | 01:44 | |
*** shuquan has joined #openstack-ironic | 01:57 | |
*** gcb has joined #openstack-ironic | 02:16 | |
*** zhangyufei has quit IRC | 02:17 | |
*** zhangyufei has joined #openstack-ironic | 02:18 | |
*** shuquan has quit IRC | 02:18 | |
*** zhangyufei has quit IRC | 02:18 | |
*** godp1301 has joined #openstack-ironic | 02:21 | |
*** cdearborn has quit IRC | 02:24 | |
*** zhangyufei has joined #openstack-ironic | 02:26 | |
*** godp1301 has quit IRC | 02:29 | |
*** amotoki has joined #openstack-ironic | 02:35 | |
*** aNuposic has joined #openstack-ironic | 02:37 | |
*** zhangyufei has quit IRC | 02:37 | |
*** zhangyufei has joined #openstack-ironic | 02:38 | |
*** aNuposic has quit IRC | 02:44 | |
*** shuquan has joined #openstack-ironic | 02:48 | |
*** yuanying has quit IRC | 02:51 | |
*** shuquan has quit IRC | 02:56 | |
*** amotoki has quit IRC | 03:12 | |
*** shuquan has joined #openstack-ironic | 03:15 | |
*** sabeen has joined #openstack-ironic | 03:18 | |
*** siva has joined #openstack-ironic | 03:29 | |
*** fragatin_ has joined #openstack-ironic | 03:32 | |
*** jerrygb has quit IRC | 03:33 | |
*** godp1301 has joined #openstack-ironic | 03:34 | |
*** fragatina has quit IRC | 03:35 | |
*** siva_ has joined #openstack-ironic | 03:35 | |
*** fragatina has joined #openstack-ironic | 03:36 | |
*** fragatin_ has quit IRC | 03:37 | |
*** siva has quit IRC | 03:37 | |
*** siva_ is now known as siva | 03:37 | |
*** Nisha_away has joined #openstack-ironic | 03:40 | |
*** fragatina has quit IRC | 03:40 | |
*** yuanying has joined #openstack-ironic | 03:49 | |
*** godp1301 has quit IRC | 03:49 | |
*** bks has joined #openstack-ironic | 03:53 | |
*** kumarb has joined #openstack-ironic | 03:58 | |
*** bks has quit IRC | 03:59 | |
*** persia has quit IRC | 03:59 | |
*** lindycoder has joined #openstack-ironic | 03:59 | |
*** godp1301 has joined #openstack-ironic | 03:59 | |
*** siva has quit IRC | 04:00 | |
*** persia has joined #openstack-ironic | 04:01 | |
*** siva has joined #openstack-ironic | 04:02 | |
*** zhangyufei has quit IRC | 04:03 | |
*** kumarb has quit IRC | 04:03 | |
*** chlong has quit IRC | 04:13 | |
*** rama_y has joined #openstack-ironic | 04:21 | |
*** godp1301 has quit IRC | 04:22 | |
*** links has joined #openstack-ironic | 04:23 | |
*** jerrygb has joined #openstack-ironic | 04:34 | |
*** chas_ has joined #openstack-ironic | 04:39 | |
*** jerrygb has quit IRC | 04:39 | |
*** aNuposic has joined #openstack-ironic | 04:39 | |
*** hoangcx_ has quit IRC | 04:42 | |
*** priteau has joined #openstack-ironic | 04:42 | |
*** chas_ has quit IRC | 04:43 | |
*** jrcloud has quit IRC | 04:45 | |
*** priteau has quit IRC | 04:47 | |
*** rama_y has quit IRC | 04:55 | |
*** ijw has quit IRC | 05:02 | |
*** zhangyufei has joined #openstack-ironic | 05:02 | |
*** Sukhdev has joined #openstack-ironic | 05:05 | |
*** aNuposic has quit IRC | 05:11 | |
*** zhangyufei has quit IRC | 05:17 | |
*** siva has quit IRC | 05:19 | |
openstackgerrit | Sukhdev Kapur proposed openstack/ironic: Documentation for Security Groups for baremetal servers https://review.openstack.org/393962 | 05:26 |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic power interface for soft reboot and soft power off https://review.openstack.org/216730 | 05:32 |
*** yuanying has quit IRC | 05:34 | |
*** aparnav has joined #openstack-ironic | 05:44 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Ipmitool power driver for soft reboot and soft power off https://review.openstack.org/216738 | 05:44 |
*** amotoki has joined #openstack-ironic | 05:45 | |
*** jaosorior has joined #openstack-ironic | 05:51 | |
*** zhangyufei has joined #openstack-ironic | 05:52 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: iRMC power driver for soft reboot and soft power off https://review.openstack.org/216743 | 06:00 |
*** vmud213 has joined #openstack-ironic | 06:05 | |
*** ChubYann has quit IRC | 06:09 | |
*** rcernin has joined #openstack-ironic | 06:17 | |
*** Sukhdev has quit IRC | 06:23 | |
*** shuquan has quit IRC | 06:25 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Update the existing APIs due to adding get_supported_power_states https://review.openstack.org/382194 | 06:28 |
*** shuquan_ has joined #openstack-ironic | 06:30 | |
*** harlowja_ has quit IRC | 06:33 | |
*** jaosorior has quit IRC | 06:34 | |
*** jaosorior has joined #openstack-ironic | 06:35 | |
*** jerrygb has joined #openstack-ironic | 06:35 | |
*** shuquan_ has quit IRC | 06:37 | |
*** sabeen has quit IRC | 06:37 | |
*** chas_ has joined #openstack-ironic | 06:40 | |
*** shuquan has joined #openstack-ironic | 06:41 | |
*** vsaienko has joined #openstack-ironic | 06:41 | |
*** jerrygb has quit IRC | 06:42 | |
*** kamtamtun has joined #openstack-ironic | 06:42 | |
*** vsaienko has quit IRC | 06:43 | |
*** chas_ has quit IRC | 06:45 | |
*** pcaruana has joined #openstack-ironic | 06:51 | |
*** rcernin has quit IRC | 06:56 | |
openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic management I/F for Inject NMI https://review.openstack.org/348191 | 06:56 |
*** rcernin has joined #openstack-ironic | 06:56 | |
*** chlong has joined #openstack-ironic | 06:58 | |
*** fxpester has joined #openstack-ironic | 07:03 | |
*** chlong has quit IRC | 07:09 | |
*** vsaienko has joined #openstack-ironic | 07:20 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip some steps for multinode case https://review.openstack.org/368646 | 07:20 |
*** tesseract has joined #openstack-ironic | 07:20 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add devstack setup_vxlan_network() https://review.openstack.org/368647 | 07:20 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update iptables rules and services IPs for multinode https://review.openstack.org/368648 | 07:20 |
*** tesseract is now known as Guest40216 | 07:20 | |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Testing multinode stuff https://review.openstack.org/368173 | 07:20 |
*** jtomasek has joined #openstack-ironic | 07:32 | |
*** pester has joined #openstack-ironic | 07:34 | |
*** vsaienko has quit IRC | 07:35 | |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Introduce VolumeConnector object https://review.openstack.org/214584 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Introduce new RPCs to support volume connector operation https://review.openstack.org/214585 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Add REST API for volume connector and volume target operation https://review.openstack.org/214586 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Introduce new RPCs to support volume target operations https://review.openstack.org/285220 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Introduce VolumeTarget object https://review.openstack.org/285219 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Add volume_targets table to save target information https://review.openstack.org/285218 | 07:36 |
openstackgerrit | Hironori Shiina proposed openstack/ironic: Add volume_connector table to save connector information https://review.openstack.org/200983 | 07:36 |
*** fxpester has quit IRC | 07:38 | |
*** vsaienko has joined #openstack-ironic | 07:45 | |
*** openstackgerrit has quit IRC | 07:48 | |
*** openstackgerrit has joined #openstack-ironic | 07:48 | |
*** rbartal has joined #openstack-ironic | 07:54 | |
*** yuanying has joined #openstack-ironic | 07:55 | |
*** watanabe_isao has joined #openstack-ironic | 07:56 | |
*** Nisha_away has quit IRC | 07:59 | |
*** ohamada has joined #openstack-ironic | 08:02 | |
*** vsaienko has quit IRC | 08:02 | |
*** ccamacho has joined #openstack-ironic | 08:04 | |
*** shuquan has quit IRC | 08:05 | |
*** vsaienko has joined #openstack-ironic | 08:07 | |
*** hoangcx has quit IRC | 08:13 | |
*** shuquan_ has joined #openstack-ironic | 08:13 | |
*** moshele has joined #openstack-ironic | 08:16 | |
*** hoangcx has joined #openstack-ironic | 08:17 | |
*** vsaienko has quit IRC | 08:18 | |
xek | morning :) | 08:23 |
*** ralonsoh has joined #openstack-ironic | 08:26 | |
*** aNuposic has joined #openstack-ironic | 08:30 | |
*** jerrygb has joined #openstack-ironic | 08:38 | |
*** jpich has joined #openstack-ironic | 08:41 | |
*** chas_ has joined #openstack-ironic | 08:42 | |
*** priteau has joined #openstack-ironic | 08:44 | |
*** jerrygb has quit IRC | 08:44 | |
*** chas_ has quit IRC | 08:47 | |
*** priteau has quit IRC | 08:48 | |
*** portdirect_away has quit IRC | 08:50 | |
*** chas_ has joined #openstack-ironic | 08:53 | |
*** rbartal has quit IRC | 08:58 | |
*** shuquan_ has quit IRC | 08:59 | |
*** betherly has joined #openstack-ironic | 08:59 | |
*** shuquan_ has joined #openstack-ironic | 08:59 | |
*** zzzeek has quit IRC | 09:00 | |
*** strigazi_AFK is now known as strigazi | 09:01 | |
*** zzzeek has joined #openstack-ironic | 09:01 | |
*** rbartal has joined #openstack-ironic | 09:02 | |
*** Nisha_away has joined #openstack-ironic | 09:15 | |
*** Linkmark has joined #openstack-ironic | 09:21 | |
*** dtantsur|afk is now known as dtantsur | 09:21 | |
dtantsur | Morning Ironic | 09:21 |
*** lucas-afk is now known as lucasagomes | 09:23 | |
lucasagomes | morning all | 09:24 |
*** Romanenko_K has joined #openstack-ironic | 09:24 | |
dtantsur | morning lucasagomes | 09:24 |
*** ohamada has quit IRC | 09:24 | |
*** ohamada has joined #openstack-ironic | 09:24 | |
*** priteau has joined #openstack-ironic | 09:25 | |
*** aNuposic has quit IRC | 09:25 | |
*** yolanda has quit IRC | 09:31 | |
*** milan has joined #openstack-ironic | 09:31 | |
*** yolanda has joined #openstack-ironic | 09:31 | |
openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: [WIP] Add a new Hardware Manager for Mellanox NICs https://review.openstack.org/334816 | 09:33 |
openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: [WIP] Add a new Hardware Manager for Mellanox NICs https://review.openstack.org/334816 | 09:35 |
*** priteau has quit IRC | 09:38 | |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: Fix several errors in LLDP handling code https://review.openstack.org/396043 | 09:41 |
*** milan has quit IRC | 09:45 | |
*** milan has joined #openstack-ironic | 09:46 | |
*** amotoki_ has joined #openstack-ironic | 09:47 | |
*** pcaruana has quit IRC | 09:49 | |
*** shuquan_ has quit IRC | 09:51 | |
*** amotoki has quit IRC | 09:51 | |
milan | morning Ironic :) | 09:52 |
*** xek has left #openstack-ironic | 09:52 | |
*** shuquan has joined #openstack-ironic | 09:53 | |
dtantsur | gah, dracclient review speed is... bad... | 09:53 |
dtantsur | lucasagomes, mind approving https://review.openstack.org/#/c/394465/ and https://review.openstack.org/#/c/394466/ please? | 09:53 |
patchbot | patch 394465 - python-dracclient - Raise a reasonable error on non-nullable flags wit... | 09:53 |
patchbot | patch 394466 - python-dracclient - Raise a specific exception instead of generic Attr... | 09:53 |
dtantsur | it doesn't seem like we have many active reviewers... | 09:53 |
lucasagomes | dtantsur, 1 sec, will take a look in a few | 09:53 |
milan | morning dtantsur, lucasagomes! :) | 09:54 |
*** MattMan has quit IRC | 09:54 | |
dtantsur | morning milan | 09:54 |
*** MattMan has joined #openstack-ironic | 09:54 | |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node provision state change notification https://review.openstack.org/348437 | 09:56 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add ironic resources CRUD notifications https://review.openstack.org/356541 | 09:57 |
openstackgerrit | Aline Bousquet proposed openstack/ironic: Add possibility to remove chassis_uuid from a node https://review.openstack.org/390558 | 09:59 |
*** amotoki has joined #openstack-ironic | 09:59 | |
*** sambetts|afk is now known as sambetts | 10:00 | |
*** shuquan has quit IRC | 10:01 | |
*** shuquan has joined #openstack-ironic | 10:01 | |
*** pcaruana has joined #openstack-ironic | 10:01 | |
openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Add new capabilities to ironic https://review.openstack.org/338138 | 10:02 |
*** amotoki_ has quit IRC | 10:03 | |
*** karthi has joined #openstack-ironic | 10:03 | |
karthi | Hello IRONICers... | 10:04 |
*** yibo has quit IRC | 10:06 | |
*** athomas has joined #openstack-ironic | 10:06 | |
lucasagomes | dtantsur, the parameters on that function sounds a bit odd. If nullable = True it will raise an exception when the value is None ?! Shouldn't it be the other way around ? | 10:07 |
lucasagomes | (I know it was there before, but still...) | 10:07 |
*** Nisha_away has quit IRC | 10:07 | |
lucasagomes | dtantsur, oh ignore me | 10:08 |
dtantsur | this whole thing is odd IMO :/ | 10:08 |
lucasagomes | heh | 10:08 |
dtantsur | I don't like the fact that it raises errors to a user when *we* don't know when a field is nullable or not | 10:08 |
dtantsur | IMO it should be the other way around: here are the mandatory fields, everything else is optional | 10:08 |
dtantsur | dunno if ifarkas would agree with that.. I can try after the release. right now I only want to merge the 3 damned patches to fix inspection >_< | 10:09 |
dtantsur | (the 3rd is https://review.openstack.org/394510, I need mgould|afk to re-review it) | 10:09 |
patchbot | patch 394510 - python-dracclient - Make name and description nullable for virtual disks | 10:09 |
*** jaosorior is now known as jaosorior_lunch | 10:10 | |
lucasagomes | dtantsur, right, yeah probably nullable should be the default | 10:11 |
*** portdirect has joined #openstack-ironic | 10:19 | |
*** portdirect is now known as portdirect_away | 10:19 | |
*** portdirect_away is now known as portdirect | 10:20 | |
sambetts | Morning all, sorry for being absent yesterday family emergency happened in the morning | 10:23 |
dtantsur | morning sambetts, I hope everything is ok now | 10:23 |
sambetts | dtantsur: things have calmed down now, thanks :) | 10:24 |
dtantsur | cool | 10:24 |
milan | morning sambetts! :) | 10:24 |
dtantsur | sambetts, I wanted to show you https://review.openstack.org/#/c/395632/ yesterday, but it is already approved. lemme know if you see any problems | 10:24 |
patchbot | patch 395632 - ironic-python-agent - Fix several errors in LLDP handling code (MERGED) | 10:24 |
sambetts | dtantsur: thanks for getting that fix into IPA, that must have takening some debugging? | 10:24 |
dtantsur | ah, you saw it :) | 10:24 |
sambetts | yup :) | 10:24 |
*** zhangjl1 has joined #openstack-ironic | 10:25 | |
dtantsur | sambetts, well, the traceback was pretty obvious, fortunately :) but it took me some time thinking why exceptions did not propagate :) | 10:25 |
* dtantsur is always forgetting __exit__ return value semantics | 10:25 | |
sambetts | dtantsur: yeah, I'm still surprised that it managed to cause the other function to return not a dict when in the function itself it never returns anything that can't be a dict | 10:26 |
dtantsur | sambetts, yeah. __exit__ silences the exception, so the context manager yields control past the "return" command | 10:27 |
sambetts | *mind splodes* | 10:27 |
dtantsur | as there are no more commands left, and we skipped "return", it returns None | 10:27 |
*** zhangjl has quit IRC | 10:27 | |
*** zhangjl1 has left #openstack-ironic | 10:30 | |
*** shuquan has quit IRC | 10:33 | |
openstackgerrit | Merged openstack/python-ironicclient: Update to hacking 0.12.0 and use new checks https://review.openstack.org/395588 | 10:39 |
*** jerrygb has joined #openstack-ironic | 10:40 | |
*** jerrygb has quit IRC | 10:45 | |
dtantsur | jroll, morning! please check the lldp fix backport https://review.openstack.org/#/c/396043/ | 10:47 |
patchbot | patch 396043 - ironic-python-agent (stable/newton) - Fix several errors in LLDP handling code | 10:47 |
*** hoangcx has quit IRC | 10:48 | |
dtantsur | vsaienk0, morning! were you planning to overtake https://review.openstack.org/#/c/357983/ ? (sorry if I confuse something) | 10:50 |
patchbot | patch 357983 - ironic - Support names for {cleaning,provisioning}_network (ABANDONED) | 10:50 |
dtantsur | I'm thinking of reviving it again | 10:50 |
*** jschwarz is now known as jschwarz|food | 10:51 | |
*** pester has quit IRC | 10:58 | |
vsaienk0 | dtantsur: morning, yeah, it is in my todo list | 10:58 |
dtantsur | cool! | 10:59 |
*** amotoki_ has joined #openstack-ironic | 11:06 | |
*** siva has joined #openstack-ironic | 11:08 | |
*** amotoki has quit IRC | 11:09 | |
*** hw_wutianwei has quit IRC | 11:10 | |
vdrok | morning ironic, dtantsur milan lucasagomes sambetts vsaienk0 | 11:11 |
*** mgould|afk is now known as mgould | 11:11 | |
vsaienk0 | morning vdrok :) and all Ironic'ers! | 11:11 |
mgould | morning dtantsur milan lucasagomes sambetts vsaienk0 vdrok and all Ironicers! | 11:11 |
sambetts | o/ vsaienk0, vdrok | 11:11 |
vdrok | morning mgould :) | 11:11 |
milan | morning mgould, vdrok, vsaienk0! :) | 11:11 |
*** xavierr has quit IRC | 11:12 | |
*** watanabe_isao has quit IRC | 11:12 | |
lucasagomes | vdrok, mgould vsaienk0 sambetts morning :-) | 11:13 |
lucasagomes | milan, morning! | 11:13 |
milan | hey lucasagomes! :) | 11:13 |
*** karthi has quit IRC | 11:14 | |
dtantsur | morning vdrok, mgould | 11:16 |
*** siva_ has joined #openstack-ironic | 11:18 | |
*** siva_ has left #openstack-ironic | 11:18 | |
*** xavierr has joined #openstack-ironic | 11:18 | |
*** siva_ has joined #openstack-ironic | 11:19 | |
*** siva_ has left #openstack-ironic | 11:19 | |
*** siva_ has joined #openstack-ironic | 11:20 | |
*** siva has quit IRC | 11:20 | |
*** siva_ is now known as siva | 11:20 | |
*** jschwarz|food is now known as jschwarz | 11:20 | |
*** jerrygb has joined #openstack-ironic | 11:25 | |
*** siva is now known as yostk | 11:27 | |
*** yostk is now known as skat | 11:28 | |
*** amotoki_ has quit IRC | 11:31 | |
*** jkilpatr has quit IRC | 11:37 | |
*** zhangyufei has quit IRC | 11:40 | |
*** chlong has joined #openstack-ironic | 11:43 | |
*** chlong has quit IRC | 11:44 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/396134 | 11:47 |
*** cdearborn has joined #openstack-ironic | 11:47 | |
*** amotoki has joined #openstack-ironic | 11:52 | |
*** jkilpatr has joined #openstack-ironic | 11:58 | |
*** amotoki has quit IRC | 12:00 | |
openstackgerrit | Karthik S proposed openstack/ironic-inspector-specs: Spec for retrieving NUMA node information https://review.openstack.org/396147 | 12:06 |
openstackgerrit | Vadim Hmyrov proposed openstack/ironic: Block some port updates https://review.openstack.org/347360 | 12:06 |
lucasagomes | vsaienk0, you were talking about deploying ubuntu images yesterday right ? I was wondering if we actually need a super tiny image which we can deploying using as little ram as possible | 12:09 |
lucasagomes | OR we could deploy raw images and stream them directly onto the disk | 12:09 |
lucasagomes | Ironic should already support it, we probably just need to config it in devstack | 12:10 |
*** fragatina has joined #openstack-ironic | 12:11 | |
tuanluong | morning ironic | 12:11 |
*** eroux has joined #openstack-ironic | 12:11 | |
mgould | morning tuanluong | 12:11 |
*** fragatina has quit IRC | 12:11 | |
sambetts | lucasagomes: I was talking to the neutron team at the summit regarding the fact we need to have a tenant image for testing that supports the features we need like vlans, bonding, etc and cloud-init, and ideally we shouldn't be building it on every devstack run | 12:12 |
*** fragatina has joined #openstack-ironic | 12:12 | |
vsaienk0 | lucasagomes: we can try that | 12:12 |
lucasagomes | sambetts, right, exactly. We could get a ubuntu/fedora image in raw format (I think the publish it in that format already) | 12:12 |
lucasagomes | cloud image* | 12:12 |
lucasagomes | and use it for our tests | 12:12 |
sambetts | lucasagomes: we concluded that we should create a repo that contains a bunch of image-templates to build and then when a merge happens run a post-job to build and publish the images to tarballs | 12:13 |
lucasagomes | sambetts, we don't need to build them as far as I understand | 12:13 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: SUPER WIP: Run all tests in one job. https://review.openstack.org/396155 | 12:13 |
lucasagomes | sambetts, cause I think the main concern of replacing cirros is the memory footprint | 12:14 |
lucasagomes | we need to run a bunch of VMs in parallel and if the image is too big it makes it harder for agent to deploy it, since we need to convert the image in memory | 12:14 |
mgould | lucasagomes: not sure about raw format, but they publish qcow2s | 12:14 |
*** eroux has quit IRC | 12:14 | |
lucasagomes | but, using raw images we could stream it directly to the disk from a swift tempurl | 12:15 |
sambetts | lucasagomes: You suggest that we switch over to something like the canonical bullt images? Can we guarentee that they support the stuff we need? | 12:15 |
*** eroux has joined #openstack-ironic | 12:15 | |
vsaienk0 | mgould we can convert qcow2 to raw when uploading image | 12:15 |
sambetts | I like the idea of having a repo with DIB templates that are built and published to tarballs | 12:15 |
lucasagomes | sambetts, well not really, we can publish some image under openstack | 12:15 |
sambetts | lucasagomes: right, thats what I'm suggesting | 12:15 |
sambetts | but we need to have something that publishes and updates it | 12:16 |
lucasagomes | sambetts, fair | 12:16 |
lucasagomes | my only concern with ubuntu images is their policy on publishing custom images... it would be fine as long as we don't customize them | 12:16 |
lucasagomes | otherwise I think we should pick another distro | 12:16 |
lucasagomes | (that was the reason why we changed the coreos IPA image from ubuntu to debian btw) | 12:17 |
mgould | nixos! | 12:17 |
sambetts | I'm good with a debian image | 12:17 |
lucasagomes | sambetts, will check for debian cloud images | 12:17 |
mgould | yeah, debian would probably be more sensible | 12:17 |
*** dprince has joined #openstack-ironic | 12:19 | |
mgould | anyone tried Alpine? https://nickjanetakis.com/blog/alpine-based-docker-images-make-a-difference-in-real-world-apps | 12:21 |
*** vsaienko has joined #openstack-ironic | 12:25 | |
lucasagomes | mgould, not really. I didn't know it to be honest | 12:26 |
lucasagomes | will take a look | 12:26 |
openstackgerrit | Karthik S proposed openstack/ironic-inspector-specs: Spec for retrieving NUMA node information https://review.openstack.org/396147 | 12:27 |
lucasagomes | "Alpine Linux is a very simple distribution"... "It uses its own package manager called apk" | 12:28 |
lucasagomes | *sigh* | 12:28 |
sambetts | XD | 12:28 |
mgould | yay for wheel-reinvention | 12:29 |
lucasagomes | yeah... package managers is the biggest flaw in the linux world. 25 years and it's far from being fixed | 12:31 |
mgould | OTOH, we're in a better place than OS X or Windows | 12:32 |
lucasagomes | I really don't know to be honest | 12:33 |
mgould | yeah, I'm probably out of date | 12:34 |
*** jaosorior_lunch is now known as jaosorior | 12:34 | |
lucasagomes | I mean, linux doens't have a universal format (you gotta use what your distro supports), it requires root acccess | 12:34 |
lucasagomes | it's not integrated with development tooling, you don't "code, build, test" | 12:34 |
lucasagomes | rolling a package is a whole new world totally separeted from "build" and worst than that, it only solves a problem for 1 distro at time | 12:35 |
lucasagomes | and it's painful | 12:35 |
sambetts | i like tinycore's package manager because packages are literally gziped squashfs's and they actually say not to run it as root | 12:35 |
lucasagomes | it's a damn embarrassment for the linux world to use packages like that | 12:35 |
sambetts | ;) | 12:36 |
lucasagomes | sambetts, it's the same problem... | 12:36 |
lucasagomes | fwiw, Arch linux uses tar.gz's | 12:36 |
lucasagomes | still... distros gotta talk to each other | 12:36 |
lucasagomes | not counting the npm, pip, etc... | 12:36 |
mgould | as does Slackware, but AIUI the toolchains aren't compatible | 12:36 |
sambetts | yeah, and friend of mine has had good experiences with Arch, especially the fact you can use a git repo as a package source | 12:36 |
* lucasagomes uses arch | 12:37 | |
mgould | I ran Arch for a while, it seemed fine | 12:37 |
mgould | no break-everything distro-version upgrades | 12:37 |
sambetts | oh nice, I've been thinking of switching from debian testing | 12:37 |
lucasagomes | sambetts, exactly that's great and it allows people to benefit from opensource by being able to change the source! | 12:38 |
lucasagomes | all without having to setup a apt/rpm/etc.. repo | 12:38 |
sambetts | ++ | 12:38 |
* mgould uses Fedora and CentOS at work, because dogfood | 12:38 | |
sambetts | heh | 12:38 |
mgould | and Ubuntu at home, because it works and I can't be arsed to change it :-) | 12:39 |
lucasagomes | I develop on arch but test on CentOS & Fedora. I used Fedora for a long time, but every upgrade (which is ~8months) was a pain | 12:39 |
sambetts | I use debian testing as a base, but I run a custom desktop evironment mixing xfce4 tools and the awesome WM | 12:40 |
* sambetts activaly avoids anything with gnome deps | 12:41 | |
dtantsur | "we're in a better place than OS X or Windows" for sure :) I still remember installing postresql on Mac one day | 12:41 |
* lucasagomes goes get some lunch | 12:41 | |
*** lucasagomes is now known as lucas-hungry | 12:41 | |
*** tuanluong has quit IRC | 12:43 | |
*** Nisha_away has joined #openstack-ironic | 12:45 | |
*** makowals has quit IRC | 12:48 | |
openstackgerrit | Merged openstack/ironic-python-agent: Add bash scripts style checking for IPA https://review.openstack.org/374147 | 12:48 |
*** aparnav has quit IRC | 12:48 | |
*** trown|outtypewww is now known as trown | 12:48 | |
*** makowals has joined #openstack-ironic | 12:49 | |
openstackgerrit | Hugo Nicodemos proposed openstack/ironic: Timing metrics: OneView drivers https://review.openstack.org/379619 | 12:56 |
*** portdirect has quit IRC | 12:57 | |
openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: [WIP] Add a new Hardware Manager for Mellanox NICs https://review.openstack.org/334816 | 12:57 |
nicodemos | good morning, ironic. o/ | 12:59 |
mgould | morning nicodemos | 13:00 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: SUPER WIP: Run all tests in one job. https://review.openstack.org/396155 | 13:00 |
*** hshiina has joined #openstack-ironic | 13:02 | |
*** gcb has quit IRC | 13:02 | |
openstackgerrit | Sam Betts proposed openstack/ironic-specs: Add spec for interface attach and detach API https://review.openstack.org/317636 | 13:03 |
*** ElCoyote_ has joined #openstack-ironic | 13:05 | |
openstackgerrit | Sam Betts proposed openstack/ironic: Add Virtual Network Interface Attach and Detach APIs https://review.openstack.org/327046 | 13:06 |
*** Nisha_away has quit IRC | 13:08 | |
vdrok | morning nicodemos | 13:09 |
*** bfournie has quit IRC | 13:11 | |
jroll | morning y'all | 13:14 |
*** roeyc has joined #openstack-ironic | 13:14 | |
jroll | sambetts: I wonder if there's any reason to make a GET nodes/ident/vifs/ to return the list of vifs | 13:14 |
jroll | if for nothing else, so that they can be deleted properly | 13:14 |
dtantsur | morning jroll | 13:15 |
jroll | (then we need to standardize how they're stored and such, hm) | 13:15 |
sambetts | jroll: I did wonder about it, then every network interface would need a list vifs function too | 13:15 |
sambetts | jroll: because they handle it differently | 13:15 |
jroll | sambetts: yeah, which is fine | 13:15 |
jroll | so the case I'm imagining | 13:16 |
jroll | is if a nova delete fails in a strange way | 13:16 |
jroll | and doesn't tear things down properly | 13:16 |
jroll | (or the ironic equivalent, I guess) | 13:16 |
jroll | and you end up with vifs stuck on the node | 13:16 |
sambetts | yeah you could list them and delete them | 13:16 |
sambetts | +! | 13:16 |
sambetts | +1 | 13:16 |
jroll | :) | 13:16 |
jroll | sambetts: other than that, spec lgtm | 13:17 |
sambetts | jroll: does the deprecation stuff look ok with the default action during the deprecation period? | 13:17 |
jroll | sambetts: so the hard thing there is, we can't tell an api client that it's deprecated | 13:18 |
jroll | but other than that it seems fine | 13:19 |
jroll | and as a nit, we wouldn't detect NotImplementedError, but rather the default implementation is just used | 13:19 |
jroll | right? | 13:19 |
sambetts | jroll: Yeah, originally I wasn't going to put the logic in the base driver functions, I was going to put it a layer up in the rpcapi, so then we would have detect the NotImplemented then called the default | 13:21 |
jroll | sambetts: right, but now it's in the base, ya? | 13:21 |
jroll | left a comment so we can clean up that wording | 13:21 |
sambetts | jroll: ya, I decided that was nicer :-P | 13:21 |
jroll | ++ | 13:21 |
jroll | cool, let's get the GET requests added and I think we can get this merged today | 13:21 |
sambetts | jroll: I also remembered when I was going through the spec this morning why I initially choose to make it a POST in both direction, originally I was going for making the VIF json schemaless, so you could provide whatever your network interface required through that interface, and you would use the whole JSON to represent the VIF not just the ID | 13:22 |
jroll | sambetts: I'd prefer we have a schema and IDs :) | 13:23 |
sambetts | yeah, even if they are just unique keys generated by the user | 13:24 |
*** shuquan has joined #openstack-ironic | 13:26 | |
openstackgerrit | Vadim Hmyrov proposed openstack/ironic: Block some port updates https://review.openstack.org/347360 | 13:26 |
*** lucas-hungry is now known as lucasagomes | 13:27 | |
*** fragatina has quit IRC | 13:33 | |
*** shuquan has quit IRC | 13:33 | |
*** vmud213 has quit IRC | 13:34 | |
*** shuquan_ has joined #openstack-ironic | 13:34 | |
*** jerrygb has quit IRC | 13:35 | |
*** jerrygb has joined #openstack-ironic | 13:36 | |
*** jerrygb has quit IRC | 13:40 | |
openstackgerrit | Sam Betts proposed openstack/ironic-specs: Add spec for interface attach and detach API https://review.openstack.org/317636 | 13:41 |
*** kamtamtun has quit IRC | 13:45 | |
*** roeyc has quit IRC | 13:45 | |
*** vsaienko has quit IRC | 13:46 | |
*** jerrygb has joined #openstack-ironic | 13:47 | |
*** vsaienko has joined #openstack-ironic | 13:47 | |
*** bfournie has joined #openstack-ironic | 13:48 | |
*** jheroux has joined #openstack-ironic | 13:48 | |
*** baoli_ has joined #openstack-ironic | 13:50 | |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic-specs: API Evolution - ETAG identifiers https://review.openstack.org/381991 | 13:53 |
*** roeyc has joined #openstack-ironic | 13:56 | |
*** jcoufal has joined #openstack-ironic | 13:57 | |
*** athomas has quit IRC | 13:59 | |
*** rbudden has joined #openstack-ironic | 13:59 | |
*** rloo has joined #openstack-ironic | 14:00 | |
*** rbrud has joined #openstack-ironic | 14:01 | |
*** Nisha_away has joined #openstack-ironic | 14:01 | |
*** shuquan_ has quit IRC | 14:02 | |
*** Goneri has joined #openstack-ironic | 14:04 | |
*** athomas has joined #openstack-ironic | 14:04 | |
*** portdirect has joined #openstack-ironic | 14:09 | |
*** shuquan has joined #openstack-ironic | 14:10 | |
*** godp1301 has joined #openstack-ironic | 14:10 | |
*** godp1301 has joined #openstack-ironic | 14:10 | |
*** ralonsoh has quit IRC | 14:11 | |
*** ralonsoh has joined #openstack-ironic | 14:11 | |
*** portdirect_ has joined #openstack-ironic | 14:13 | |
vdrok | morning jroll | 14:13 |
jroll | \o vdrok | 14:13 |
*** cdearborn has quit IRC | 14:14 | |
*** portdirect has quit IRC | 14:15 | |
vdrok | I'm currently looking at the bug - https://bugs.launchpad.net/ironic/+bug/1639982 and it seems, if we pass complex objects that cannot be serialized by json.dumps into IronicException constructor, we get the problem described there | 14:16 |
openstack | Launchpad bug 1639982 in Ironic "DRAC set_bios_config() vendor passthru times out waiting for reply" [Undecided,In progress] - Assigned to Richard G. Pioso (richard.pioso) | 14:16 |
*** lindycoder has joined #openstack-ironic | 14:16 | |
mariojv | JayF: thanks i'll take a look today | 14:16 |
vdrok | while trying to send them over rpc | 14:16 |
jroll | mmm | 14:16 |
*** shuquan has quit IRC | 14:16 | |
*** Nisha_away has quit IRC | 14:16 | |
vdrok | this brings a question - maybe we should always try to json.dumps kwargs in the __init__ of the exception? | 14:16 |
vdrok | this already blows up in a couple of cases | 14:17 |
openstackgerrit | Milan Kováčik proposed openstack/python-ironic-inspector-client: UUID, started_at, finished_at in the status https://review.openstack.org/395609 | 14:17 |
vdrok | the easy way to check this is https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L2138-L2139 | 14:17 |
jroll | and handle it differently if we can't dumps() it? | 14:17 |
vdrok | do set raid config with fake_ipmitool driver | 14:17 |
milan | morning jroll! :) | 14:17 |
vdrok | jroll: maybe just forbid it even | 14:18 |
jroll | hmm | 14:18 |
vdrok | or try to convert to string first | 14:18 |
jroll | ah yeah, that conductor link you gave is a good example | 14:18 |
jroll | that's what I was thinking, call str() first | 14:18 |
*** jerrygb has quit IRC | 14:19 | |
vdrok | so, json.dumps first, if it fails try str(), if that fails too, raise BadExceptionError :) | 14:19 |
jroll | heh | 14:20 |
*** jkilpatr has quit IRC | 14:20 | |
*** jkilpatr has joined #openstack-ironic | 14:22 | |
*** jkilpatr has quit IRC | 14:22 | |
dtantsur | mgould, mind rereviewing https://review.openstack.org/#/c/394510/ please? | 14:24 |
patchbot | patch 394510 - python-dracclient - Make name and description nullable for virtual disks | 14:24 |
* dtantsur thinks of self-approving it and just releasing >_< | 14:24 | |
jroll | vdrok: but yeah, something like that seems sane | 14:24 |
vdrok | ++, thanks | 14:24 |
*** hshiina has quit IRC | 14:25 | |
openstackgerrit | Sam Betts proposed openstack/ironic: Rework Ironic devstack baremetal network simulation https://review.openstack.org/392959 | 14:29 |
*** jkilpatr has joined #openstack-ironic | 14:29 | |
openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: [WIP] Add a new Hardware Manager for Mellanox NICs https://review.openstack.org/334816 | 14:29 |
*** links has quit IRC | 14:31 | |
*** zackf has joined #openstack-ironic | 14:32 | |
openstackgerrit | Mario Villaplana proposed openstack/ironic: Add notification documentation to install guide https://review.openstack.org/395787 | 14:32 |
mariojv | JayF: there's the nit fix 6 | 14:32 |
mariojv | 6 was supposed to be ^ | 14:32 |
openstackgerrit | Karthik S proposed openstack/ironic-inspector-specs: Spec for retrieving NUMA node information https://review.openstack.org/396147 | 14:32 |
openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node maintenance notifications https://review.openstack.org/396239 | 14:33 |
mariojv | i'd appreciate some ironic reviews on this nova patch if anyone has a minute - 3 line change: https://review.openstack.org/#/c/395144/ | 14:34 |
patchbot | patch 395144 - nova - Remove ironic instance resize from support matrix doc | 14:34 |
*** jaosorior has quit IRC | 14:34 | |
*** jaosorior has joined #openstack-ironic | 14:35 | |
*** rama_y has joined #openstack-ironic | 14:35 | |
jroll | I have a +2 on the attach/detach spec https://review.openstack.org/#/c/317636/ | 14:35 |
patchbot | patch 317636 - ironic-specs - Add spec for interface attach and detach API | 14:35 |
jroll | need to get that merged this week if possible | 14:35 |
*** causten has joined #openstack-ironic | 14:35 | |
jroll | dtantsur: I'll probably have some bandwidth in the next week or two to write some code - is there a way we can split up the driver comp work? | 14:38 |
*** ElCoyote_ has left #openstack-ironic | 14:38 | |
dtantsur | hmm | 14:39 |
*** jcoufal has quit IRC | 14:39 | |
dtantsur | well, you try to think about API bits | 14:39 |
jroll | I could do that | 14:40 |
dtantsur | awesome! it's of course a bit hard without actual hardware types in place.. | 14:40 |
jroll | heh | 14:41 |
jroll | I can start dropping in routes and such though | 14:41 |
dtantsur | jroll, do you have ideas on better naming than "noop"? | 14:41 |
mat128 | jroll: re attach/detach api, I also thought we would need a GET /nodes/<node_ident>/vifs/<vif_id> but found out a vif has no property other than it's name | 14:41 |
rloo | mariojv: you have any idea why that nova matrix has 'only certain ironic drivers support this'? | 14:42 |
jroll | dtantsur: hrm, yeah, was thinking about that | 14:42 |
rloo | hi and morning all you hard working ironic'ers! | 14:42 |
jroll | morning rloo :) | 14:42 |
dtantsur | morning rloo | 14:42 |
mariojv | rloo: no idea. i should ask in #openstack-nova | 14:42 |
jroll | rloo: I'd bet on copy pasta | 14:42 |
mariojv | i can do a git blame and ask the person who did it maybe | 14:43 |
mat128 | rloo: that was from the wiki -- https://github.com/openstack/nova/commit/4837c42127b87acdaf6f68d1e8956bba9a6b3fd2 | 14:43 |
jlvillal | mariojv, What editor do you use? | 14:43 |
mat128 | mariojv: ^ | 14:43 |
mariojv | jlvillal: vim, why? | 14:43 |
jlvillal | mariojv, To you use vim-fugitive plugin? | 14:43 |
jlvillal | s/To/Do/ | 14:43 |
* mariojv checks | 14:43 | |
lucasagomes | vsaienk0, 2016-11-10 14:42:34.462 393 INFO ironic_python_agent.extensions.standby [-] Image streamed onto device /dev/sda in 93.9221189022 seconds (the node has 350MiB of ram and the image deployed 1G) | 14:43 |
mat128 | lucasagomes: yay for streaming :D | 14:44 |
mariojv | jlvillal: no, i used to, but i disabled it | 14:44 |
lucasagomes | mat128, :D | 14:44 |
rloo | mat128: thx. probably not worth pursing any more. | 14:44 |
* dtantsur uses git the old school way | 14:44 | |
jlvillal | mariojv, Oh. I saw you mention 'git blame" :) | 14:44 |
mariojv | i do use a plugin manager that included it by default | 14:44 |
* jlvillal loves Gblame with vim-fugitive | 14:44 | |
lucasagomes | mat128, btw, was suppose to as you. Are you planning on having virtualPDU in the gate and keeping the snmp driver in tree ? | 14:45 |
vsaienk0 | lucasagomes: how long it will take to boot that image with 350MB ram? | 14:45 |
mat128 | lucasagomes: yup | 14:45 |
lucasagomes | mat128, awesome | 14:45 |
sambetts | jlvillal: I have moved the topology from my devstack patch into the actual dev docs | 14:45 |
mat128 | lucasagomes: we fixed the last bugs, then we will introduce -nv jobs on it, then bring virtualpdu under ironic governance | 14:45 |
jlvillal | sambetts, Awesome. I do like that diagram and hated to think it might disappear | 14:46 |
lucasagomes | mat128, gotcha, there was a patch for adding supporting for a new snmp model | 14:46 |
mat128 | lucasagomes: yeah, all of this work is from friends at Internap | 14:46 |
lucasagomes | I was unsure whether we should cause it's atm marked as an unsupported | 14:46 |
lucasagomes | mat128, oh right | 14:46 |
lucasagomes | vsaienk0, hmmm I will deploy again and see, it should be OK (def slowed than cirros) | 14:46 |
lucasagomes | but it's just a xenial cloud image | 14:46 |
*** vsaienko has quit IRC | 14:47 | |
sambetts | how big is that raw? | 14:48 |
*** vsaienko has joined #openstack-ironic | 14:48 | |
lucasagomes | sambetts, stack@ubuntu-Standard-PC-i440FX-PIIX-1996:~/devstack$ du -sh xenial-server-cloudimg-amd64-disk1.img.raw | 14:48 |
lucasagomes | 972M xenial-server-cloudimg-amd64-disk1.img.raw | 14:48 |
sambetts | oooo wow mega big | 14:48 |
lucasagomes | right | 14:48 |
mgould | morning rloo mat128 jlvillal mariojv | 14:48 |
openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: Retry connecting to vm after power-on https://review.openstack.org/394901 | 14:48 |
sambetts | what is even in that iamge to make it that big | 14:48 |
jlvillal | Good morning :) | 14:49 |
lucasagomes | sambetts, hah idk I just got it from https://cloud-images.ubuntu.com/xenial/current/ | 14:49 |
lucasagomes | for testing | 14:49 |
mariojv | o/ mgould | 14:49 |
mat128 | rloo: Walking back the wiki history, the "notes" numbers were shifted | 14:49 |
mat128 | rloo: Devananda had noted "this feature does not make sense for baremetal" on the migration, and it got shifted to "enabled by some ironic drivers" | 14:50 |
*** jcoufal has joined #openstack-ironic | 14:50 | |
rloo | mat128: ha ha. good detective work, thx! | 14:51 |
rloo | mat128: you could mention that in the patch | 14:51 |
mat128 | yes I'm updating it | 14:52 |
mat128 | https://wiki.openstack.org/w/index.php?title=HypervisorSupportMatrix&diff=63839&oldid=62351 | 14:52 |
mgould | dtantsur: done | 14:52 |
mat128 | rloo: done | 14:53 |
*** derekh has joined #openstack-ironic | 14:53 | |
vdrok | morning rloo mariojv jlvillal mat128 ! | 14:53 |
jlvillal | :) | 14:54 |
rloo | hi vdrok, mgould, jlvillal | 14:55 |
jlvillal | :) | 14:55 |
*** wajdi has joined #openstack-ironic | 14:55 | |
*** makowals has quit IRC | 14:55 | |
jroll | dtantsur: friend recommended "NullConsoleDriver or NoConsoleDriver or MissingConsoleDriver or AbsentConsoleDriver or something like that" | 14:55 |
jroll | wdyt about those? | 14:55 |
JayF | I like "NoConsoleDriver" | 14:56 |
mat128 | jroll, dtantsur: I like NullConsoleDriver tbh | 14:56 |
JayF | I like that name simply because of how silly someone will sound if they complain about their console not working | 14:56 |
jroll | mat128: I feel like Null is more for devs | 14:56 |
jroll | heh | 14:57 |
JayF | "I'm using the NoConsoleDriver but for some reason I have no console" | 14:57 |
mat128 | this ^ ++ | 14:57 |
mat128 | I change my vote | 14:57 |
sambetts | ++ | 14:57 |
*** makowals has joined #openstack-ironic | 14:57 | |
mat128 | jroll: agreed | 14:57 |
jroll | dtantsur: ^^ | 14:57 |
jroll | I agree | 14:57 |
dtantsur | NullConsoleDriver has a better shortening though | 14:57 |
jroll | shortening? | 14:57 |
dtantsur | well, we can short NoConsoleDriver as no-console | 14:57 |
dtantsur | sorry, entrypoint name | 14:57 |
jroll | ah | 14:57 |
* dtantsur has no idea what he's typing | 14:57 | |
jroll | I don't hate 'no' for the entrypoint | 14:58 |
mat128 | having the operator use the "no-console" console interface is perfectly logical IMO | 14:58 |
mat128 | if they want to disable the console | 14:58 |
mat128 | they will understand that their action will disable it | 14:58 |
jroll | no_console, because entrypoints can't have - :/ | 14:58 |
mat128 | meh | 14:58 |
jroll | but yeah | 14:58 |
mat128 | wtv ;) | 14:58 |
jroll | :D | 14:58 |
mgould | yeah, NoConsoleDriver distinguishes between "there deliberately isn't one" and "we forgot to set one" | 14:58 |
* jroll gets pedantic sometimes | 14:58 | |
*** mgoddard_ has joined #openstack-ironic | 15:01 | |
*** Linkmark has quit IRC | 15:02 | |
openstackgerrit | Merged openstack/ironic-python-agent: Fix several errors in LLDP handling code https://review.openstack.org/396043 | 15:02 |
dtantsur | jroll, they can iirc | 15:02 |
jroll | oh? | 15:02 |
jroll | TIL | 15:02 |
jroll | I seem to remember they couldn't | 15:03 |
dtantsur | jroll, well, this seems to work https://github.com/openstack/ironic-inspector/blob/master/setup.cfg#L54-L55 | 15:03 |
mat128 | dtantsur, jroll: I can confirm, "docker-compose" is a console entry point | 15:03 |
jroll | huh, ok | 15:03 |
mat128 | and it has one | 15:03 |
* jroll wonders what he was missing then | 15:03 | |
dtantsur | right, console scripts also :) https://github.com/openstack/ironic-inspector/blob/master/setup.cfg#L23-L25 | 15:03 |
* dtantsur will submit a (yet another) updated to the driver composition spec | 15:04 | |
dtantsur | I wonder how many of ^^^ we're going to have before we call it done? :) | 15:04 |
* jroll hates updating specs | 15:04 | |
jroll | I didn't realize that was in the spec, though | 15:04 |
* dtantsur doesn't quite like writing specs, updating specs and reviewing specs | 15:04 | |
*** mgoddard has quit IRC | 15:04 | |
jroll | it isn't, other than NoopInspect | 15:04 |
jroll | used as an example | 15:05 |
jlvillal | dtantsur, s/doesn't quite like/loves/ :) | 15:05 |
dtantsur | jroll, we do mention it here http://specs.openstack.org/openstack/ironic-specs/specs/approved/driver-composition-reform.html#hardware-types | 15:05 |
dtantsur | but yeah, if nobody insists, I won't update it :) | 15:05 |
* jlvillal gets ready to go to the office | 15:05 | |
jroll | dtantsur: yeah, I just read that as an example | 15:06 |
jroll | I wouldn't worry about it | 15:06 |
dtantsur | ack | 15:06 |
* dtantsur -> short break | 15:06 | |
lucasagomes | vsaienk0, it's super slow to boot :-( | 15:07 |
lucasagomes | vsaienk0, ubuntu@testing:~$ systemd-analyze | 15:07 |
lucasagomes | Startup finished in 31.517s (kernel) + 5min 57.757s (userspace) = 6min 29.274s | 15:07 |
*** roeyc has quit IRC | 15:07 | |
rloo | anyone else want to review attach/detach spec? Otherwise I'll +A https://review.openstack.org/#/c/317636/ | 15:07 |
patchbot | patch 317636 - ironic-specs - Add spec for interface attach and detach API | 15:07 |
vsaienk0 | lucasagomes: we still need to have more RAM for normal operations | 15:08 |
jroll | rloo++ | 15:08 |
JayF | rloo: I'll take a quick look then land it | 15:08 |
rloo | thx JayF | 15:08 |
lucasagomes | vsaienk0, what would be the minimal acceptable you think ? 512 ? | 15:08 |
jroll | JayF: ping me when you do land that so I can bug nova :D | 15:08 |
vsaienk0 | lucasagomes: with 512 it should be much faster I think | 15:08 |
lucasagomes | vsaienk0, but would 512 be acceptable for the multiple VMs jobs ? | 15:09 |
vsaienk0 | jroll: could you please review nova portgroup spec https://review.openstack.org/#/c/387534/ | 15:09 |
patchbot | patch 387534 - nova-specs - Proposal for ironic static portgroups support | 15:09 |
*** godp1301 has quit IRC | 15:09 | |
* lucasagomes will check how much it takes to boot on 512 | 15:09 | |
JayF | that attach/detach spec is one of the nicest, most complete ones I've read | 15:10 |
JayF | A++ | 15:10 |
sambetts | :D thanks | 15:10 |
jroll | vsaienk0: it's on my list | 15:10 |
* jroll looks quikcly | 15:10 | |
JayF | jroll: cleared for landing | 15:11 |
*** yolanda has quit IRC | 15:11 | |
rloo | vsaienk0: you beat me to it. please look jroll :) | 15:11 |
rloo | JayF: should we have a 'model' spec award? | 15:11 |
JayF | If we have a devdocs page about writing specs, linking a few model specs would not be a bad idea | 15:12 |
JayF | I know a team I was on at Rackspace did that with PRs, describe what a good PR looked like (with a long description and such) and linked several examples | 15:12 |
jroll | sambetts: sanity check me on the portgroup nova spec https://review.openstack.org/#/c/387534/3 | 15:13 |
patchbot | patch 387534 - nova-specs - Proposal for ironic static portgroups support | 15:13 |
*** godp1301 has joined #openstack-ironic | 15:13 | |
jroll | sambetts: the attach/detach thing just turns that into "needs metadata", yeah? | 15:13 |
rloo | JayF: maybe here? | 15:13 |
rloo | JayF: http://docs.openstack.org/developer/ironic/dev/code-contribution-guide.html#adding-new-features | 15:13 |
JayF | rloo: I would think we could have a separaet entry on specs, that link up to there | 15:14 |
jroll | vsaienk0: reviewed, btw, see ^ my questions for sam | 15:14 |
rloo | JayF: or http://docs.openstack.org/developer/ironic/dev/code-contribution-guide.html#ironic-specs-process | 15:14 |
JayF | that's more what I was thinking | 15:14 |
openstackgerrit | Merged openstack/ironic-specs: Add spec for interface attach and detach API https://review.openstack.org/317636 | 15:14 |
JayF | I'll take a note to do that | 15:14 |
sambetts | jroll: yeah I believe we'd just need to populate the configdrive and thats it | 15:14 |
jroll | \o/ | 15:14 |
rloo | thx JayF. | 15:14 |
JayF | link that 2nd specs-process back to adding-new-features and then add some links to good model specs | 15:14 |
JayF | proably monday though. I go down the meetnig rabbithole in 15 minutes and I have tomorrow off :D | 15:15 |
rloo | JayF: sounds good. now we can bikeshed on good model specs :D | 15:15 |
rloo | JayF: no hurry | 15:15 |
sambetts | jroll: although I thought we decided that we were going to push for PG before attach/detach? | 15:15 |
JayF | lol that's exactly what I was thinking | 15:15 |
rloo | JayF: nice that you have tomorrow off! Enjoy! | 15:15 |
jroll | sambetts: oh, I thought it was the opposite :/ I could be wrong | 15:16 |
jroll | but | 15:16 |
JayF | rloo: well, my wife is a teacher, she gets veterans day off, and not much choice as to what days she gets off, so we're going to go to a museum | 15:16 |
rloo | I thought pg had to be done first before attach/detach? | 15:16 |
jroll | sambetts: it's so much easier with attach/detach, I wonder what our confidence is on getting both done in ocata | 15:16 |
rloo | but i thought we were going to do both so... | 15:16 |
rloo | JayF: ah, right. So she gets the summer off, you take that off too? ;) | 15:17 |
JayF | rloo: :( working from home while your partner has months off is particularly awful, lol | 15:17 |
sambetts | jroll, rloo: I think it should be pretty straight forward the hard part is landing the spec for nova before they freeze | 15:17 |
vdrok | sambetts: jroll: by the way, regarding configdrive generation, currently some of the parameters are hardcoded - https://review.openstack.org/#/c/388756/2/nova/virt/ironic/driver.py@712, ideally thos should be added as fields in the portgroup right? | 15:17 |
patchbot | patch 388756 - nova - Generate necessary network metadata for portgroups | 15:17 |
jroll | sambetts: if the spec is just 'generate configdrive stuff' it should be quick, right? | 15:18 |
sambetts | vdrok: Yes, things like the hash mode etc would be nice to define on the PG in ironic | 15:18 |
rloo | sambetts: jroll is going to do that hard part ;) | 15:18 |
sambetts | jroll: yeah I'd hope so, we have until the 17th right? | 15:18 |
jroll | sambetts: yeah, but the sooner the better | 15:18 |
sambetts | ++ | 15:18 |
rloo | ++ | 15:19 |
jroll | vsaienk0: are you good with nova's portgroups stuff relying on attach/detach? | 15:19 |
jroll | sambetts: that means you get to handle how these map to ports/portgroups/both | 15:19 |
jroll | :D | 15:19 |
vsaienk0 | jroll: yeah they are dependent | 15:19 |
jroll | vsaienk0: you're good with attach/detach first, right? | 15:19 |
vsaienk0 | jroll: yes, it will be better to rely on that, I will update nova patch | 15:20 |
jroll | and everyone, that means we need to get attach/detach done early this cycle | 15:20 |
jroll | vsaienk0: awesome, thanks :) | 15:20 |
*** rbartal has quit IRC | 15:20 | |
*** jaosorior has quit IRC | 15:20 | |
rloo | sambetts: what's the status of the ironic code for attach/detach? | 15:21 |
vsaienk0 | jroll, sambetts: we may generate network_metadata on ironic side also | 15:21 |
jroll | vsaienk0: and return it to nova? dunno if it's worth it | 15:21 |
jroll | hrm, GET nodes/ident/vifs/vif_id/config | 15:21 |
jroll | heh | 15:21 |
* jroll resists rabbit holing | 15:22 | |
*** galyna has quit IRC | 15:22 | |
vdrok | :) | 15:22 |
jroll | that's an interesting idea though... | 15:22 |
sambetts | jroll: the main reason I see for doing it is that otherwise nova has to inspect the ports looking for vif_port_ids or whatever tying it back to the implementation again | 15:22 |
*** galyna has joined #openstack-ironic | 15:22 | |
sambetts | jroll: but I don't think we have an API for it right now | 15:23 |
jroll | sambetts: don't think it does... it just needs macs yeah? | 15:23 |
jroll | or, I guess it needs to know which mac maps to which port, huh | 15:23 |
jroll | because IP address | 15:23 |
jroll | wow, this just got hard :| | 15:23 |
sambetts | rloo: mostly done, I need to finish updating it with the API changes though | 15:24 |
rloo | sambetts: ok, so maybe next week we can start reviewing. after you all agree on ^^ too :) | 15:25 |
sambetts | rloo: yeah that would be awesome :) | 15:25 |
jroll | sambetts: so yeah, in the new attach/detach world, how do we do configdrive, does nova just get that info from neutron? | 15:26 |
*** yolanda has joined #openstack-ironic | 15:26 | |
rloo | sambetts: that's the Royal We. I will plan on it :) | 15:26 |
vdrok | maybe 'POST nodes/ident/network_config -d some_metadata_from_nova', it will update it and return back to nova, and nova puts it into configdrive? | 15:26 |
jroll | and then for portgroups, it would have the mac for the portgroup, so it would need to ask ironic about the underlying ports? | 15:26 |
sambetts | jroll: that assumes that we've done the neutron port update before that gets generated which I don't think is the case :/ | 15:27 |
jroll | erm | 15:28 |
jroll | so wat do | 15:28 |
sambetts | just thinking | 15:28 |
jroll | wait, we'd be doing the update at the same time as plug_vifs runs now, right? | 15:28 |
jroll | (or should be) | 15:28 |
sambetts | no, the update happens at the same time as the other update happens | 15:28 |
sambetts | at the configure_X_network step | 15:29 |
*** sabeen has joined #openstack-ironic | 15:29 | |
jroll | so after configdrive is written? | 15:30 |
jroll | need to figure something out then | 15:30 |
sambetts | yeah but before the tenant image is booted | 15:30 |
jroll | yeah | 15:30 |
jroll | but we need the right info in configdrive | 15:30 |
*** moshele has quit IRC | 15:31 | |
*** jerrygb has joined #openstack-ironic | 15:31 | |
lucasagomes | vsaienk0, fyi, 512MiB didn't help much | 15:31 |
*** jerrygb has quit IRC | 15:31 | |
lucasagomes | vsaienk0, http://paste.openstack.org/show/588756/ | 15:31 |
*** jerrygb has joined #openstack-ironic | 15:32 | |
*** rbrud has quit IRC | 15:33 | |
*** portdirect_ is now known as portdirect | 15:33 | |
sambetts | jroll: trying to decide if we need the vif_port_id or not, looking at the generation code here https://review.openstack.org/#/c/388756/2/nova/virt/ironic/driver.py we only use it to decide which physical ports are bound and which aren't | 15:34 |
patchbot | patch 388756 - nova - Generate necessary network metadata for portgroups | 15:34 |
vdrok | jroll: sambetts hm, we first do _plug_vifs, then _generate_configdrive, so in _generate_configdrive we can just call to ironic to ask it for updates to network metadata? | 15:34 |
*** adreznec has quit IRC | 15:34 | |
* mat128 wouldnt object to ^ | 15:34 | |
mat128 | sound good | 15:35 |
mat128 | sounds* | 15:35 |
sambetts | jroll: I was thinking about making Ironic generate it for a while, because of cases where the physical ports might not exist until deployment | 15:35 |
* sambetts need more coffee | 15:35 | |
mat128 | sambetts: you would need to be passed everything the client passed in too | 15:35 |
mat128 | metadata, etc. | 15:36 |
mat128 | sounds like a pita :( | 15:36 |
jroll | sambetts: yeah, that's fine, but we need a spec update to add nodes/ident/vifs/vif_if/config or whatever | 15:36 |
jroll | or | 15:36 |
jroll | nova needs to build it | 15:36 |
JayF | sambetts: we should avoid that at all costs if we can, not just becasue of what mat128 pointed out, but also because if we don't assemble it in exactly the same way as nova we can introduce incompatibilities | 15:36 |
mat128 | oh yes | 15:36 |
mat128 | ++ | 15:36 |
sambetts | yeah | 15:37 |
*** adreznec has joined #openstack-ironic | 15:37 | |
vdrok | JayF: but even if done in nova, we still fiddle with configdrive kind of manually, and it still can break | 15:37 |
vdrok | adding links and such | 15:38 |
*** jerrygb has quit IRC | 15:39 | |
mat128 | vdrok: we do? | 15:39 |
vdrok | mat128: that's the link - https://review.openstack.org/#/c/388756/2/nova/virt/ironic/driver.py | 15:39 |
patchbot | patch 388756 - nova - Generate necessary network metadata for portgroups | 15:39 |
mat128 | vdrok: Ironic write it on disk, but never modifies the content iirc | 15:39 |
mat128 | vdrok: thats fine because it's in nova | 15:40 |
mat128 | we have something very similar downstream | 15:40 |
*** mnaser has joined #openstack-ironic | 15:40 | |
sambetts | jroll: what if we introduced the GET /node/<ident>/vif/<ident> like we talked about earlier and that could return the mapping for that particular VIF? | 15:40 |
mnaser | does this channel cover bifrost or is it strictly ironic :3 | 15:40 |
mnaser | :> | 15:40 |
mat128 | sambetts: does that mean you would be able to create it with a mapping too? | 15:40 |
mat128 | mnaser: both | 15:40 |
jroll | sambetts: well, we need the mac address too ya | 15:41 |
mat128 | mnaser: everything under the ironic governance | 15:41 |
mat128 | and more :) | 15:41 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Configure nodes/environment to boot in UEFI mode https://review.openstack.org/374988 | 15:41 |
mnaser | mat128: cool, we're looking at using bifrost to scale our hardware deployment for our public cloud but i'm interesting in knowing if its possible to add existing physical infra to be "managed" but not deployed | 15:41 |
mat128 | mnaser: using an Ironic driver without a "deploy" driver will probably do what you want (e.g. manage power but not deploy) | 15:42 |
mnaser | so we have a whole bunch of new servers we want to deploy as well as a few that we want to set as "managed" by bifrost so we can redeploy later (or for consistency sake) | 15:42 |
mat128 | mnaser: look into the adopt stuff | 15:42 |
mat128 | let me get the docs for you | 15:42 |
mnaser | mat128 : http://docs.openstack.org/developer/ironic/deploy/adoption.html ? :P | 15:42 |
mat128 | this, yes :) | 15:42 |
jroll | yep, that's exactly what this is for | 15:42 |
sambetts | jroll, mat128: what about if we did something like GET /node/<ident>/vif/<ident> returns the mac address that its been mapped too, then you can follow it up with a GET /ports?address= or /portgroups?address ? | 15:43 |
*** Nisha_away has joined #openstack-ironic | 15:43 | |
sambetts | ;/ | 15:43 |
mat128 | sambetts: where is the MAC-VIF mapping stored? | 15:44 |
mat128 | sambetts: and if GET returns it, I would expect the create call would also take it | 15:44 |
lucasagomes | jroll, dtantsur if you have time, mind taking a peek at the patch switching our jobs to xenial/ipmitool: https://review.openstack.org/#/c/392890/ ? | 15:44 |
patchbot | patch 392890 - openstack-infra/project-config - Switch ironic-ipa jobs to Xenial and make ipmitool... | 15:44 |
sambetts | mat128: its not stored its produced by a function in the network interface | 15:44 |
lucasagomes | I think the patch is good now, but just to make sure | 15:44 |
mat128 | which calls neutron? | 15:44 |
sambetts | mat128: hmmm well it sort of does, the mac address is on the neutron port we pass in | 15:44 |
mat128 | Nova could get the VIF list, then query neutron about it? | 15:45 |
mat128 | but it breaks non-neutron setups | 15:45 |
* mat128 shrugs | 15:45 | |
jroll | sambetts: well, yeah, that would work, but is it the best way | 15:45 |
mat128 | brb, coffee | 15:46 |
sambetts | jroll: basically we need a standard way to get the mapping back out of ironic right?, We can list the attached vifs, but we can't see what they are attached too | 15:48 |
sambetts | jroll: and they might not be attached to anything yet, if its a flex-nic card that generates the NICs on the fly | 15:48 |
dtantsur | lucasagomes, will do | 15:49 |
lucasagomes | ty | 15:49 |
sambetts | jroll: so we need the network interfaces to be able to say I have this VIF and it will become this mac-address or whatever | 15:49 |
*** mgoddard_ has quit IRC | 15:50 | |
sambetts | jroll: or in the case of trunked vifs, we need to be able to indicate they are all on a single interface | 15:51 |
*** syed__ has joined #openstack-ironic | 15:51 | |
*** vsaienko has quit IRC | 15:51 | |
*** vsaienko has joined #openstack-ironic | 15:52 | |
jroll | sambetts: sorry, meeting'd :/ | 15:52 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Create dummy interfaces for use with hardware types https://review.openstack.org/393867 | 15:52 |
dtantsur | updated ^^^^ | 15:52 |
jroll | sambetts: tl;dr before nova tells ironic to deploy the thing, we need to have the configdrive properly sorted | 15:52 |
sambetts | jroll: yup, which makes it hard for hardware that changes during deployment because we need to know what the final state is before we do anything | 15:54 |
jroll | sambetts: seems like those magic nics should be created at plug_vifs time, then | 15:54 |
*** Guest40216 has quit IRC | 15:56 | |
sambetts | jroll: yeah potentially | 15:57 |
*** mgoddard has joined #openstack-ironic | 15:59 | |
*** rama_y has quit IRC | 15:59 | |
*** rbrud has joined #openstack-ironic | 16:00 | |
openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Introducing node introspection state management https://review.openstack.org/348943 | 16:00 |
*** cdaniel has joined #openstack-ironic | 16:00 | |
*** rajinir has joined #openstack-ironic | 16:02 | |
*** BLOODBATH has joined #openstack-ironic | 16:02 | |
*** BLOODBATH is now known as Guest13773 | 16:03 | |
milan | dtantsur, sambetts I've updated https://review.openstack.org/#/c/395609/ if you guys feel like a review, it's a short one ;) | 16:03 |
patchbot | patch 395609 - python-ironic-inspector-client - UUID, started_at, finished_at in the status | 16:03 |
dtantsur | k | 16:03 |
*** jkilpatr_ has joined #openstack-ironic | 16:03 | |
rajinir | mjturek: Do you have any additional info on the Ironic image configuration for devstack-gate? | 16:04 |
*** baoli has joined #openstack-ironic | 16:04 | |
*** vsaienko has quit IRC | 16:05 | |
vsaienk0 | jroll, dtantsur, jlvillal, lucasagomes and other bash experts please help to review multinode patches https://review.openstack.org/#/q/topic:bug/1622653 (only 6 patches to ironic are left, they are not so big, and 1 to devstack-gate) | 16:06 |
* milan could use a coffee | 16:06 | |
*** baoli_ has quit IRC | 16:07 | |
lucasagomes | vsaienk0, will take a look, but I'm far for being a bash expert heh | 16:07 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Abort cleaning and deployment if node is in maintenance https://review.openstack.org/366828 | 16:07 |
vsaienk0 | sambetts ^ | 16:08 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: DNM: Test patch to verify bond https://review.openstack.org/388660 | 16:08 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ironic devstack portgroup support https://review.openstack.org/381743 | 16:08 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Rely on portgroup standalone_ports_supported https://review.openstack.org/360647 | 16:08 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: Tempest tests for portgroups https://review.openstack.org/382476 | 16:08 |
mjturek | rajinir: not written up. I can get you our disk image builder commands though | 16:08 |
rajinir | mjturek : That will help | 16:09 |
*** links has joined #openstack-ironic | 16:09 | |
JayF | dtantsur: that patch, 366828, has a potentially unintended side effect | 16:10 |
rajinir | https://www.irccloud.com/pastebin/bNXgRke3/ironic-image | 16:10 |
* jlvillal grumbles that irccloud requires javascript | 16:10 | |
JayF | dtantsur: imagine I have a dozen cleaning steps, after step 7, the BMC switch goes out and maintenances the node. today, when I unmaint that node, it'll start back cleaning where it left off. After your change, it'll have to start back over from the beginning | 16:11 |
jlvillal | Ah it does allow seeing the "raw" version | 16:11 |
*** rcernin has quit IRC | 16:11 | |
lucasagomes | vsaienk0, why the mtu has to be 1400 for the vm bridge ? /me remembers we had to set mtu to 1450 for the public bridge awhile ago | 16:12 |
*** yufei has joined #openstack-ironic | 16:12 | |
dtantsur | JayF, iirc for some reason it does not continue cleaning after getting back.. | 16:12 |
dtantsur | it was quite some time ago though, I may not remember precisely | 16:12 |
JayF | dtantsur: as long as you left the agent running and heartbeating it does | 16:12 |
vsaienk0 | lucasagomes, because on multinode job we need to take into account additional 50 bytes for vxlan header | 16:13 |
JayF | dtantsur: to be clear, I'm probably OK with that, but it does suck a lot for a large deployer with lots of clean steps -- basically some temporary network issue could have a denial of service effect by causing long-running clean steps to have to repeat | 16:13 |
dtantsur | hmm | 16:14 |
* dtantsur tries to remember why cleaning did not continue in his case.. | 16:14 | |
lucasagomes | vsaienk0, gotcha | 16:14 |
JayF | dtantsur: my knowledge is from downstream, where we run with dhcp_provider=None, so there could be headahce somewhere in there around booting an agent | 16:15 |
JayF | dtantsur: whereas in our environment the agent can boot anytime the node is powered on in the right network | 16:15 |
dtantsur | well, I think I just found it confusing that it stays in clean wait forever, until you figure out it's due to maintenance | 16:15 |
JayF | Absolutely, like I said, I'm probably OK with the change in behavior | 16:15 |
JayF | I just wanted to make sure you were aware of the negative side effect of the change | 16:16 |
JayF | I just know anything that requires a bunch of machines to re-clean, when you have an extensive cleaning process, is really obnoxious :) | 16:16 |
dtantsur | JayF, should we make it configurable then? | 16:17 |
*** rpioso has joined #openstack-ironic | 16:18 | |
*** [1]cdearborn has joined #openstack-ironic | 16:18 | |
milan | https://review.openstack.org/#/c/383862/ got merged! \o/ | 16:18 |
patchbot | patch 383862 - api-wg - Specify time intervals based filtering queries (MERGED) | 16:18 |
JayF | dtantsur: Eh. I just think it means we need to think hard about the anomoly resolution stuff I'm designing. For instance, in that case, if Ironic kicked a node into a bmc fault, I'd also want ironic to recover it from the associated clean fail, but I think that's easy enough | 16:19 |
JayF | dtantsur: if anything maybe indicates a potential need for the ability to "resume" cleaning on a clean fail vs restarting | 16:19 |
dtantsur | this is a good catch, yeah | 16:20 |
JayF | Madasi: ^ what would you think about that? The ability to `ironic node-set-provision-state NODE resume` a cleanfail node, resuming from the step that failed rather than restarting cleaning from scratch | 16:22 |
dtantsur | I'd rather see 'clean --continue' to avoid a new provision action | 16:22 |
JayF | clean is used only for manual cleaning | 16:22 |
dtantsur | or 'provide --continue-cleaning' | 16:22 |
dtantsur | :) | 16:22 |
sambetts | mat128: what do you think of this https://etherpad.openstack.org/p/network-interface-vifs-configdrive ? | 16:22 |
JayF | automated cleaning is started using manage/provide | 16:22 |
JayF | dtantsur: yeah, I'm not an API design guy, and don't care about the details :P | 16:23 |
JayF | dtantsur: well, I care that the details are good. I just make no claim to knowing what is good in an api, lol | 16:23 |
dtantsur | anyway, this makes sense for people with a lot of steps | 16:23 |
openstackgerrit | Vasyl Saienko proposed openstack/ironic: SUPER WIP: Run all tests in one job. https://review.openstack.org/396155 | 16:23 |
vsaienk0 | jroll, sambetts: do we have a spec to nova related to attach/deatach changes? | 16:25 |
jroll | vsaienk0: we have a BP which is now approved without a spec | 16:25 |
jroll | vsaienk0: https://blueprints.launchpad.net/nova/+spec/ironic-plug-unplug-vifs-update | 16:26 |
vsaienk0 | jroll: wow, that is nice | 16:27 |
yufei | hi, JayF, could you please help review this doc modification when you are free? https://review.openstack.org/#/c/387082/ | 16:27 |
jroll | :) | 16:27 |
patchbot | patch 387082 - ironic - Add the way to get the deploy ram disks | 16:27 |
JayF | vsaienk0: nova devs <3 jroll, they're trying to sucker him into working on that instead ;) | 16:27 |
JayF | /s | 16:27 |
jroll | ಠ_ಠ | 16:27 |
yufei | thank you | 16:28 |
vsaienk0 | :) | 16:28 |
NobodyCam | Good Morning Ironic'ers | 16:30 |
*** harlowja has joined #openstack-ironic | 16:31 | |
*** harlowja has quit IRC | 16:32 | |
*** rcernin has joined #openstack-ironic | 16:32 | |
*** harlowja has joined #openstack-ironic | 16:34 | |
vdrok | morning NobodyCam :) | 16:35 |
openstackgerrit | Aline Bousquet proposed openstack/ironic: Add possibility to remove chassis_uuid from a node https://review.openstack.org/390558 | 16:35 |
*** links has quit IRC | 16:37 | |
NobodyCam | :) | 16:37 |
*** baoli has quit IRC | 16:37 | |
dtantsur | morning NobodyCam | 16:39 |
NobodyCam | morning dtantsur :) | 16:40 |
lucasagomes | NobodyCam, sup! | 16:43 |
openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Abort cleaning and deployment if node is in maintenance https://review.openstack.org/366828 | 16:43 |
dtantsur | JayF, first attempt at proper wording ^^^ :) | 16:43 |
NobodyCam | morning lucasagomes :) just back home after a unexpected trip to Omaha :) | 16:45 |
lucasagomes | NobodyCam, ah nice, home's good. I spent ~1 month away and came back 2 weeks ago... Glad to be back | 16:47 |
NobodyCam | yes! Home is Good :) | 16:47 |
yufei | Hello,ironicers, when I try to follow this document(http://docs.openstack.org/developer/ironic/deploy/multitenancy.html) to configure my multi-tenant network ironic at newton version, I find that when I want to create a bare metal instance from nova, nova create a interface with port mac address for it, then ironic asks neutron to create a port with same mac, so i get a mac conflict error, I’m not sure whether there is any configuration for | 16:50 |
yufei | this, I try to google but find no messages for this,does someone has meet the same problem or has any suggestion for me?Thanks a lot. | 16:50 |
*** ccamacho has quit IRC | 16:51 | |
sambetts | yufei: are you trying to deploy your instance into your provisioning network? | 16:51 |
*** mgoddard_ has joined #openstack-ironic | 16:51 | |
yufei | yes | 16:51 |
yufei | that’s not supported? | 16:51 |
*** cdaniel has quit IRC | 16:52 | |
yufei | Oh, thanks a lot! | 16:52 |
sambetts | yufei: that isn't supported, the node should be deployed into a separate network from the provisioning network, nodes should only be connected to the provisioning network by Ironic during provisioning and then moved to a separate network once its completed :) | 16:53 |
*** cdaniel has joined #openstack-ironic | 16:53 | |
*** mgoddard has quit IRC | 16:54 | |
yufei | Thanks a lot, I have spent hours on this. If this is not methoned in docs now, I think I can push a patch for this. | 16:56 |
*** ohamada has quit IRC | 16:56 | |
sambetts | yufei: I think we added something to the docs about it recently | 16:56 |
sambetts | vsaienk0, jroll: Do you remember? ^ | 16:56 |
vsaienk0 | sambetts: i've opened a bug for this | 16:57 |
Madasi | JayF: right chan, I like it :) | 16:57 |
vsaienk0 | sambetts, jroll: we didn't decided if we should allow to build instances in provisioning network https://bugs.launchpad.net/ironic/+bug/1634573 | 16:58 |
openstack | Launchpad bug 1634573 in Ironic "Allow to spawn instances in provisioning network" [Undecided,New] - Assigned to Vasyl Saienko (vsaienko) | 16:58 |
vsaienk0 | sambetts, I've opened to any option, allow or do not allow. But I think we shouldn't allow spawn instances in provisioning network, we receive flat network case than | 17:00 |
*** sabeen has quit IRC | 17:02 | |
*** chuckC_ has joined #openstack-ironic | 17:03 | |
yufei | I agree that this feature is not worthy of being supported | 17:03 |
jlvillal | Agree with documenting current situation. If decide to change in future then update documentation. | 17:03 |
sambetts | ++ | 17:03 |
sambetts | mat128, jroll: I think if we ensure that the network interfaces update the mac address information in neutron at time of _plug_vifs then we should be OK to just handle things this way: https://etherpad.openstack.org/p/network-interface-vifs-configdrive | 17:05 |
*** jpich has quit IRC | 17:06 | |
dtantsur | o/ | 17:07 |
*** dtantsur is now known as dtantsur|afk | 17:07 | |
*** rama_y has joined #openstack-ironic | 17:07 | |
NobodyCam | have a good night dtantsur|afk :0 | 17:07 |
sambetts | jroll: basically instead of checking/having to return a mapping, the code is rather just checking if the VIF is or is not a bond by cross checking it with the port groups list | 17:07 |
NobodyCam | s/0/)/ | 17:07 |
sambetts | night dtantsur|afk | 17:07 |
*** ccamacho has joined #openstack-ironic | 17:08 | |
*** sabeen has joined #openstack-ironic | 17:08 | |
mat128 | sambetts: checking | 17:09 |
*** eroux has quit IRC | 17:10 | |
*** mgoddard has joined #openstack-ironic | 17:11 | |
mat128 | sambetts: fixed a typo (pg vs portgroup) but lgtm | 17:11 |
mat128 | sambetts: this will work if plug_vifs does its job | 17:11 |
sambetts | yeah, I think its the only correct way to do it | 17:11 |
*** mgoddard_ has quit IRC | 17:11 | |
*** eroux has joined #openstack-ironic | 17:13 | |
sambetts | ideally we shouldn't have to query the /vifs end point, ideally we shouldn't have to query the /ports or /portgroups endpoints either, but as bonds are a model not supported by neutron we can't rely on using their datastrcutures to how the information for that | 17:13 |
*** strigazi is now known as strigazi_AFK | 17:18 | |
*** _milan_ has joined #openstack-ironic | 17:19 | |
*** dougs1 has joined #openstack-ironic | 17:19 | |
openstackgerrit | Merged openstack/virtualbmc: Updated from global requirements https://review.openstack.org/389959 | 17:19 |
dougs1 | Anyone know if it's safe to set disable notify_nova_on_port_status_changes to false in neutron.conf when using Ironic exclusively with Nova? It seems to work, but not sure if i missed something. | 17:21 |
*** milan has quit IRC | 17:22 | |
sambetts | dougs1: it should be because we don't support nova's interface attach and detach functions post deployment which is what is triggered by the neutron nova notifier | 17:22 |
sambetts | however I've not tried it personally | 17:22 |
dougs1 | sambetts - thanks, that makes sense | 17:24 |
openstackgerrit | Merged openstack/ironic: Fix typo in release note filename https://review.openstack.org/389629 | 17:24 |
*** yee379 has quit IRC | 17:24 | |
*** skat has quit IRC | 17:26 | |
dougs1 | sambetts - I noticed the notification also gets triggered at the start of an Ironic deploy. I'll have a read of the code. | 17:26 |
*** baoli has joined #openstack-ironic | 17:29 | |
sambetts | dougs1: that is likely because of the creation of the neutron ports by nova as part of the deployment process, and if your using ironic + multitenancy when we update the neutron port binding infomation during provisioning might affect it too | 17:30 |
dougs1 | sambetts: I'll look forward to the day when we can use that multitenancy functionality. We're still on kilo :) | 17:31 |
sambetts | dougs1: :) ah ok then it'll be because of the nova port creation | 17:31 |
*** _milan_ has quit IRC | 17:32 | |
dougs1 | alright, thanks for your help | 17:32 |
sambetts | :) | 17:32 |
*** causten has quit IRC | 17:32 | |
*** ccamacho has quit IRC | 17:34 | |
*** rbrud has quit IRC | 17:35 | |
*** iurygregory has joined #openstack-ironic | 17:35 | |
*** baoli has quit IRC | 17:36 | |
*** ccamacho has joined #openstack-ironic | 17:36 | |
vdrok | sambetts: so it seems we're kinda decided with configdrive on that etherpad? I'll rebase it on top of your attach/detach stuff and change as suggested | 17:39 |
sambetts | vdrok: I think that makes the most sense, what do you think? | 17:39 |
vdrok | well, yeah, the code on the etherpad is clear, and by reusing attach/detach we get rid of a bunch of code, so I'm all for it :) | 17:40 |
sambetts | :D | 17:40 |
*** baoli has joined #openstack-ironic | 17:40 | |
*** lucasagomes is now known as lucas-afk | 17:41 | |
lucas-afk | have a great evening all | 17:41 |
sambetts | o/ lucas-afk | 17:41 |
lucas-afk | calling it a day | 17:41 |
vdrok | night lucas-afk | 17:42 |
JayF | vdrok: sambetts: I'm not sure I grok how the code path works for non-pg? | 17:44 |
mat128 | JayF: oh that was you :) | 17:44 |
mat128 | just found out | 17:44 |
*** aNuposic has joined #openstack-ironic | 17:44 | |
JayF | I'm always white-on-black in etherpads, for future reference :P | 17:44 |
sambetts | JayF: Non-PG doesn't need any modifications to the metadata | 17:45 |
mat128 | heh, my color keeps changing for reasons I can't explain | 17:45 |
*** trown is now known as trown|lunch | 17:45 | |
JayF | sambetts: ohh, so the base_metadata= just gets returned unmodified in the end | 17:45 |
mat128 | we also have etherpad downstream so I think maybe it's conflicting | 17:45 |
mat128 | we also have etherpad downstream so I think maybe it's conflicting | 17:45 |
sambetts | JayF: yup | 17:45 |
JayF | sambetts: ok, cool | 17:46 |
JayF | this seems fairly sane to me then | 17:46 |
sambetts | JayF: similarly for mixed PG and non-PG environments only the links that match PGs will get changed and the others will remain unchanged | 17:46 |
*** godp1301 has quit IRC | 17:47 | |
sambetts | JayF: we might need to update the mac addresses actually :/ because they might have been updated by Ironic | 17:47 |
sambetts | JayF: thats better | 17:48 |
vdrok | sambetts: nope, we're safe I think | 17:50 |
vdrok | sambetts: nova puts instance_uuid on a node before doing plug vifs and configdrive generation | 17:50 |
vdrok | we can't update address on a node with instance_uuid | 17:51 |
vdrok | or you mean a period of time between build request and PATCH from nova? | 17:51 |
sambetts | vdrok: during plug vifs ironic will update the neutron port to be the right MAC address, so the network_info in that function might be out of date | 17:52 |
sambetts | vdrok: so we should refetch the neutron port and update the mac address for all he links | 17:52 |
vdrok | ah, gotcha | 17:52 |
vdrok | thanks sambetts | 17:52 |
*** causten has joined #openstack-ironic | 17:53 | |
sambetts | vdrok: the simple change of moving the link['ethernet_mac_address'] outside of the PG logic should do it like I've done on the ehterpad | 17:53 |
*** causten_ has joined #openstack-ironic | 17:54 | |
JayF | vdrok: sambetts: just thought of something that's unpleasant ... in this sample code we're using very linux-specific identifiers | 17:54 |
JayF | vdrok: sambetts: Should we make the logic more ... os-agnostic? | 17:54 |
sambetts | JayF: which part? | 17:55 |
mat128 | JayF, vdrok, sambetts: if you found out the amount of time we spent choosing "lacp" internally for bond_mode | 17:55 |
JayF | sambetts: literally the bond_mode, bond_xmit_hash_policy, etc | 17:55 |
JayF | those are the names of attributes on the bonding module in linux | 17:55 |
mat128 | I suggest we stick to one set of terminology :) | 17:55 |
*** derekh has quit IRC | 17:55 | |
vdrok | JayF: these are passed further as is in cloud-init :( | 17:55 |
JayF | and particularly if we're going to identify bond_mode by the number (4) rather than name (lacp) | 17:55 |
sambetts | JayF: thats not from us, that is from here https://specs.openstack.org/openstack/nova-specs/specs/liberty/implemented/metadata-service-network-info.html | 17:56 |
JayF | sambetts: ah, okay. So it's somewhat my fault anyway. lol | 17:56 |
*** causten__ has joined #openstack-ironic | 17:56 | |
mat128 | JayF: what abt 802.3ad | 17:56 |
mat128 | lol | 17:56 |
openstackgerrit | Julia Kreger proposed openstack/bifrost: Allow bifrost to work with selinux enabled https://review.openstack.org/393745 | 17:56 |
JayF | I think it'll be fine then, just going to suck for someone implementing this for windows/bsd/etc | 17:56 |
JayF | but at least this config style has been static in linux for a long time | 17:57 |
mat128 | JayF: we did and it works | 17:57 |
*** causten has quit IRC | 17:57 | |
mat128 | cloudbase-init reads 4 or 802.3ad and behaves correctly | 17:57 |
sambetts | JayF: hopfully cloud-init for windows will do the transltion for us ;) | 17:57 |
JayF | mat128: nice, cool | 17:57 |
mat128 | I mean, with our stuff | 17:57 |
mat128 | not exactly sure where this code is | 17:57 |
mat128 | lemme check | 17:57 |
JayF | you really don't have to, it sounds like we arleady hit "it's no big deal" with this :) | 17:58 |
*** mgoddard_ has joined #openstack-ironic | 17:58 | |
*** causten_ has quit IRC | 17:58 | |
*** baoli has quit IRC | 18:00 | |
*** causten__ has quit IRC | 18:00 | |
*** ralonsoh has quit IRC | 18:00 | |
*** mgoddard has quit IRC | 18:02 | |
JayF | dtantsur|afk: reviewed that and added some suggested wording | 18:02 |
*** baoli has joined #openstack-ironic | 18:02 | |
TheJulia | good morning everyone | 18:03 |
*** causten has joined #openstack-ironic | 18:03 | |
vdrok | morning TheJulia | 18:03 |
mat128 | JayF: found it (thank to mgagne): https://review.openstack.org/#/c/347540/ | 18:06 |
patchbot | patch 347540 - cloudbase-init - Add support for OpenStack json network metadata | 18:06 |
*** vsaienko has joined #openstack-ironic | 18:10 | |
*** rbrud has joined #openstack-ironic | 18:11 | |
mjturek | rajinir: I haven't forgotten about you! just had to finish something. Getting our DIB commands now | 18:12 |
rajinir | mjturek: Thanks no hurry | 18:12 |
*** sabeen has quit IRC | 18:12 | |
mjturek | rajinir: "disk-image-create -a ppc64el -u ubuntu ironic-agent dhcp-all-interfaces" for our deploy images | 18:13 |
mjturek | note the -a though | 18:13 |
mjturek | I'm assuming you're x86_64? | 18:13 |
rajinir | yes | 18:14 |
mjturek | then I think -a would be amd64 | 18:14 |
mjturek | rajinir: "disk-image-create -a ppc64el -u ubuntu baremetal dhcp-all-interface simple-init -o $OUTPUT_PATH" for our user image | 18:15 |
*** baoli_ has joined #openstack-ironic | 18:16 | |
*** baoli has quit IRC | 18:16 | |
rajinir | mjturek: ok | 18:16 |
mjturek | rajinir: let me know if you have any other questions | 18:17 |
rajinir | mjturek: How about the devstack yaml variables? | 18:17 |
*** godp1301 has joined #openstack-ironic | 18:18 | |
mjturek | rajinir: you mean for the jenkins yaml file?? | 18:18 |
rajinir | yes | 18:18 |
mjturek | ah, I'll get that together for you | 18:18 |
*** rama_y has quit IRC | 18:22 | |
*** baoli_ has quit IRC | 18:23 | |
mjturek | rajinir: this is what our jenkins yaml looks like for our ironic job. FYI - I've expunged some values which look like this <our_*> (just in case it confuses you) http://paste.openstack.org/show/588796/ | 18:23 |
rajinir | mjturek: Great , will look at it. Thank you | 18:25 |
*** baoli has joined #openstack-ironic | 18:27 | |
*** yufei has quit IRC | 18:32 | |
jroll | sambetts: your etherpad seems fine to me, it's a lot of API calls but I think it's a good start. and means no changes for pre-portgroups attach/detach yeah? just need to update the portgroup spec? | 18:33 |
jroll | (now reading more, it looks like you confirmed that with jay) | 18:33 |
jroll | thanks for that | 18:34 |
*** rama_y has joined #openstack-ironic | 18:34 | |
*** cdearborn has joined #openstack-ironic | 18:35 | |
* sambetts -> home | 18:36 | |
sambetts | jroll: yeah hopefully that'll all work out, we just need to make sure to do the neutron port-update for the mac_address in attach_vif | 18:37 |
*** sambetts is now known as sambetts|afk | 18:37 | |
jroll | sambetts|afk: indeed | 18:37 |
jroll | night \o | 18:37 |
sambetts|afk | o/ cya | 18:37 |
*** [1]cdearborn has quit IRC | 18:39 | |
*** Sukhdev has joined #openstack-ironic | 18:40 | |
*** makowals_ has joined #openstack-ironic | 18:41 | |
*** makowals has quit IRC | 18:44 | |
*** Romanenko_K has quit IRC | 18:44 | |
*** krtaylor has quit IRC | 18:46 | |
*** dprince has quit IRC | 18:48 | |
*** trown|lunch is now known as trown | 18:48 | |
*** Guest13773 has quit IRC | 18:51 | |
*** jkilpatr_ has quit IRC | 18:51 | |
*** athomas has quit IRC | 18:52 | |
*** jkilpatr has quit IRC | 18:52 | |
*** ChubYann has joined #openstack-ironic | 18:53 | |
openstackgerrit | Vladyslav Drok proposed openstack/ironic: Make all IronicExceptions RPC-serializable https://review.openstack.org/396355 | 18:53 |
vdrok | jroll: that should be the fix for the exceptions issue^ (not tested yet on a real ironic tho) | 18:57 |
vdrok | thanks for review mat128 :) | 18:59 |
mat128 | np | 18:59 |
vdrok | and good night everyone! | 18:59 |
mat128 | o/ | 18:59 |
jlvillal | jroll: Regarding the Nova spec for portgroups. I see https://review.openstack.org/#/c/387534/ | 19:00 |
patchbot | patch 387534 - nova-specs - Proposal for ironic static portgroups support | 19:00 |
jlvillal | jroll: And then the Nova schedule has a freeze on specs next week: https://wiki.openstack.org/wiki/Nova/Ocata_Release_Schedule | 19:00 |
jroll | jlvillal: indeed. | 19:00 |
jlvillal | vdrok: Good night | 19:01 |
*** dprince has joined #openstack-ironic | 19:03 | |
*** jkilpatr_ has joined #openstack-ironic | 19:04 | |
*** Guest13773 has joined #openstack-ironic | 19:05 | |
*** jkilpatr has joined #openstack-ironic | 19:06 | |
*** fragatina has joined #openstack-ironic | 19:08 | |
jroll | jlvillal: did you have a question or? | 19:11 |
*** krtaylor has joined #openstack-ironic | 19:23 | |
*** vsaienko has quit IRC | 19:26 | |
*** MattMan has quit IRC | 19:27 | |
*** MattMan has joined #openstack-ironic | 19:28 | |
*** MattMan has quit IRC | 19:28 | |
*** Nisha_away has quit IRC | 19:33 | |
mat128 | jlvillal: https://github.com/internap/sbtest/blob/master/sbtest.sh#L19 | 19:40 |
mat128 | :) | 19:41 |
*** jerrygb has joined #openstack-ironic | 19:43 | |
*** mgoddard__ has joined #openstack-ironic | 19:44 | |
*** mgoddard_ has quit IRC | 19:48 | |
*** openstackgerrit has quit IRC | 19:48 | |
*** openstackgerrit has joined #openstack-ironic | 19:49 | |
*** mgoddard__ has quit IRC | 19:49 | |
*** chuckC_ has quit IRC | 19:55 | |
openstackgerrit | Merged openstack/ironic-specs: Update ironic-ml2-integration spec https://review.openstack.org/393660 | 19:56 |
*** vsaienko has joined #openstack-ironic | 19:58 | |
*** ijw has joined #openstack-ironic | 20:00 | |
*** cppforlife_ has quit IRC | 20:00 | |
*** sabeen has joined #openstack-ironic | 20:02 | |
*** krtaylor has quit IRC | 20:02 | |
*** cppforlife_ has joined #openstack-ironic | 20:03 | |
*** vsaienko has quit IRC | 20:04 | |
*** baoli_ has joined #openstack-ironic | 20:06 | |
*** baoli has quit IRC | 20:07 | |
openstackgerrit | Philippe Godin proposed openstack/ironic: Adds snmp driver auto discovery for APC and baytech https://review.openstack.org/389736 | 20:09 |
openstackgerrit | Philippe Godin proposed openstack/ironic: Baytech MRP27 PDU support https://review.openstack.org/387718 | 20:09 |
*** johnthetubaguy has quit IRC | 20:10 | |
*** baoli_ has quit IRC | 20:11 | |
*** johnthetubaguy has joined #openstack-ironic | 20:11 | |
*** baoli has joined #openstack-ironic | 20:12 | |
*** krtaylor has joined #openstack-ironic | 20:15 | |
*** crinkle has quit IRC | 20:16 | |
*** chas_ has quit IRC | 20:18 | |
*** ccamacho1 has joined #openstack-ironic | 20:18 | |
jlvillal | mat128: :) | 20:20 |
jlvillal | jroll: I wasn't sure if people were aware of Nova freeze date. But I think everyone is | 20:20 |
*** ccamacho has quit IRC | 20:20 | |
jlvillal | mat128: Hopefully that helps | 20:20 |
*** godp1301 has quit IRC | 20:22 | |
*** hshiina has joined #openstack-ironic | 20:27 | |
*** aNuposic has quit IRC | 20:27 | |
rloo | jlvillal: see the subteam status report in the whiteboard, we remembered to capture that there :) | 20:28 |
*** moshele has joined #openstack-ironic | 20:29 | |
jlvillal | Thanks rloo | 20:29 |
*** baoli has quit IRC | 20:30 | |
*** baoli has joined #openstack-ironic | 20:30 | |
*** betherly is now known as betherly_ | 20:33 | |
*** betherly_ is now known as betherly__ | 20:33 | |
*** eroux has quit IRC | 20:34 | |
*** godp1301 has joined #openstack-ironic | 20:35 | |
openstackgerrit | Merged openstack/bifrost: Updated from global requirements https://review.openstack.org/395305 | 20:36 |
*** rcernin has quit IRC | 20:36 | |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 20:38 |
*** jcoufal has quit IRC | 20:38 | |
krtaylor | jroll, when you get a chance, you mentioned getting together and working on the CI requirements/docs | 20:39 |
jroll | krtaylor: yeah, my thought was we just go back and forth on reviews there | 20:40 |
jroll | and then bribe^W force^W ask someone else to +A | 20:40 |
*** jcoufal has joined #openstack-ironic | 20:40 | |
jroll | but we could continue adding to that | 20:40 |
krtaylor | oh, sorry, I thought we had agreed to get an outling merged then sections, my bad | 20:41 |
jroll | no no, we did | 20:41 |
jroll | I see your outline | 20:41 |
jroll | but if we're both here, I can review and you can change quickly | 20:41 |
jroll | and we can get it mergeable | 20:41 |
krtaylor | hm, ok, I have a couple of sections in mind too | 20:42 |
krtaylor | let me push the previous then | 20:42 |
jroll | krtaylor: no, this one is fine :) | 20:42 |
* jroll isn't speaking well | 20:43 | |
krtaylor | well I removed a couple of sections that were getting too many remarks | 20:43 |
jroll | right, so let's add those in a second patch | 20:43 |
jroll | but I have at least one change for this one | 20:43 |
krtaylor | gotcha | 20:43 |
jroll | krtaylor: two quick comments there | 20:44 |
jroll | well, first is quick, dunno about the second :) | 20:44 |
krtaylor | first done, second can be clarified, it is for the previous section head "Hardware Pool Management" | 20:46 |
krtaylor | fixing now | 20:46 |
*** vsaienko has joined #openstack-ironic | 20:46 | |
jroll | ah, so it's an alternative to using just one machine? | 20:46 |
krtaylor | no, I think solutions is a better title | 20:47 |
jroll | heh, cool | 20:47 |
krtaylor | Possible solutions maybe | 20:47 |
jroll | yeah, works for me | 20:47 |
jroll | or even just 'solutions' | 20:47 |
krtaylor | done | 20:48 |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 20:48 |
JayF | I've had recruiters recruit me to be a "Solutioneer" | 20:48 |
JayF | is that what you're solving with this solutioneering session? | 20:48 |
jroll | awesome, thanks :D | 20:48 |
JayF | :P | 20:48 |
krtaylor | hehheh, thats a great business card title | 20:48 |
jroll | JayF: now you popped your head in, you have to review it | 20:48 |
krtaylor | right? | 20:49 |
JayF | jroll: I'm technically on lunch, hence why you got the funny but not anything worthwhile | 20:49 |
jroll | krtaylor: happy to work through adding some more sections if you're up for it | 20:49 |
jroll | or collaborate on them | 20:49 |
* jroll steals JayF's fork | 20:49 | |
JayF | jroll: I'm going to -1 it | 20:49 |
krtaylor | jroll, it will really up your neg count! | 20:50 |
JayF | jroll: seriously | 20:50 |
*** mjturek has quit IRC | 20:50 | |
*** dprince has quit IRC | 20:50 | |
jroll | JayF: even if I give you your fork back? | 20:50 |
JayF | I had chicken tenders, I didn't need a fork | 20:50 |
JayF | and I'm at the "watch funny/sad internet videos about tuesday night" part of my lunch | 20:50 |
JayF | lol | 20:50 |
*** Guest13773 has quit IRC | 20:51 | |
jroll | ok | 20:51 |
*** jkilpatr has quit IRC | 20:51 | |
JayF | jroll: krtaylor: my -1 was simply -- can you put a todo in the empty secionds | 20:51 |
krtaylor | JayF, it's kinda covered in the commit message, but I'll do it if you think it helps | 20:51 |
JayF | *sections | 20:51 |
jroll | makes sense | 20:51 |
JayF | krtaylor: the people finding that page in google won't see the commit message | 20:51 |
*** jkilpatr_ has quit IRC | 20:51 | |
krtaylor | ah, well said | 20:51 |
krtaylor | how about a global TODO at the top? | 20:52 |
JayF | Sure, like a note saying the page is a work in progress and unfilled sections are gonig to be filled | 20:52 |
JayF | I'm fine with that | 20:52 |
JayF | just wanted to make sure it wasn't just sitting there 1/3rd done to confuse folks | 20:52 |
openstackgerrit | Xavier proposed openstack/ironic: Avoid name errors in oneview periodics https://review.openstack.org/396399 | 20:53 |
krtaylor | exactly | 20:53 |
*** jkilpatr has joined #openstack-ironic | 20:55 | |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 20:57 |
*** mjturek has joined #openstack-ironic | 20:58 | |
krtaylor | brb, nature break... | 20:59 |
jroll | krtaylor: left a comment there for formatting | 20:59 |
JayF | jroll: yep was looking for an example to link, you must remember rst better than me | 21:00 |
jroll | JayF: nah, I just google faster :) | 21:00 |
*** MattMan has joined #openstack-ironic | 21:02 | |
openstackgerrit | Félix Bouliane proposed openstack/ironic: Make snmp udp transport settings configurable https://review.openstack.org/396400 | 21:02 |
*** jkilpatr has quit IRC | 21:03 | |
krtaylor | sure, good idea | 21:04 |
mrda | Morning Ironic | 21:05 |
*** baoli has quit IRC | 21:07 | |
*** causten has quit IRC | 21:07 | |
*** baoli has joined #openstack-ironic | 21:07 | |
openstackgerrit | Kurt Taylor proposed openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 21:08 |
* krtaylor didn't know about Admonitions, very cool | 21:09 | |
*** moshele has quit IRC | 21:11 | |
*** jkilpatr has joined #openstack-ironic | 21:17 | |
*** moshele has joined #openstack-ironic | 21:17 | |
*** chas_ has joined #openstack-ironic | 21:19 | |
openstackgerrit | Félix Bouliane proposed openstack/ironic: Make snmp udp transport settings configurable https://review.openstack.org/396400 | 21:19 |
*** jkilpatr has quit IRC | 21:21 | |
*** Goneri has quit IRC | 21:22 | |
xhku | jayF: I Already know for the RFE, I was just trying to make what we use available to others, Im still working on testing this on devstack so we could get it removed from deprecated drivers. | 21:23 |
*** portdirect is now known as portdirect_away | 21:23 | |
*** chas_ has quit IRC | 21:23 | |
xhku | JayF: Thanks to have taken the time to respond | 21:24 |
xhku | JayF: im Félix Bouliane btw | 21:24 |
JayF | xhku: I'm aware :) Just putting the -2 on it for procedural reasons. Just ping me if/when there is CI and it's un-deprecated and I'll remove it. | 21:25 |
xhku | JayF Thanks :) | 21:25 |
*** ccamacho1 has quit IRC | 21:31 | |
openstackgerrit | Annie Lezil proposed openstack/ironic-python-agent: Trying to create a package for biosdevname https://review.openstack.org/396406 | 21:31 |
*** godp1301 has quit IRC | 21:33 | |
openstackgerrit | John L. Villalovos proposed openstack/ironic: WIP/DNM: Playground https://review.openstack.org/396407 | 21:36 |
aslezil | Hey Ironic, I trying to create biosedevname package for tiny OS. I am facing few issues. This is a sample patch https://review.openstack.org/#/c/396406/1. Since biosdevname needs pre requisits, I have to add pci-utils.tcz which is a tiny core OS package. But I am not able to dowloas it gives me 404 error | 21:36 |
patchbot | patch 396406 - ironic-python-agent - Trying to create a package for biosdevname | 21:36 |
aslezil | Could anyone please help to figure it out | 21:36 |
*** mjturek has quit IRC | 21:36 | |
*** ijw has quit IRC | 21:37 | |
JayF | aslezil: I'm curious to see what'll happen with the gate job when it tries to build tinyipa. once it gets to that error i'll try to find some time to take a look | 21:37 |
*** moshele has quit IRC | 21:37 | |
mat128 | aslezil: also, it will need to be added to finalreqs.lst for it to be installed in the final ramdisk | 21:37 |
mat128 | aslezil: follow tgt/qemu examples | 21:37 |
jlvillal | rloo: https://review.openstack.org/#/c/396407/1/ironic/conductor/rpcapi.py@89 I was wondering if that would be okay. What do you think of the idea? | 21:38 |
patchbot | patch 396407 - ironic - WIP/DNM: Playground | 21:38 |
aslezil | http://paste.openstack.org/show/588839/ this is the error | 21:38 |
*** baoli has quit IRC | 21:39 | |
aslezil | mat128 I tried adding in finalreqs.lst but did not work | 21:39 |
rloo | jlvillal: it might. i remember thinking that but i don't recall why i didn't do it. maybe never got around to doing it. oh. it won't work. | 21:40 |
rloo | jlvillal: well, my mind is a bit numb from reviewing. the api service is separate from conductor service. | 21:40 |
JayF | aslezil: it's really weird that the error doesn't have the url it attempted downloading from in it :/ | 21:40 |
jlvillal | rloo: Ah thanks | 21:41 |
rloo | jlvillal: although the codebase would be the same for an api service so it could get the value from the manager code. just might not be the same as what is being used by conductor service. but that is a separate thing i think. | 21:41 |
aslezil | JayF I checked the tiny core OS downloads it has the pci-utils.tcz | 21:41 |
*** baoli has joined #openstack-ironic | 21:42 | |
rloo | jlvillal: it may also have been like that because a bit more work was going to be done to deal with sync'ing them (between api & conductor). but it was so long ago. | 21:42 |
jlvillal | rloo: Okay. I was just noticing how we have to update both files each time when making change. But there is a unit test to ensure they stay the same. So it works. | 21:43 |
rloo | jlvillal: rolling upgrades will be doing something with that rpc value i believe... | 21:43 |
*** baoli_ has joined #openstack-ironic | 21:43 | |
*** hshiina has quit IRC | 21:46 | |
*** mjturek has joined #openstack-ironic | 21:46 | |
openstackgerrit | Annie Lezil proposed openstack/ironic-python-agent: Trying to create a package for biosdevname https://review.openstack.org/396406 | 21:47 |
*** baoli has quit IRC | 21:47 | |
*** xavierr is now known as xavierr|afk | 21:48 | |
*** jkilpatr has joined #openstack-ironic | 21:50 | |
*** jkilpatr has quit IRC | 21:51 | |
*** jkilpatr has joined #openstack-ironic | 21:51 | |
*** ccamacho has joined #openstack-ironic | 21:51 | |
mat128 | aslezil: good news (kinda). I was able to reproduce the issue in a much quicker way | 21:52 |
mat128 | docker run -it tatsushid/tinycore:7.1-x86_64 | 21:52 |
mat128 | tce-load -wci pci-utils.tcz | 21:52 |
mat128 | aslezil: exact same message | 21:52 |
mat128 | my guess is that the file is missing on the repo | 21:52 |
aslezil | mat128: thanks so is there a way to get that file | 21:53 |
mat128 | aslezil: thats what I'm trying to do | 21:53 |
*** causten has joined #openstack-ironic | 21:54 | |
*** wajdi_ has joined #openstack-ironic | 21:56 | |
*** baoli_ has quit IRC | 21:56 | |
mat128 | aslezil: I can download the file just file.. http://repo.tinycorelinux.net/7.x/x86/tcz/pci-utils.tcz.md5.txt | 21:56 |
*** baoli has joined #openstack-ironic | 21:56 | |
mat128 | aslezil: for some reason, it's trying to get /7.x/x86_64/tcz/pci-utils.tcz.dep | 21:57 |
mat128 | (_64) | 21:57 |
mat128 | aslezil: http://paste.openstack.org/show/588849/ | 21:58 |
mat128 | tcpdump to the rescue | 21:58 |
*** ijw has joined #openstack-ironic | 21:58 | |
aslezil | mat128: oops | 21:58 |
mat128 | aslezil: http://paste.openstack.org/show/588853/ | 21:59 |
aslezil | mat128: is a way to get pci-utils | 21:59 |
mat128 | x86 vs x86_64 | 21:59 |
krtaylor | jroll, any other comments? https://review.openstack.org/#/c/353102 | 21:59 |
patchbot | patch 353102 - ironic - Add CI documentation outline | 21:59 |
mat128 | not sure why the package is not there | 21:59 |
*** wajdi has quit IRC | 21:59 | |
*** yee379 has joined #openstack-ironic | 22:00 | |
mat128 | aslezil: the other packages are also available under x86_64 | 22:00 |
mat128 | aslezil: take http://repo.tinycorelinux.net/7.x/x86_64/tcz/hdparm.tcz.md5.txt for example | 22:00 |
jroll | krtaylor: sorry, was pulled into a meeting | 22:01 |
* jroll clicks | 22:01 | |
*** ijw has quit IRC | 22:01 | |
krtaylor | jroll, no worries | 22:02 |
*** ijw has joined #openstack-ironic | 22:02 | |
aslezil | mat128: biosdevname has a pre requisit on pci-utils that my problem | 22:02 |
jroll | krtaylor: +W | 22:02 |
jroll | thanks! | 22:02 |
mat128 | aslezil: the x86_64 has a much shorter list than x86: 2291 x86. vs 1839 x86_64.lst | 22:02 |
jroll | krtaylor: it's 1700 here but if you end up with additions to that I can review tomorrow :) | 22:02 |
mat128 | aslezil: I understand that, maybe there is a packaging/distribution problem in TCL | 22:03 |
mat128 | they are normally very helpful on the forums | 22:03 |
mat128 | s/they/the tinycore people | 22:03 |
rama_y | aslezil: just my 2 cents: is it pciutils.tcz, not pic-utils.tcz? | 22:03 |
krtaylor | jroll, thanks, I'll get with mjturek and see if we can crank out a section or two | 22:03 |
rama_y | aslezil: just my 2 cents: is it pciutils.tcz, not pci-utils.tcz? | 22:03 |
mat128 | rama_y: ting ting ting | 22:03 |
mat128 | rama_y: that was it | 22:03 |
jroll | krtaylor: sounds good :) | 22:04 |
mat128 | rama_y: pciutils.tcz for x86_64 | 22:04 |
JayF | mat128: why the hell would it have a different name in one than the other | 22:04 |
mjturek | very cool krtaylor, I'll take a look | 22:04 |
JayF | that's insanity | 22:04 |
mat128 | rama_y, aslezil: pciutils.tcz for x86_64, pci-utils.tcz for x86 | 22:04 |
mat128 | JayF: no clue :( | 22:04 |
rama_y | That's what I saw: pciutils.tcz without a hyphen | 22:04 |
mat128 | http://repo.tinycorelinux.net/7.x/x86_64/tcz/pciutils.tcz.md5.txt | 22:04 |
JayF | egad | 22:04 |
mat128 | http://repo.tinycorelinux.net/7.x/x86/tcz/pci-utils.tcz.md5.txt | 22:04 |
aslezil | thanks mat128 and rama_y | 22:04 |
* jroll see y'all tomorrow | 22:04 | |
JayF | we don't support !64bit for tinyipa, do we? lol | 22:04 |
krtaylor | mjturek, we can start with the pool management and I have a references section from the third-party docs | 22:05 |
*** rbudden has quit IRC | 22:05 | |
mat128 | JayF: thats correct | 22:05 |
*** ijw has quit IRC | 22:05 | |
krtaylor | mjturek, we also don't want to loose the comments in the outline patchset, we need to roll those in, good stuff | 22:05 |
mat128 | JayF: it's hardcoded https://github.com/openstack/ironic-python-agent/blob/master/imagebuild/tinyipa/build-tinyipa.sh#L36-L37 | 22:05 |
mjturek | krtaylor: absolutely | 22:05 |
*** ijw has joined #openstack-ironic | 22:05 | |
mariojv | good night ironic - i'm on pto tomorrow so hope everyone has a great weekend | 22:07 |
*** rbudden has joined #openstack-ironic | 22:07 | |
*** portdirect_away is now known as portdirect | 22:07 | |
openstackgerrit | Merged openstack/ironic: Add CI documentation outline https://review.openstack.org/353102 | 22:08 |
aslezil | ls | 22:12 |
mat128 | . .. .gitconfig projects | 22:12 |
mat128 | # | 22:12 |
JayF | lol | 22:13 |
aslezil | :P | 22:13 |
*** jcoufal has quit IRC | 22:15 | |
*** trown is now known as trown|outtypewww | 22:15 | |
*** aNuposic has joined #openstack-ironic | 22:16 | |
*** ijw has quit IRC | 22:16 | |
*** baoli has quit IRC | 22:17 | |
*** baoli has joined #openstack-ironic | 22:17 | |
*** ijw has joined #openstack-ironic | 22:18 | |
*** godp1301 has joined #openstack-ironic | 22:20 | |
*** jerrygb_ has joined #openstack-ironic | 22:21 | |
*** crinkle has joined #openstack-ironic | 22:21 | |
*** ccamacho has quit IRC | 22:22 | |
*** godp1301_ has joined #openstack-ironic | 22:23 | |
*** jerrygb has quit IRC | 22:24 | |
*** mjturek has quit IRC | 22:24 | |
*** jerrygb has joined #openstack-ironic | 22:25 | |
*** jerrygb_ has quit IRC | 22:26 | |
*** godp1301 has quit IRC | 22:26 | |
*** jerrygb_ has joined #openstack-ironic | 22:27 | |
*** jheroux has quit IRC | 22:28 | |
*** jerrygb has quit IRC | 22:30 | |
*** sabeen has quit IRC | 22:30 | |
*** chuckC_ has joined #openstack-ironic | 22:30 | |
*** jerrygb_ has quit IRC | 22:31 | |
*** fragatin_ has joined #openstack-ironic | 22:34 | |
*** aNuposic has quit IRC | 22:37 | |
*** wajdi_ has quit IRC | 22:37 | |
*** fragatina has quit IRC | 22:37 | |
*** baoli has quit IRC | 22:38 | |
*** baoli has joined #openstack-ironic | 22:39 | |
*** chas_ has joined #openstack-ironic | 22:58 | |
*** baoli has quit IRC | 23:00 | |
*** baoli has joined #openstack-ironic | 23:00 | |
*** rama_y has quit IRC | 23:02 | |
*** chas_ has quit IRC | 23:02 | |
*** chuckC_ has quit IRC | 23:03 | |
*** lindycoder has quit IRC | 23:03 | |
*** chas_ has joined #openstack-ironic | 23:04 | |
*** fragatin_ has quit IRC | 23:04 | |
*** fragatina has joined #openstack-ironic | 23:05 | |
*** chas_ has quit IRC | 23:09 | |
*** baoli has quit IRC | 23:11 | |
*** rama_y has joined #openstack-ironic | 23:15 | |
*** Sukhdev has quit IRC | 23:16 | |
*** rloo has quit IRC | 23:16 | |
*** pmannidi has quit IRC | 23:23 | |
*** pmannidi has joined #openstack-ironic | 23:24 | |
*** e0ne has joined #openstack-ironic | 23:26 | |
*** rbudden has quit IRC | 23:30 | |
*** yuanying has quit IRC | 23:34 | |
*** rbrud has quit IRC | 23:35 | |
*** godp1301 has joined #openstack-ironic | 23:38 | |
*** godp1301_ has quit IRC | 23:42 | |
*** godp1301 has quit IRC | 23:43 | |
*** ijw has quit IRC | 23:44 | |
*** ijw has joined #openstack-ironic | 23:44 | |
*** yuanying has joined #openstack-ironic | 23:44 | |
*** ijw has quit IRC | 23:46 | |
*** ijw has joined #openstack-ironic | 23:46 | |
*** vsaienko has quit IRC | 23:46 | |
*** e0ne has quit IRC | 23:48 | |
*** krtaylor has quit IRC | 23:54 | |
*** amotoki has joined #openstack-ironic | 23:58 | |
*** amotoki has quit IRC | 23:58 | |
*** amotoki has joined #openstack-ironic | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!