| *** romcheg has quit IRC | 00:11 | |
| *** Marga_ has quit IRC | 00:31 | |
| *** david-lyle has joined #openstack-ironic | 00:35 | |
| *** Masahiro has joined #openstack-ironic | 00:35 | |
| *** ryanpetrello has joined #openstack-ironic | 00:39 | |
| *** Masahiro has quit IRC | 00:40 | |
| *** Haomeng has joined #openstack-ironic | 00:41 | |
| *** Haomeng|2 has quit IRC | 00:42 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic-specs: Add a new OS Reboot driver https://review.openstack.org/146361 | 00:52 |
|---|---|---|
| *** ryanpetrello has quit IRC | 00:58 | |
| *** jiangfei has quit IRC | 01:00 | |
| *** ryanpetrello has joined #openstack-ironic | 01:17 | |
| *** david-lyle has quit IRC | 01:18 | |
| *** chenglch has joined #openstack-ironic | 01:25 | |
| *** david-lyle has joined #openstack-ironic | 01:25 | |
| *** david-lyle has quit IRC | 01:27 | |
| *** david-lyle has joined #openstack-ironic | 01:28 | |
| openstackgerrit | John Trowbridge proposed stackforge/ironic-discoverd: Changes utils.get_keystone(token) to utils.is_admin(token) https://review.openstack.org/145657 | 01:43 |
| *** ryanpetrello has quit IRC | 01:48 | |
| *** Masahiro has joined #openstack-ironic | 01:51 | |
| *** Masahiro has quit IRC | 01:56 | |
| *** jerryz has joined #openstack-ironic | 01:58 | |
| *** david-lyle has quit IRC | 01:59 | |
| *** killer_prince has joined #openstack-ironic | 02:14 | |
| *** lazy_prince has quit IRC | 02:15 | |
| *** killer_prince is now known as lazy_prince | 02:15 | |
| *** Marga_ has joined #openstack-ironic | 02:29 | |
| *** r-daneel has joined #openstack-ironic | 02:42 | |
| *** ramineni has joined #openstack-ironic | 02:44 | |
| *** Marga_ has quit IRC | 02:44 | |
| *** Marga_ has joined #openstack-ironic | 02:46 | |
| *** pcrews has joined #openstack-ironic | 02:48 | |
| *** Marga_ has quit IRC | 03:11 | |
| *** Marga_ has joined #openstack-ironic | 03:19 | |
| *** r-daneel has quit IRC | 03:22 | |
| *** r-daneel has joined #openstack-ironic | 03:23 | |
| *** jrist has quit IRC | 03:34 | |
| *** Masahiro has joined #openstack-ironic | 03:40 | |
| *** david-lyle has joined #openstack-ironic | 03:44 | |
| *** Masahiro has quit IRC | 03:45 | |
| *** xxj has joined #openstack-ironic | 03:45 | |
| *** Nisha has joined #openstack-ironic | 04:09 | |
| *** Marga_ has quit IRC | 04:13 | |
| *** rameshg87 has joined #openstack-ironic | 04:16 | |
| rameshg87 | Haomeng, hi | 04:21 |
| Haomeng | rameshg87: yes:) | 04:21 |
| rameshg87 | Haomeng, regarding https://review.openstack.org/#/c/145996/ | 04:22 |
| rameshg87 | Haomeng, i am just checking the dbapi | 04:22 |
| rameshg87 | Haomeng, but at a higher level, i was wondering if we should make the change at all | 04:23 |
| Haomeng | rameshg87: yes, I am not sure how to implement the 'and condition' filter | 04:23 |
| rameshg87 | Haomeng, when node is DEPLOYING it is under the control of some conductor | 04:23 |
| Haomeng | rameshg87: any idea? | 04:23 |
| rameshg87 | Haomeng, if we make a change for conductor B to pickup a node reserved by conductor A | 04:24 |
| rameshg87 | Haomeng, then we would open way to race conditions | 04:24 |
| rameshg87 | Haomeng, i think a node stuck at DEPLOYING is possible only if conductor is dead | 04:24 |
| rameshg87 | Haomeng, am i correct ? | 04:24 |
| Haomeng | rameshg87: yes, make sense | 04:25 |
| Haomeng | rameshg87: no for some time, it is networking issue, such as the iscsi can not be connected at that time | 04:26 |
| rameshg87 | Haomeng, yes, but the control of the node is still with another conductor, right ? | 04:27 |
| Haomeng | rameshg87: or some case that not working on local disk to create paration or something | 04:27 |
| Haomeng | rameshg87: I just run the case which is networking issue, the conductor can not connect to bare metal via iscsi | 04:28 |
| rameshg87 | Haomeng, hmm .. okay | 04:28 |
| Haomeng | rameshg87: maybe the route is wrong or firewall issue | 04:28 |
| rameshg87 | Haomeng, so imagine if conductor A sets a node to DEPLOYING and is stuck | 04:28 |
| rameshg87 | Haomeng, if conductor B finds out that deploy has timed out and takes back control | 04:29 |
| rameshg87 | Haomeng, how do we make sure that conductor A doesn't wake up at some point of time later and takes back control of node ? | 04:29 |
| *** coolsvap|afk is now known as coolsvap | 04:31 | |
| Haomeng | rameshg87: yes | 04:31 |
| Haomeng | rameshg87: I have just one conductor | 04:32 |
| Haomeng | rameshg87: not the *another conductor* | 04:32 |
| rameshg87 | Haomeng, yeah i guess it will be a bit risky in the multi-conductor scenario | 04:33 |
| rameshg87 | Haomeng, i guess the best place we can act upon is when the conductor is dead while doing deployment of node | 04:33 |
| Haomeng | rameshg87: yes, this is new question | 04:36 |
| Haomeng | rameshg87: ok, let me hold on here and resume if we have better solution | 04:37 |
| Haomeng | rameshg87: thank you! | 04:37 |
| rameshg87 | Haomeng, sure .. thanks | 04:37 |
| Haomeng | rameshg87: :) | 04:37 |
| Haomeng | rameshg87: another question | 04:42 |
| rameshg87 | Haomeng, sure | 04:42 |
| Haomeng | rameshg87: do you think if our ironic can work with neutron to support vlan networking? | 04:42 |
| Haomeng | rameshg87: I just tested with flat networking, works well | 04:42 |
| Haomeng | rameshg87: and commit some doc patch to fix the flat networking guide | 04:43 |
| rameshg87 | Haomeng, i have very limited idea in neutron space :( | 04:43 |
| Haomeng | rameshg87: np, I am too:) | 04:43 |
| Haomeng | rameshg87: I just think the production env need vlan networking but flat | 04:44 |
| Haomeng | rameshg87: and this is the new area which is how to work with neutron to control the physical network | 04:44 |
| rameshg87 | Haomeng, in layman terms | 04:45 |
| Haomeng | rameshg87: :( | 04:45 |
| rameshg87 | Haomeng, is vlan networking something like this ? | 04:45 |
| Haomeng | rameshg87: I just think about the customer network env - three or more vlans: manage vlan, api vlan, tenant vlan etc:) | 04:46 |
| rameshg87 | Haomeng, we have 3-4 nics on a bare metal, and you need to control how to provide connectivity to different networks through a single network ? | 04:46 |
| rameshg87 | Haomeng, yeah so if we take a bare metal | 04:46 |
| Haomeng | rameshg87: yes not just single network | 04:46 |
| Haomeng | rameshg87: did you see such plugin - https://github.com/rackerlabs/ironic-neutron-plugin | 04:46 |
| Haomeng | rameshg87: looks like this plugin supports vlan | 04:47 |
| Haomeng | which can manage physical switch | 04:47 |
| Haomeng | and this one - https://review.openstack.org/#/c/90422/2/specs/juno/ml2-mechanism-snmp-vlan.rst | 04:47 |
| rameshg87 | Haomeng, oh okay :) | 04:48 |
| Haomeng | rameshg87: :) | 04:48 |
| rameshg87 | Haomeng, i think i should read them before even planning to ask some questions :) | 04:48 |
| Haomeng | rameshg87: np, dont worry:) | 04:49 |
| Haomeng | rameshg87: :) | 04:49 |
| Haomeng | rameshg87: maybe racespace guys know this plugin, and maybe they run this in their production env | 04:49 |
| rameshg87 | Haomeng, yeah may be .. | 04:49 |
| Haomeng | rameshg87: :) | 04:50 |
| *** david-lyle has quit IRC | 04:50 | |
| *** david-lyle has joined #openstack-ironic | 04:50 | |
| Haomeng | and this spec - https://review.openstack.org/#/c/143480/2/specs/kilo/network-provider.rst | 04:51 |
| rameshg87 | Haomeng, yeah i had seen this spec infact | 04:52 |
| Haomeng | rameshg87: but it not mention with how to control vlan | 04:52 |
| rameshg87 | Haomeng, yeah it was just talking about providing different network connectivity for bare metals during deploy time and after-deploy | 04:53 |
| Haomeng | rameshg87: yes | 04:53 |
| Haomeng | rameshg87: to secure the network to switch networking is the core idea | 04:55 |
| Haomeng | rameshg87: but not sure if can support vlan, so I left comments about vlan concern:) | 04:55 |
| Haomeng | :) | 04:55 |
| rameshg87 | Haomeng, :) | 04:56 |
| *** david-lyle has quit IRC | 05:06 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic-specs: Implement Cleaning Operations for iLO drivers https://review.openstack.org/142401 | 05:09 |
| *** pcrews has quit IRC | 05:11 | |
| *** teju has joined #openstack-ironic | 05:22 | |
| *** pensu has joined #openstack-ironic | 05:26 | |
| *** Masahiro has joined #openstack-ironic | 05:29 | |
| *** Masahiro has quit IRC | 05:33 | |
| *** pradipta_away has quit IRC | 05:39 | |
| *** pradipta has joined #openstack-ironic | 06:01 | |
| *** sambetts has quit IRC | 06:23 | |
| *** r-daneel has quit IRC | 06:25 | |
| *** Masahiro has joined #openstack-ironic | 06:30 | |
| *** Haomeng has quit IRC | 06:31 | |
| *** Nisha has quit IRC | 06:32 | |
| *** Masahiro has quit IRC | 06:34 | |
| *** rakesh_hs2 has joined #openstack-ironic | 06:48 | |
| *** jerryz has quit IRC | 06:52 | |
| *** jerryz has joined #openstack-ironic | 06:53 | |
| *** lazy_prince has quit IRC | 07:03 | |
| *** lazy_prince has joined #openstack-ironic | 07:03 | |
| *** dlpartain has joined #openstack-ironic | 07:10 | |
| *** dlpartain has left #openstack-ironic | 07:10 | |
| *** Nisha has joined #openstack-ironic | 07:19 | |
| *** coolsvap is now known as coolsvap|afk | 07:20 | |
| *** coolsvap has joined #openstack-ironic | 07:29 | |
| *** subscope has joined #openstack-ironic | 07:37 | |
| *** coolsvap has quit IRC | 07:38 | |
| *** ijw has joined #openstack-ironic | 08:04 | |
| *** chlong has quit IRC | 08:10 | |
| *** Nisha has quit IRC | 08:18 | |
| *** Masahiro has joined #openstack-ironic | 08:18 | |
| *** Masahiro has quit IRC | 08:23 | |
| *** lazy_prince has quit IRC | 08:25 | |
| *** Nisha has joined #openstack-ironic | 08:34 | |
| dtantsur | Morning folks! | 08:39 |
| *** coolsvap|afk is now known as coolsvap | 08:42 | |
| *** ifarkas has joined #openstack-ironic | 08:43 | |
| *** romcheg has joined #openstack-ironic | 08:43 | |
| *** jcoufal has joined #openstack-ironic | 08:48 | |
| *** ifarkas has quit IRC | 08:51 | |
| *** romcheg has quit IRC | 08:52 | |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Add tests for ilo_deploy driver https://review.openstack.org/142985 | 08:53 |
| *** ifarkas has joined #openstack-ironic | 08:55 | |
| *** andreykurilin has joined #openstack-ironic | 08:59 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Allow drivers to have their own periodic tasks https://review.openstack.org/135589 | 09:01 |
| *** teju has quit IRC | 09:02 | |
| *** achanda has joined #openstack-ironic | 09:06 | |
| *** derekh has joined #openstack-ironic | 09:07 | |
| *** Nisha has quit IRC | 09:10 | |
| *** Nisha has joined #openstack-ironic | 09:15 | |
| *** romcheg has joined #openstack-ironic | 09:18 | |
| *** ndipanov has joined #openstack-ironic | 09:22 | |
| *** MattMan has joined #openstack-ironic | 09:25 | |
| *** athomas has joined #openstack-ironic | 09:27 | |
| *** andreykurilin has quit IRC | 09:28 | |
| *** andreykurilin has joined #openstack-ironic | 09:29 | |
| *** achanda has quit IRC | 09:30 | |
| *** killer_prince has joined #openstack-ironic | 09:32 | |
| *** killer_prince is now known as lazy_prince | 09:32 | |
| *** lazy_prince has quit IRC | 09:34 | |
| *** lazy_prince has joined #openstack-ironic | 09:35 | |
| *** ijw has quit IRC | 09:37 | |
| *** ijw has joined #openstack-ironic | 09:38 | |
| *** pelix has joined #openstack-ironic | 09:40 | |
| *** jcoufal_ has joined #openstack-ironic | 09:41 | |
| *** jcoufal has quit IRC | 09:44 | |
| *** andreykurilin has quit IRC | 09:50 | |
| *** jistr has joined #openstack-ironic | 09:58 | |
| *** bauwser is now known as bauzas | 10:01 | |
| *** achanda has joined #openstack-ironic | 10:02 | |
| *** Masahiro has joined #openstack-ironic | 10:07 | |
| *** ijw has quit IRC | 10:08 | |
| *** ijw has joined #openstack-ironic | 10:09 | |
| *** Masahiro has quit IRC | 10:11 | |
| *** achanda has quit IRC | 10:15 | |
| *** Nisha has quit IRC | 10:22 | |
| *** jerryz has quit IRC | 10:34 | |
| *** Jatin360 has joined #openstack-ironic | 10:42 | |
| *** Nisha has joined #openstack-ironic | 10:44 | |
| *** Nisha has quit IRC | 10:49 | |
| *** jerryz has joined #openstack-ironic | 10:49 | |
| *** andreykurilin has joined #openstack-ironic | 10:56 | |
| *** pcaruana has joined #openstack-ironic | 10:59 | |
| *** ramineni has quit IRC | 11:03 | |
| *** rameshg87 has quit IRC | 11:03 | |
| *** lucasagomes has joined #openstack-ironic | 11:04 | |
| *** pradipta has quit IRC | 11:05 | |
| *** pradipta has joined #openstack-ironic | 11:08 | |
| *** Masahiro has joined #openstack-ironic | 11:08 | |
| *** andreykurilin has quit IRC | 11:10 | |
| *** gilllliard is now known as gilliard | 11:11 | |
| *** Masahiro has quit IRC | 11:13 | |
| *** pradipta is now known as pradipta_away | 11:21 | |
| *** david-lyle has joined #openstack-ironic | 11:32 | |
| *** lsmola_ has quit IRC | 11:32 | |
| *** Marga_ has joined #openstack-ironic | 11:39 | |
| *** vdrok has joined #openstack-ironic | 11:40 | |
| *** chlong has joined #openstack-ironic | 11:41 | |
| dtantsur | ifarkas, hi! please review https://review.openstack.org/#/c/146148 when you have some time | 11:46 |
| *** chlong has quit IRC | 11:47 | |
| *** chlong has joined #openstack-ironic | 11:47 | |
| openstackgerrit | Merged openstack/ironic: Delete unnecessary document files https://review.openstack.org/145699 | 11:56 |
| *** Jatin360 has quit IRC | 12:02 | |
| *** pensu has quit IRC | 12:02 | |
| *** chenglch has quit IRC | 12:05 | |
| *** chenglch has joined #openstack-ironic | 12:05 | |
| *** EmilienM|afk is now known as EmilienM | 12:10 | |
| ifarkas | dtantsur, ok | 12:11 |
| openstackgerrit | Merged openstack/ironic: Adjust the help strings to better reflect usage https://review.openstack.org/146299 | 12:18 |
| *** vdrok has quit IRC | 12:22 | |
| *** vdrok has joined #openstack-ironic | 12:25 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Support new Kilo state machine https://review.openstack.org/137101 | 12:31 |
| dtantsur | lucasagomes, hey, mind having a look at this patch ^^^? I want to be sure I understood things correctly | 12:31 |
| lucasagomes | dtantsur, hey, sure | 12:32 |
| lucasagomes | dtantsur, hmm it seems that inspection can only be triggered from MANAGED not from ENROLL | 12:37 |
| lucasagomes | tho it's considering enroll as a valid state | 12:37 |
| dtantsur | lucasagomes, it's from Ironic point of view :) from out point of view, we have that strange scenario with setting IPMI creds | 12:37 |
| dtantsur | lucasagomes, which means we have to support enroll... | 12:37 |
| lucasagomes | a-ha, gotcha | 12:38 |
| lucasagomes | also I really wanted to change the name MANAGED, to MANAGE | 12:38 |
| lucasagomes | cause there's that -ED -ING sufix states to be temporary state | 12:38 |
| lucasagomes | temporary = active... | 12:39 |
| dtantsur | hmmm.... | 12:39 |
| dtantsur | is there any consensus on this change? | 12:39 |
| lucasagomes | but it's not part of the discovered thing | 12:39 |
| lucasagomes | I think devananda have agreed with it before | 12:39 |
| *** alexpilotti has joined #openstack-ironic | 12:39 | |
| lucasagomes | I think I will propose a patch in the specs, see what people think | 12:40 |
| lucasagomes | dtantsur, but anyway, other than that the patch lgtm | 12:40 |
| dtantsur | aha I see... | 12:40 |
| lucasagomes | you may want to copy the states.py from ironic to discovered later | 12:40 |
| lucasagomes | and keep that in sync, like we do for the nova driver | 12:40 |
| dtantsur | maybe. anyway it's a manual process | 12:41 |
| dtantsur | lucasagomes, please try to propose&discuss the change asap, as I plan on release discoverd 1.0.0 with Kilo-2 | 12:42 |
| lucasagomes | will do it right now | 12:42 |
| dtantsur | thanks | 12:42 |
| *** alexpilotti has quit IRC | 12:44 | |
| *** Marga_ has quit IRC | 12:46 | |
| *** lsmola has joined #openstack-ironic | 12:46 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Rename MANAGED to MANAGE https://review.openstack.org/146452 | 12:47 |
| lucasagomes | ^ | 12:47 |
| * dtantsur is looking | 12:49 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Rename MANAGED to MANAGE https://review.openstack.org/146452 | 12:49 |
| lucasagomes | fixing the diagram that was a bit out | 12:49 |
| lucasagomes | dtantsur, but anyway, that's just a nit... the logic in the discoverd patch seems correct | 12:50 |
| dtantsur | lucasagomes, hmm. in some places we use verbs, in some not. maybe it should be MANAGEABLE? (like AVAILABLE, ACTIVE etc) | 12:51 |
| dtantsur | yeah, thanks | 12:51 |
| * dtantsur is leaving a comments on the spec too | 12:51 | |
| * dtantsur -> lunch | 12:52 | |
| lucasagomes | yeah, MANAGEABLE seems good too | 12:52 |
| *** chlong has quit IRC | 12:55 | |
| *** Masahiro has joined #openstack-ironic | 12:57 | |
| *** alexpilotti has joined #openstack-ironic | 12:58 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Ensure that image link points to the correct image https://review.openstack.org/145254 | 13:01 |
| *** Masahiro has quit IRC | 13:01 | |
| *** dprince has joined #openstack-ironic | 13:02 | |
| *** Marga_ has joined #openstack-ironic | 13:17 | |
| *** ryanpetrello has joined #openstack-ironic | 13:17 | |
| *** agordeev has joined #openstack-ironic | 13:20 | |
| *** ifarkas has quit IRC | 13:21 | |
| *** Marga_ has quit IRC | 13:22 | |
| *** ijw has quit IRC | 13:37 | |
| *** ijw has joined #openstack-ironic | 13:38 | |
| *** ifarkas has joined #openstack-ironic | 13:39 | |
| Shrews | lucasagomes: morning | 13:39 |
| lucasagomes | Shrews, hi there, good morning | 13:40 |
| Shrews | lucasagomes: qq... won't https://review.openstack.org/144265 break any out of tree drivers? | 13:40 |
| * lucasagomes clicks | 13:40 | |
| lucasagomes | Shrews, that was my concern on previous reviews | 13:41 |
| lucasagomes | so I suggest to make method a kwargs | 13:41 |
| lucasagomes | method=None | 13:41 |
| lucasagomes | just to make it more explicit | 13:41 |
| Shrews | oh, i should have read more closely :) | 13:41 |
| * Shrews needs coffee | 13:41 | |
| lucasagomes | Shrews, :) it's all good | 13:41 |
| *** chenglch has quit IRC | 13:46 | |
| *** jjulien has quit IRC | 13:46 | |
| *** chenglch has joined #openstack-ironic | 13:47 | |
| *** ijw has quit IRC | 13:47 | |
| *** ijw has joined #openstack-ironic | 13:47 | |
| *** ifarkas has quit IRC | 13:57 | |
| *** ifarkas has joined #openstack-ironic | 13:58 | |
| *** rloo has joined #openstack-ironic | 13:59 | |
| *** chenglch has quit IRC | 14:02 | |
| *** jjohnson2 has joined #openstack-ironic | 14:08 | |
| *** rameshg87 has joined #openstack-ironic | 14:08 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Rework node cache clean up according to recent changes https://review.openstack.org/146148 | 14:09 |
| *** ijw has quit IRC | 14:10 | |
| *** ijw has joined #openstack-ironic | 14:10 | |
| *** Marga_ has joined #openstack-ironic | 14:18 | |
| vdrok | hi all | 14:20 |
| vdrok | rameshg87, around? | 14:20 |
| rameshg87 | vdrok, o/ | 14:20 |
| vdrok | rameshg87, hi :) | 14:20 |
| rameshg87 | vdrok, hi :) | 14:20 |
| vdrok | rameshg87, so, as for your comment here https://review.openstack.org/#/c/136741/9/ironic/drivers/modules/ilo/deploy.py | 14:20 |
| vdrok | rameshg87, it seems that there is only 2 options | 14:21 |
| vdrok | http or glance | 14:21 |
| rameshg87 | vdrok, i tried with http, that's when it failed | 14:21 |
| rameshg87 | vdrok, i just changed get_glance_image_property to be generic and not use glance by default | 14:21 |
| rameshg87 | vdrok, as commented in https://review.openstack.org/#/c/136741/9/ironic/common/images.py | 14:22 |
| vdrok | rameshg87, and you provided ilo_boot_iso in instance_info? | 14:22 |
| rameshg87 | vdrok, no i didn't | 14:22 |
| rameshg87 | vdrok, i provided kernel and ramdisk in instance_info | 14:22 |
| vdrok | rameshg87, ah, I see | 14:22 |
| *** Marga_ has quit IRC | 14:22 | |
| vdrok | rameshg87, yup, got it, will fix | 14:23 |
| rameshg87 | vdrok, thanks | 14:23 |
| rameshg87 | NobodyCam, rloo, can you please have a look at https://review.openstack.org/137291 | 14:26 |
| rameshg87 | NobodyCam, rloo, both of you have already taken a look at it a couple of times :) | 14:26 |
| rameshg87 | NobodyCam, rloo, it has been lying idle for some 20 days :( | 14:26 |
| *** rameshg87 is now known as rameshg87-dinner | 14:27 | |
| *** rakesh_hs2 has quit IRC | 14:31 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Refactor stable API to be /v1/introspection/<UUID> https://review.openstack.org/146475 | 14:33 |
| *** kbs1 has joined #openstack-ironic | 14:34 | |
| *** jerryz has quit IRC | 14:35 | |
| *** Marga_ has joined #openstack-ironic | 14:35 | |
| *** lazy_prince is now known as killer_prince | 14:37 | |
| *** stendulker has joined #openstack-ironic | 14:40 | |
| *** Marga_ has quit IRC | 14:44 | |
| *** Marga_ has joined #openstack-ironic | 14:44 | |
| *** Masahiro has joined #openstack-ironic | 14:45 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add introspect client call and switch functest to it https://review.openstack.org/146479 | 14:48 |
| dtantsur | ifarkas, could you please also review at least https://review.openstack.org/#/c/137101/ today? it's pretty small | 14:48 |
| ifarkas | dtantsur, sure thing | 14:49 |
| *** Masahiro has quit IRC | 14:50 | |
| *** rameshg87-dinner is now known as rameshg87 | 14:50 | |
| stendulker | dtantsur, NobodyCam: Can you please review these secure boot related specs https://review.openstack.org/#/c/135228 and https://review.openstack.org/#/c/135845 | 14:50 |
| dtantsur | I'll try to find some time asap | 14:51 |
| ifarkas | dtantsur, there's a sticky -2 on that patch | 14:51 |
| stendulker | thanks dtantsur | 14:51 |
| dtantsur | ifarkas, ooooops :) removed | 14:52 |
| *** ndipanov has quit IRC | 14:58 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Support for non-Glance image references https://review.openstack.org/136741 | 14:58 |
| NobodyCam | good morning Ironic, | 15:01 |
| NobodyCam | rameshg87: will take a look after I has some coffee | 15:02 |
| *** stendulker has quit IRC | 15:02 | |
| rameshg87 | NobodyCam, thanks | 15:02 |
| NobodyCam | :) | 15:02 |
| *** ndipanov has joined #openstack-ironic | 15:02 | |
| *** ifarkas has quit IRC | 15:07 | |
| *** ifarkas has joined #openstack-ironic | 15:08 | |
| *** stendulker has joined #openstack-ironic | 15:12 | |
| *** lucasagomes is now known as lucas-hungry | 15:12 | |
| lucas-hungry | NobodyCam, morning :) | 15:12 |
| lucas-hungry | rameshg87, morning | 15:12 |
| * lucas-hungry brb | 15:12 | |
| NobodyCam | morning lucas-hungry | 15:12 |
| dtantsur | NobodyCam, morning | 15:13 |
| rameshg87 | lucas-hungry, morning :) | 15:13 |
| Shrews | hey NobodyCam | 15:15 |
| NobodyCam | mornign Shrews | 15:15 |
| NobodyCam | how was your weekend | 15:15 |
| Shrews | looooots of football... otherwise, pretty boring :) | 15:15 |
| NobodyCam | lol :) that sounds good to me | 15:16 |
| openstackgerrit | Merged stackforge/ironic-discoverd: Support new Kilo state machine https://review.openstack.org/137101 | 15:16 |
| *** beekneemech is now known as bnemec | 15:16 | |
| NobodyCam | bnemec: did devananda get in touch with you friday? I beleieve he was looking for ya | 15:17 |
| stendulker | NobodyCam: Hi | 15:20 |
| NobodyCam | Hi stendulker | 15:20 |
| bnemec | NobodyCam: No, I saw that he was looking for me, but I never heard anything after you pointed out my nick. | 15:20 |
| NobodyCam | ack was just checking | 15:20 |
| stendulker | NobodyCam: Can you please review these specs related to secure boot https://review.openstack.org/#/c/135845 and https://review.openstack.org/#/c/135228 | 15:20 |
| NobodyCam | stendulker: one already on the list ahead of them but I will add them to my list for today :) | 15:21 |
| openstackgerrit | Jarrod Johnson proposed stackforge/pyghmi: Implement server side IPMI protocol https://review.openstack.org/138109 | 15:22 |
| jjohnson2 | NobodyCam, fyi, that one I'm ready to say good enough for now | 15:22 |
| jjohnson2 | NobodyCam, of particular interest to me is if fakebmc illustrates the sort of interface that a developer would want | 15:23 |
| NobodyCam | jjohnson2: I looked at it on saturday. wow thats some slick stuff... :) | 15:23 |
| NobodyCam | ya | 15:23 |
| NobodyCam | will have to add set next boot device | 15:23 |
| stendulker | thanks NobodyCam | 15:23 |
| jjohnson2 | NobodyCam, to what I was doing or to something else? fakebmc sets/gets boot device | 15:24 |
| NobodyCam | but its looking supper jjohnson2 thank you so much | 15:24 |
| NobodyCam | oh did I miss that... :( | 15:24 |
| jjohnson2 | setting boot device to network | 15:24 |
| jjohnson2 | asked for boot device, saying network | 15:24 |
| jjohnson2 | that's what fakebmc printed out in response to | 15:24 |
| jjohnson2 | chassis bootdev pxe and bootparam get 5 | 15:25 |
| jjohnson2 | that was in the most recent review | 15:25 |
| jjohnson2 | anyway, my concept was a 'bmc.py' base class that is serviceable and fakebmc.py to demonstrate precisely which functions warrant attention as a developer | 15:26 |
| jjohnson2 | I'm most curious as to whether the current scheme of so many power_ functions is good or if people would rather have a single set_power entry point | 15:27 |
| NobodyCam | jjohnson2: yep its there. I must have been looking at rev 10 :( | 15:27 |
| NobodyCam | aswesome-ness | 15:27 |
| *** jrist has joined #openstack-ironic | 15:29 | |
| *** jrist has quit IRC | 15:29 | |
| *** jrist has joined #openstack-ironic | 15:29 | |
| *** ndipanov has quit IRC | 15:31 | |
| jjohnson2 | NobodyCam, anyway, I'll let it sit there, and if you have time or comfort to review fakebmc and maybe bmc.py at least, would appreciate if that is the sort of interface | 15:31 |
| jjohnson2 | now back to my regularly scheduled stuff | 15:33 |
| NobodyCam | jjohnson2: for sure. it may be a little later today. But I deff want to take a good look | 15:33 |
| NobodyCam | jjohnson2: Thank you again | 15:33 |
| jjohnson2 | no problem, I know I stalled a bit | 15:33 |
| NobodyCam | not at all :) | 15:34 |
| *** stendulker has quit IRC | 15:34 | |
| *** Marga_ has quit IRC | 15:36 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Add VirtualBox drivers and its modules https://review.openstack.org/146305 | 15:39 |
| victor_lowther | Good morning, Ironic | 15:40 |
| NobodyCam | morning victor_lowther :) | 15:40 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Refactor stable API to be /v1/introspection/<UUID> https://review.openstack.org/146475 | 15:40 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add introspect client call and switch functest to it https://review.openstack.org/146479 | 15:40 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Update HTTP API documentation to reflect the current reality https://review.openstack.org/146498 | 15:40 |
| *** zz_jgrimm is now known as jgrimm | 15:44 | |
| *** ijw has quit IRC | 15:44 | |
| *** ndipanov has joined #openstack-ironic | 15:44 | |
| NobodyCam | brb | 15:45 |
| *** rameshg87 is now known as rameshg87-brb | 15:49 | |
| NobodyCam | rameshg87-brb: can you ping me when your back | 15:52 |
| *** rameshg87-brb has quit IRC | 15:59 | |
| *** ndipanov has quit IRC | 16:00 | |
| *** ijw has joined #openstack-ironic | 16:02 | |
| *** ndipanov has joined #openstack-ironic | 16:03 | |
| *** ijw has quit IRC | 16:05 | |
| *** ijw has joined #openstack-ironic | 16:05 | |
| *** lucas-hungry is now known as lucasagomes | 16:08 | |
| *** pcrews has joined #openstack-ironic | 16:09 | |
| *** jcoufal_ has quit IRC | 16:10 | |
| NobodyCam | brb | 16:12 |
| *** rameshg87 has joined #openstack-ironic | 16:13 | |
| * NobodyCam back | 16:20 | |
| *** Nisha has joined #openstack-ironic | 16:23 | |
| *** coolsvap is now known as coolsvap|afk | 16:25 | |
| rameshg87 | NobodyCam, hi | 16:25 |
| NobodyCam | hey rameshg87 | 16:27 |
| rameshg87 | hi NobodyCam | 16:27 |
| NobodyCam | rameshg87: I had a question on the hashed value for the boot file name | 16:28 |
| rameshg87 | NobodyCam, yes | 16:28 |
| NobodyCam | The concern I am thinking of is with the file name being a hashed value a operator would not easily be able to tell when a image is nolonger needed (other then ref_count) would there be a way to have a "check" type script that could verfy the hashed names are still in use / needed? | 16:30 |
| NobodyCam | I am concerned that ref_count could get out of sync | 16:31 |
| openstackgerrit | Alexis Lee proposed openstack/ironic: Avoid querying the power state twice https://review.openstack.org/146554 | 16:31 |
| rameshg87 | NobodyCam, okay | 16:32 |
| rameshg87 | NobodyCam, we cannot decrypt hash value and get back the images | 16:33 |
| rameshg87 | NobodyCam, so to tell whether it is in use or not | 16:33 |
| rameshg87 | NobodyCam, the best way i can think of is to go through all nodes and check if there is a reference to it in instance_info | 16:33 |
| rameshg87 | NobodyCam, as you said, may be a tool can do it easily | 16:34 |
| *** Masahiro has joined #openstack-ironic | 16:34 | |
| NobodyCam | rameshg87: Cool :) I think you would be worth it. but would love to hear what others think? | 16:35 |
| rameshg87 | NobodyCam, but can that be pursued as a separate item ? or would you look the tool to be part of this spec itself ? what do you think ? | 16:36 |
| NobodyCam | rameshg87: I'm ok with either. but do want to hear what other are thinking | 16:37 |
| rameshg87 | NobodyCam, okay .. | 16:37 |
| NobodyCam | lucasagomes: j* (others) any thoughts on ^^^^^^^ | 16:38 |
| *** Masahiro has quit IRC | 16:39 | |
| lucasagomes | there's a patch or something doing that? ^ I think I'm a bit out of context | 16:40 |
| NobodyCam | lucasagomes: this is the Enable boot images sharing for iLO driver(s) spec... https://review.openstack.org/#/c/137291/8/specs/kilo/ilo-drivers-share-boot-images.rst | 16:41 |
| *** slagle has quit IRC | 16:41 | |
| *** dhellmann has quit IRC | 16:42 | |
| *** dhellmann has joined #openstack-ironic | 16:42 | |
| *** coolsvap|afk has quit IRC | 16:42 | |
| NobodyCam | lucasagomes: my concern is that ref_count could get out of sync and then a operator would never know | 16:42 |
| lucasagomes | right lemme read the spec | 16:43 |
| NobodyCam | :) Ty | 16:43 |
| *** slagle has joined #openstack-ironic | 16:43 | |
| *** ndipanov has quit IRC | 16:44 | |
| jroll | storing ref count in swift? hmm | 16:45 |
| jroll | yeah, that scares me | 16:45 |
| openstackgerrit | Alexis Lee proposed openstack/ironic: Distinguish between prepare + deploy errors https://review.openstack.org/146135 | 16:45 |
| NobodyCam | jroll: would a tool to check the value make it less scarry? | 16:46 |
| *** coolsvap has joined #openstack-ironic | 16:47 | |
| clif_h | /join #openstack-meeting-3 | 16:47 |
| devananda | g'morning, all | 16:47 |
| clif_h | whoops | 16:47 |
| clif_h | morning ironic :) | 16:47 |
| NobodyCam | good morning devananda | 16:47 |
| NobodyCam | morning clif_h :) | 16:47 |
| *** stendulker has joined #openstack-ironic | 16:48 | |
| jroll | NobodyCam: no, it's http calls to an external service to do ref counting that worries me | 16:49 |
| jroll | morning devananda | 16:49 |
| rameshg87 | jroll, the calls are protected by a lock, so no two threads/conductors could modify at a single point of time | 16:50 |
| jroll | rameshg87: right, but networks fail, swift fails, etc | 16:50 |
| rameshg87 | jroll, but in that case atleast wouldn't we know about it ? | 16:51 |
| jroll | idk, I feel like we could do this without ref counting | 16:51 |
| jroll | rameshg87: maybe :) | 16:51 |
| *** Marga_ has joined #openstack-ironic | 16:51 | |
| jroll | rameshg87: how would you know about it? | 16:51 |
| rameshg87 | jroll, i mean if incrementing ref count in Swift failed due to network error | 16:52 |
| jroll | rameshg87: and as an operator, how would you fix it if one of those calls failed | 16:52 |
| rameshg87 | jroll, the swift client which made the call would fail causing the deploy to fail | 16:52 |
| jroll | rameshg87: right, how would you find that failure? log messages? | 16:52 |
| *** r-daneel has joined #openstack-ironic | 16:52 | |
| jroll | see | 16:52 |
| rameshg87 | jroll, yes | 16:53 |
| jroll | I think this shouldn't make a deploy fail | 16:53 |
| devananda | what refs are we counting? | 16:53 |
| jroll | also, operators don't read every log message | 16:53 |
| JayF | JoshNang: jroll: anything else for IPA update other than HardwareManager change that went in? | 16:53 |
| jroll | also, operators don't have the same lock to "fix" it | 16:53 |
| jroll | JayF: I have nothing | 16:53 |
| jroll | devananda: https://review.openstack.org/#/c/137291/8/specs/kilo/ilo-drivers-share-boot-images.rst | 16:53 |
| JayF | JoshNang: Want to plug your agent client update? | 16:53 |
| JoshNang | JayF: standalone mode! | 16:53 |
| JayF | oh yeah | 16:53 |
| lucasagomes | devananda, https://review.openstack.org/#/c/137291/8/specs/kilo/ilo-drivers-share-boot-images.rst | 16:53 |
| JoshNang | JayF: and yeah, agent client should be pretty close to done | 16:54 |
| jroll | rameshg87: it just seems like this could be really fragile, I'd like to see a better way of doing it | 16:54 |
| jroll | rameshg87: if you'd like, I can elaborate further on the spec | 16:54 |
| lucasagomes | and morning all | 16:55 |
| rameshg87 | jroll, oh okay .. :) | 16:55 |
| devananda | it seems like that should be solved by data deduplication in swift | 16:56 |
| *** ijw has quit IRC | 16:56 | |
| devananda | not reference counting in Ironic | 16:56 |
| rameshg87 | devananda, but at some point of time we would need to clean up the boot image | 16:57 |
| rloo | hi everbody. (I need to scroll back to see the chatter but...) is there a meeting in a few minutes? | 16:57 |
| *** ndipanov has joined #openstack-ironic | 16:57 | |
| jroll | rloo: yep! | 16:57 |
| rameshg87 | devananda, reference counting will help to track how many instances are using that image | 16:57 |
| lucasagomes | hmm yeah, we have to check that building the ISO with the same content results in the exactly same ISO as well | 16:57 |
| rloo | jroll: thx. I looked at the agenda and it doesn't seem to have been updated so I was a bit confused ;) | 16:57 |
| lucasagomes | idk if it will create a hash for each ISO etc... | 16:57 |
| JayF | JoshNang: feel free to add a line to whiteboard about agent client | 16:57 |
| JayF | JoshNang: I added the rest | 16:58 |
| lucasagomes | ideally the ref count should live in our db, but I'm not sure where yet | 16:58 |
| jroll | rloo: maybe there's nothing to talk about, then :P | 16:58 |
| devananda | rameshg87: ref counting in a distributed system is not sustainable | 16:58 |
| lucasagomes | if we had a model for drivers with some extra metadata | 16:58 |
| rameshg87 | devananda, once all the instances using that image are teared down, we wouuld ideally want to cleanup that image | 16:58 |
| rloo | jroll: knowing you guys, doubtful ;) | 16:59 |
| NobodyCam | rameshg87: maybe bring this up in Open dissucision | 16:59 |
| jroll | so we're going to try to do ref counting in a distributed system to save a few gigs of space in swift | 16:59 |
| devananda | rameshg87: how large is each boot image? | 16:59 |
| rameshg87 | NobodyCam, yeah .. if there's time .. | 16:59 |
| dtantsur | rloo, agenda was updated at least a bit - I added one thing | 16:59 |
| dtantsur | (about driver-specific periodic tasks) | 17:00 |
| rameshg87 | devananda, = (kernel size + ramdisk size) for the instance image | 17:00 |
| *** stendulker has quit IRC | 17:00 | |
| *** jay-s-b has joined #openstack-ironic | 17:01 | |
| *** mikedillion has joined #openstack-ironic | 17:02 | |
| *** BertieFulton has joined #openstack-ironic | 17:05 | |
| *** ijw has joined #openstack-ironic | 17:05 | |
| *** mikedillion has quit IRC | 17:06 | |
| *** jay-s-b has quit IRC | 17:06 | |
| *** jay-s-b has joined #openstack-ironic | 17:07 | |
| *** rwsu has joined #openstack-ironic | 17:09 | |
| *** rwsu has quit IRC | 17:09 | |
| *** dlaube has joined #openstack-ironic | 17:16 | |
| *** pcaruana is now known as pcaruana|off| | 17:27 | |
| *** rwsu has joined #openstack-ironic | 17:27 | |
| *** romcheg has quit IRC | 17:29 | |
| *** romcheg has joined #openstack-ironic | 17:30 | |
| *** romcheg has quit IRC | 17:30 | |
| *** ijw has quit IRC | 17:34 | |
| *** andreykurilin has joined #openstack-ironic | 17:34 | |
| *** ijw has joined #openstack-ironic | 17:34 | |
| *** penick has joined #openstack-ironic | 17:46 | |
| *** dlaube has quit IRC | 17:54 | |
| *** Marga_ has quit IRC | 17:55 | |
| NobodyCam | great meeting all | 18:00 |
| NobodyCam | thank you | 18:01 |
| rameshg87 | lucasagomes, what will happen to maintanence ? | 18:01 |
| dtantsur | g'night | 18:01 |
| *** dtantsur is now known as dtantsur|afk | 18:01 | |
| lucasagomes | rameshg87, why? it's a flag not a state | 18:01 |
| rameshg87 | lucasagomes, maintanence=True in manage ? | 18:01 |
| rloo | night dtantsur|afk | 18:01 |
| NobodyCam | night dtantsur|afk | 18:01 |
| rameshg87 | lucasagomes, yeah so will maintanence flag still be there ? | 18:01 |
| devananda | rameshg87: yes | 18:01 |
| lucasagomes | yes it won't change that | 18:01 |
| rameshg87 | devananda, why would we need maintanence flag then ? | 18:02 |
| NobodyCam | what about administer instead of manage? | 18:03 |
| devananda | rameshg87: there may still be periodic tasks which run on nodes in MANAGE[D} state | 18:03 |
| lucasagomes | we need it to indicate that someone (a real person) is/need to look[ing] at that node | 18:03 |
| *** jistr has quit IRC | 18:03 | |
| rameshg87 | lucasagomes, devananda, okay | 18:04 |
| rameshg87 | lucasagomes, devananda, may be things like sync_power_state will still continue to run in managed state ? | 18:04 |
| * NobodyCam has to run and pick up the kids from boarding... maybe even get some starbucks | 18:04 | |
| *** Marga_ has joined #openstack-ironic | 18:05 | |
| NobodyCam | brb | 18:05 |
| lucasagomes | yeah, but even if not... someone may want to put that node in mantenance manually | 18:05 |
| lucasagomes | because he wants to test it or something | 18:05 |
| lucasagomes | so it won't be picked by the scheduler | 18:05 |
| jroll | ++ | 18:05 |
| jroll | we'll always need maintenance flag in every state | 18:05 |
| rameshg87 | lucasagomes, but scheduler will pick only available nodes, right ? | 18:06 |
| jroll | there's a variety of reasons to put nodes in maintenance mode | 18:06 |
| *** jcoufal has joined #openstack-ironic | 18:06 | |
| jroll | maintenance means ironic should not do anything to that node | 18:06 |
| lucasagomes | rameshg87, oh yeah true... | 18:06 |
| lucasagomes | yeah the mantenance is something to indicate to Ironic to not touch that node | 18:07 |
| rameshg87 | jroll, yeah .. | 18:07 |
| lucasagomes | someone is looking at it | 18:07 |
| jroll | like maintenance could mean do not move from managed to available | 18:07 |
| rameshg87 | lucasagomes, okay | 18:07 |
| lucasagomes | so I don't see any harm in keeping it | 18:07 |
| jroll | when JayF tries to change the node I'm debugging to available | 18:07 |
| * rameshg87 goes to sleep | 18:08 | |
| devananda | jroll: +1 | 18:08 |
| rameshg87 | good night everyone :) | 18:08 |
| *** rameshg87 has quit IRC | 18:08 | |
| devananda | rameshg87: g'night! | 18:08 |
| *** spandhe has joined #openstack-ironic | 18:08 | |
| lucasagomes | good night | 18:08 |
| lucasagomes | jroll, lemme see if the config-drive labeling concern is resolved for you at #143510 (when you get a time) | 18:09 |
| *** derekh has quit IRC | 18:10 | |
| jroll | lucasagomes: ah, right, I knew I was missing something there :) | 18:10 |
| *** athomas has quit IRC | 18:11 | |
| lucasagomes | cool | 18:11 |
| *** athomas has joined #openstack-ironic | 18:18 | |
| *** Marga_ has quit IRC | 18:18 | |
| *** Masahiro has joined #openstack-ironic | 18:23 | |
| *** eghobo has joined #openstack-ironic | 18:25 | |
| *** Masahiro has quit IRC | 18:28 | |
| *** ijw has quit IRC | 18:29 | |
| *** Marga_ has joined #openstack-ironic | 18:30 | |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/146156 | 18:37 |
| openstackgerrit | Frederic Lepied proposed stackforge/ironic-discoverd: added an eDeploy plugin https://review.openstack.org/146599 | 18:37 |
| *** harlowja_away is now known as harlowja | 18:39 | |
| openstackgerrit | Frederic Lepied proposed stackforge/ironic-discoverd: added an eDeploy plugin https://review.openstack.org/146599 | 18:43 |
| * NobodyCam is back | 18:48 | |
| *** igordcard has joined #openstack-ironic | 18:52 | |
| *** eghobo has quit IRC | 18:58 | |
| *** harlowja has quit IRC | 19:00 | |
| *** harlowja has joined #openstack-ironic | 19:00 | |
| *** harlowja has quit IRC | 19:00 | |
| *** ijw has joined #openstack-ironic | 19:00 | |
| *** ijw_ has joined #openstack-ironic | 19:02 | |
| NobodyCam | oh the Security impact of http://review.openstack.org/#/c/135845/7/specs/kilo/uefi-secure-boot-management-interfaces.rst seems like a really big issue to me? anyone have anythouhts on it? | 19:04 |
| *** ijw has quit IRC | 19:05 | |
| *** Lotus907efi has joined #openstack-ironic | 19:08 | |
| Lotus907efi | hello all | 19:08 |
| NobodyCam | hello Lotus907efi | 19:08 |
| Lotus907efi | I am trying to pass a kernel command line arg to a baremetal installed overcloud compute node | 19:08 |
| Lotus907efi | I tried using a glance image-update command to add the arg as an extra arg | 19:09 |
| Lotus907efi | but i have been told by glance folks that is only for VMs brought up through libvirt | 19:09 |
| Lotus907efi | and they suggested I talk to the ironic folks since this is a baremetal node | 19:09 |
| Lotus907efi | if I do an ironic node-update can I pass a kernel arg like "intel_iommu=on" to a overcloud compute node that way? | 19:09 |
| Lotus907efi | or is there some other way I can accomplish that??? | 19:09 |
| NobodyCam | Lotus907efi: sounds like youd be intrested in this spec https://review.openstack.org/#/c/144235 | 19:10 |
| Lotus907efi | hmmm, that looks like really recent code | 19:10 |
| Lotus907efi | I am guessing that feature would not be in a juno code release? | 19:11 |
| NobodyCam | correct the spec has not yet landed | 19:11 |
| devananda | Lotus907efi: that's a design spec proposed for kilo | 19:11 |
| Lotus907efi | particularly since it looks like it is still under review | 19:11 |
| NobodyCam | it is | 19:11 |
| Lotus907efi | crud | 19:11 |
| Lotus907efi | so any idea how I can accomplish this in a juno release setup? | 19:11 |
| lucasagomes | Lotus907efi, you can add new kernel cmdlines via Ironic config | 19:12 |
| JayF | You can pass kernel command line options as a global argument, just not per node. | 19:12 |
| Lotus907efi | I have my complete overcloud up and running all on baremetal | 19:12 |
| lucasagomes | there's extra data u can add | 19:12 |
| JayF | or based on per image | 19:12 |
| lucasagomes | lemme find the config | 19:12 |
| Lotus907efi | would I have to completely rebuild/recreate a node or nodes? | 19:12 |
| Lotus907efi | or can I make a running change and then just reboot the node? | 19:13 |
| lucasagomes | Lotus907efi, under [pxe] pxe_append_params | 19:13 |
| *** igordcard has quit IRC | 19:13 | |
| *** corvus is now known as jeblair | 19:13 | |
| lucasagomes | Lotus907efi, you will need to rebuild the instance if it's already deployed | 19:13 |
| lucasagomes | so it regenrate the pxe config files | 19:13 |
| lucasagomes | regenerate* | 19:13 |
| Lotus907efi | nuts | 19:14 |
| devananda | lucasagomes: isn't that a global config? not per-node | 19:14 |
| lucasagomes | oh that's global yea | 19:14 |
| * lucasagomes re-read the whole tihng | 19:14 | |
| devananda | I think Lotus907efi is asking for per-node kernel params, which Juno does not, afaik, support | 19:14 |
| NobodyCam | ya thats global | 19:14 |
| Lotus907efi | I have a separate glance image for all of the overcloud compute nodes | 19:14 |
| lucasagomes | oh per node hmmm | 19:14 |
| Lotus907efi | or would those pxe configs apply across all of the overcloud systems? | 19:15 |
| Lotus907efi | which really would not concern me | 19:15 |
| Lotus907efi | but it would be difficult for me to completely regenerate & rebuild my whole overcloud | 19:15 |
| devananda | Lotus907efi: if you use the pxe_append_params option in the [pxe] configuration group, yes, it would apply to all nodes managed by trhat instance of Ironic | 19:15 |
| Lotus907efi | but I would have to somehow regen all the instances and then rebuild the overcloud? | 19:16 |
| devananda | Lotus907efi: if you're looking for an untested workaround, I think I have an idea | 19:16 |
| * Lotus907efi is all ears | 19:16 | |
| *** Hefeweizen has quit IRC | 19:16 | |
| devananda | Lotus907efi: 1) change that config option for all your ironic-conductor services (eg, in the undercloud) | 19:16 |
| adam_g | not sure it helps now, but there is a spec up to address this to some degree https://review.openstack.org/#/c/144235/ | 19:17 |
| devananda | Lotus907efi: then restart those services so it takes effect for any NEW deploynments | 19:17 |
| devananda | Lotus907efi: then write a script to edit inline all the existing config files | 19:17 |
| devananda | Lotus907efi: for each existing instance | 19:17 |
| Lotus907efi | devananda: hmm, this still sounds a little "hairy" :) | 19:18 |
| devananda | Lotus907efi: adding the option you want to the "append initrd" line of the "label boot" section of the pxe config files | 19:18 |
| devananda | Lotus907efi: then restart each node directly in Ironic | 19:19 |
| devananda | Lotus907efi: oh - totally hairy. and untested :) | 19:19 |
| Lotus907efi | hehehehe | 19:19 |
| devananda | Lotus907efi: not recommending it -- but I think it would work | 19:19 |
| Lotus907efi | I need iommu to be turned on to do some PCI pass through stuff I am trying to accomplish | 19:19 |
| JayF | The other option would be | 19:22 |
| JayF | if oyu need different per-node kernel params in Juno | 19:22 |
| *** pcrews has quit IRC | 19:22 | |
| JayF | and you only need those per-node kernel params in the Instance, not in the ramdisk | 19:22 |
| JayF | you could use IPA w/a full disk image, as the full disk image would have all bootloader config (including kernel command line) baked into the image | 19:22 |
| Lotus907efi | hmmm? | 19:22 |
| JayF | I assume you're using the pxe_* driver in Juno, right? | 19:23 |
| JayF | like pxe_ipmitool | 19:23 |
| Lotus907efi | yes | 19:23 |
| JayF | That supports partial disk images, and always booting via PXE | 19:23 |
| JayF | another deploy driver, agent_ipmitool, supports full disk images (with partitions and bootloaders baked inside) | 19:23 |
| JayF | Since the bootloader is baked into the image, you can configure it with whatever kernel command line you want | 19:24 |
| Lotus907efi | hmmm, yeah but all of my overcloud nodes are currently using qcow2 image that pxe boot the kernel & initrd files | 19:24 |
| JayF | but that'd require reconfiguring your Ironic and changing images, at a minimum | 19:24 |
| Lotus907efi | yeah | 19:24 |
| devananda | Lotus907efi: are you using something like tripleo to deploy/manage your under & over cloud? | 19:25 |
| devananda | Lotus907efi: or phrased another way, how easily can you change settings in the undercloud? | 19:25 |
| Lotus907efi | devananda: where are the pxe config files for each instance stored? | 19:26 |
| Lotus907efi | i.e. how do I find them? | 19:26 |
| devananda | Lotus907efi: default location is /tftpboot -- but check your ironic config file for tftp_root and tftp_master_path config options | 19:26 |
| * lucasagomes wants to find time to work on the ipxe driver, so the generation of config would be dynamic, all that would be needed for this case ^ would be to reboot the instances | 19:46 | |
| Lotus907efi | that would be cool | 19:48 |
| lucasagomes | folks have you guys seem sfaizan (Syed Ismail Faizan) online lately? | 19:48 |
| devananda | lucasagomes: I have not | 19:50 |
| NobodyCam | nore I | 19:51 |
| lucasagomes | :( right thanks | 19:52 |
| *** harlowja has joined #openstack-ironic | 19:52 | |
| lucasagomes | I wanna to talk to him about his boot from local disk spec, I made a PoC slightly diff than what he proposed (I need it so I couldn't wait). And it's working, for those interested the patches are: https://review.openstack.org/146619 (Nova); https://review.openstack.org/146097 (DIB); https://review.openstack.org/146189 (Ironic) | 19:54 |
| *** david-lyle has quit IRC | 20:02 | |
| *** Marga_ has quit IRC | 20:05 | |
| *** Masahiro has joined #openstack-ironic | 20:12 | |
| *** morgabra has quit IRC | 20:13 | |
| *** morgabra has joined #openstack-ironic | 20:15 | |
| *** Masahiro has quit IRC | 20:16 | |
| *** ryanpetrello has quit IRC | 20:17 | |
| *** ryanpetrello has joined #openstack-ironic | 20:17 | |
| lucasagomes | folks, I will call it a day | 20:21 |
| openstackgerrit | Adam Gandelman proposed openstack/ironic: Provided backward compat for enforcing admin policy https://review.openstack.org/145984 | 20:21 |
| lucasagomes | have a great night! | 20:21 |
| *** lucasagomes is now known as lucas-dinner | 20:21 | |
| *** ijw_ has quit IRC | 20:22 | |
| *** rloo has quit IRC | 20:22 | |
| *** pcrews has joined #openstack-ironic | 20:23 | |
| *** ijw has joined #openstack-ironic | 20:23 | |
| *** aweeks has quit IRC | 20:23 | |
| *** rainya has quit IRC | 20:23 | |
| *** victor_lowther has quit IRC | 20:23 | |
| *** rloo has joined #openstack-ironic | 20:23 | |
| *** PaulCzar has quit IRC | 20:23 | |
| *** bigjools has quit IRC | 20:23 | |
| *** russell_h has quit IRC | 20:24 | |
| *** wendar has quit IRC | 20:24 | |
| *** wendar has joined #openstack-ironic | 20:25 | |
| *** krtaylor has quit IRC | 20:25 | |
| *** jroll has quit IRC | 20:26 | |
| *** morgabra has quit IRC | 20:26 | |
| *** morgabra has joined #openstack-ironic | 20:26 | |
| *** Ng has quit IRC | 20:26 | |
| *** victor_lowther has joined #openstack-ironic | 20:26 | |
| *** Guest58319 has quit IRC | 20:27 | |
| *** aweeks has joined #openstack-ironic | 20:27 | |
| NobodyCam | night lucas-dinner | 20:28 |
| *** PaulCzar has joined #openstack-ironic | 20:28 | |
| *** ijw has quit IRC | 20:28 | |
| *** bigjools has joined #openstack-ironic | 20:29 | |
| *** mgagne has joined #openstack-ironic | 20:29 | |
| *** rainya has joined #openstack-ironic | 20:29 | |
| *** russell_h has joined #openstack-ironic | 20:29 | |
| *** mgagne is now known as Guest36580 | 20:30 | |
| *** jroll has joined #openstack-ironic | 20:30 | |
| *** russell_h has quit IRC | 20:30 | |
| *** russell_h has joined #openstack-ironic | 20:30 | |
| *** andreykurilin has quit IRC | 20:31 | |
| *** krtaylor has joined #openstack-ironic | 20:31 | |
| *** Ng has joined #openstack-ironic | 20:31 | |
| *** Marga_ has joined #openstack-ironic | 20:36 | |
| *** alexpilotti has quit IRC | 20:41 | |
| *** Marga_ has quit IRC | 20:41 | |
| *** alexm__ has joined #openstack-ironic | 20:41 | |
| alexm__ | Hi guys, base ironic with pxe_ipmitool driver works well, but the agent_ipmitool just loops infinitely during deploy phase. Anything can explain that? | 20:45 |
| NobodyCam | alexm__: need a little more to work with there | 20:47 |
| NobodyCam | where are you looping? | 20:48 |
| *** Nisha has quit IRC | 20:48 | |
| alexm__ | NobodyCam: CoreOS seems to boot fine until the ssh key phase, then it just loops on the NIC changing the MTU from 1500 to 1400 back and forth | 20:51 |
| NobodyCam | oh that a new one for /me... maybe one of the IPA folk have seen that before? | 20:52 |
| *** pcrews has quit IRC | 20:52 | |
| *** Marga_ has joined #openstack-ironic | 20:54 | |
| *** david-lyle has joined #openstack-ironic | 20:57 | |
| alexm__ | I see that the LVM tasks fail during boot | 20:58 |
| alexm__ | I used the pre-built ipa ramdisk images, will try to build them from scratch | 21:02 |
| *** andreykurilin has joined #openstack-ironic | 21:02 | |
| *** Guest36580 is now known as mgagne | 21:04 | |
| *** mgagne has joined #openstack-ironic | 21:04 | |
| openstackgerrit | Frederic Lepied proposed stackforge/ironic-discoverd: added an eDeploy plugin https://review.openstack.org/146599 | 21:11 |
| *** BertieFulton has quit IRC | 21:13 | |
| *** Marga_ has quit IRC | 21:15 | |
| *** david-lyle has quit IRC | 21:16 | |
| *** david-ly_ has joined #openstack-ironic | 21:16 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-python-agent: Error classes invoke their parent's __init__() https://review.openstack.org/146643 | 21:17 |
| *** Marga_ has joined #openstack-ironic | 21:22 | |
| *** yjiang5 has quit IRC | 21:25 | |
| *** pcrews has joined #openstack-ironic | 21:28 | |
| *** yjiang5 has joined #openstack-ironic | 21:32 | |
| *** dprince has quit IRC | 21:33 | |
| *** eghobo has joined #openstack-ironic | 21:34 | |
| *** jroll has quit IRC | 21:38 | |
| *** jroll has joined #openstack-ironic | 21:38 | |
| *** david-ly_ has quit IRC | 21:40 | |
| *** david-lyle has joined #openstack-ironic | 21:41 | |
| *** david-lyle has quit IRC | 21:49 | |
| *** david-lyle has joined #openstack-ironic | 21:50 | |
| *** jgrimm is now known as zz_jgrimm | 21:52 | |
| jjohnson2 | NobodyCam, get a chance to ponder whether the interface makes sense or if I should change? | 21:55 |
| NobodyCam | jjohnson2: I haven't sorry... also on a conf call atm? oh mondays | 21:59 |
| *** eghobo has quit IRC | 22:00 | |
| *** Masahiro has joined #openstack-ironic | 22:01 | |
| jjohnson2 | no problem | 22:01 |
| jjohnson2 | I'm about to add oem extensibility on client side | 22:01 |
| NobodyCam | :) | 22:01 |
| jjohnson2 | because I have an oem extension to implement | 22:01 |
| jjohnson2 | but time to head home | 22:02 |
| *** eghobo has joined #openstack-ironic | 22:02 | |
| *** ifarkas has quit IRC | 22:05 | |
| *** Masahiro has quit IRC | 22:05 | |
| *** andreykurilin has quit IRC | 22:06 | |
| *** andreykurilin has joined #openstack-ironic | 22:07 | |
| jroll | 19:04:56 NobodyCam | oh the Security impact of http://review.openstack.org/#/c/135845/7/specs/kilo/uefi-secure-boot-management-interfaces.rst seems like a really big issue to me? <- commented on the review, but that isn't a security issue at all, nor is it relevant to this spec imo | 22:09 |
| *** jjohnson2 has quit IRC | 22:10 | |
| *** Marga_ has quit IRC | 22:12 | |
| NobodyCam | I had a comment (not sure it got posted) that looked like a deployer impact.. but I was taken back buy the " | 22:13 |
| NobodyCam | Ironic does not have and mechanism to know if the deployed node has booted up successfully" | 22:14 |
| NobodyCam | comment | 22:14 |
| *** pelix has quit IRC | 22:17 | |
| *** r-daneel has quit IRC | 22:23 | |
| jroll | NobodyCam: that's true today, for all deployment methods | 22:24 |
| jroll | (without putting some sort of agent etc in your images) | 22:24 |
| *** Marga_ has joined #openstack-ironic | 22:27 | |
| NobodyCam | :/ | 22:27 |
| *** martini has joined #openstack-ironic | 22:34 | |
| NobodyCam | brb | 22:41 |
| *** jcoufal has quit IRC | 22:43 | |
| *** eghobo has quit IRC | 22:43 | |
| *** Marga_ has quit IRC | 22:49 | |
| *** eghobo has joined #openstack-ironic | 22:51 | |
| *** ryanpetrello_ has joined #openstack-ironic | 22:56 | |
| *** ryanpetrello has quit IRC | 22:59 | |
| *** ryanpetrello_ is now known as ryanpetrello | 22:59 | |
| *** penick has quit IRC | 23:00 | |
| *** Marga_ has joined #openstack-ironic | 23:00 | |
| *** athomas has quit IRC | 23:02 | |
| *** penick has joined #openstack-ironic | 23:04 | |
| devananda | if anyone hasn't voted in the openstack board and bylaws changes -- please go do it! | 23:17 |
| devananda | its actually really important we pass the amendments to the bylaws | 23:17 |
| *** david-lyle has quit IRC | 23:20 | |
| *** Haomeng has joined #openstack-ironic | 23:27 | |
| *** chlong has joined #openstack-ironic | 23:29 | |
| *** spandhe has quit IRC | 23:35 | |
| *** andreykurilin has quit IRC | 23:35 | |
| *** dlaube has joined #openstack-ironic | 23:37 | |
| *** spandhe has joined #openstack-ironic | 23:43 | |
| *** lucas-dinner has quit IRC | 23:46 | |
| *** martini has quit IRC | 23:47 | |
| *** Masahiro has joined #openstack-ironic | 23:49 | |
| jroll | lucas' stuff for local boot looks good | 23:50 |
| * jroll tries to remember who was working on pxe boot for ipa | 23:50 | |
| *** Marga_ has quit IRC | 23:51 | |
| *** naohirot has joined #openstack-ironic | 23:54 | |
| *** Masahiro has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!