*** rongze has quit IRC | 00:00 | |
*** NobodyCam_afk is now known as NobodyCam | 00:04 | |
NobodyCam | and back | 00:04 |
---|---|---|
NobodyCam | this is cool stuff just not sure I like it. http://www.boston.com/business/technology/2013/12/02/amazon-com-developing-delivery-drones/0AhjTYBG83ZEnDb4uA50DN/story.html | 00:09 |
*** datajerk has quit IRC | 00:12 | |
NobodyCam | great meeting guys :) | 00:21 |
*** matsuhashi has joined #openstack-ironic | 00:29 | |
*** datajerk has joined #openstack-ironic | 00:32 | |
*** kui has quit IRC | 00:47 | |
*** matsuhashi has quit IRC | 00:50 | |
*** matsuhashi has joined #openstack-ironic | 00:50 | |
*** michchap_ has quit IRC | 00:53 | |
*** matsuhas_ has joined #openstack-ironic | 00:53 | |
*** michchap has joined #openstack-ironic | 00:53 | |
*** matsuhashi has quit IRC | 00:54 | |
*** datajerk has quit IRC | 01:06 | |
*** nosnos has joined #openstack-ironic | 01:26 | |
*** rongze has joined #openstack-ironic | 01:27 | |
*** rongze has quit IRC | 01:31 | |
*** jcooley_ has joined #openstack-ironic | 01:37 | |
*** sjing has joined #openstack-ironic | 01:39 | |
*** kui has joined #openstack-ironic | 01:43 | |
*** jcooley_ has quit IRC | 01:45 | |
*** datajerk has joined #openstack-ironic | 01:49 | |
*** arata has joined #openstack-ironic | 01:54 | |
*** datajerk has quit IRC | 01:57 | |
*** jcooley_ has joined #openstack-ironic | 02:01 | |
*** jcooley_ has quit IRC | 02:02 | |
*** rongze has joined #openstack-ironic | 02:06 | |
*** vkozhukalov has joined #openstack-ironic | 02:19 | |
*** datajerk has joined #openstack-ironic | 02:25 | |
*** datajerk has quit IRC | 02:29 | |
*** matsuhas_ has quit IRC | 02:31 | |
*** matsuhashi has joined #openstack-ironic | 02:31 | |
*** datajerk has joined #openstack-ironic | 02:56 | |
Haomeng | devananda: ping | 02:56 |
devananda | Haomeng: pong | 02:57 |
*** nosnos has quit IRC | 02:58 | |
Haomeng | devananda: just see yesterday's IRC chat history, found you ping me at my local time around 1:30am:) | 02:58 |
devananda | Haomeng: regarding https://review.openstack.org/#/c/53627/ | 02:58 |
devananda | Haomeng: are you going to continue working on this patch? | 02:58 |
*** nosnos has joined #openstack-ironic | 02:59 | |
*** ndipanov has quit IRC | 02:59 | |
Haomeng | devananda: ok, let me check first | 02:59 |
Haomeng | devananda: yes, I try debug this issue, but still can not find the root cause why the node.patch_json() isn't populating node.updated_at, but for other object such as port and chasiss can work for patch_json to update updated_at field value | 03:00 |
Haomeng | devananda: I think maybe it will be our conductor issue with unittest | 03:00 |
Haomeng | devananda: node update will call conductor code, this is difference with other objects | 03:01 |
*** datajerk has quit IRC | 03:01 | |
Haomeng | devananda: I think for unittest, our conductor is disabled, and we have no such node update validation case in our existing unittest code | 03:02 |
devananda | Haomeng: afaik, the API unit tests are not using a conductor service, RPC layer, or database | 03:03 |
Haomeng | devananda: yes, I think so, so for such case, I dont know how to handle | 03:04 |
Haomeng | devananda: if we just mock the update result, it just tests mocking | 03:04 |
Haomeng | devananda: so can we restore it and add my comments to take next round review? | 03:04 |
devananda | Haomeng: I -1'd because the patch is not comparing time validly, not because of whether the time is real or not :) | 03:05 |
devananda | it is mocking many things so taht we can test the API code paths independently from other components | 03:05 |
devananda | that is the nature of a unit test -- test only one component at a time, and mock everything else | 03:05 |
Haomeng | devananda: yes | 03:06 |
Haomeng | devananda: let me try to modify the unittest code to test our api only | 03:06 |
devananda | thanks :) | 03:06 |
Haomeng | devananda: any time:) | 03:06 |
Haomeng | devananda: will contiue to work on this issue today:) thank for your reminder:) | 03:07 |
Haomeng | devananda: one more question, how do you think if I need to mock our conductor for node update unittest case? | 03:18 |
devananda | Haomeng: mock the rpcapi layer | 03:18 |
Haomeng | devananda: ok | 03:18 |
Haomeng | devananda: thank you:) | 03:18 |
devananda | Haomeng: so that you can fake what the conductor would return, if there were an RPC service running | 03:18 |
devananda | welcome :) | 03:18 |
Haomeng | devananda: got it | 03:19 |
Haomeng | devananda: enjoy your night:) | 03:19 |
*** datajerk has joined #openstack-ironic | 03:26 | |
*** jcooley_ has joined #openstack-ironic | 03:29 | |
*** datajerk has quit IRC | 03:31 | |
openstackgerrit | A change was merged to openstack/ironic: Apply comments to 58558/4 post-landing https://review.openstack.org/59357 | 03:33 |
openstackgerrit | A change was merged to openstack/ironic: Allow Pecan to use 'debuginfo' response field https://review.openstack.org/58807 | 03:33 |
openstackgerrit | A change was merged to openstack/ironic: Replace removed xrange in Python3 https://review.openstack.org/59114 | 03:36 |
*** matsuhashi has quit IRC | 03:42 | |
*** arata has left #openstack-ironic | 03:55 | |
*** datajerk has joined #openstack-ironic | 03:57 | |
*** rloo has quit IRC | 04:01 | |
*** datajerk has quit IRC | 04:01 | |
*** michchap has quit IRC | 04:21 | |
*** michchap has joined #openstack-ironic | 04:21 | |
*** matsuhashi has joined #openstack-ironic | 04:25 | |
*** datajerk has joined #openstack-ironic | 04:27 | |
*** rongze has quit IRC | 04:31 | |
*** datajerk has quit IRC | 04:32 | |
*** datajerk has joined #openstack-ironic | 04:58 | |
*** matsuhashi has quit IRC | 05:00 | |
*** datajerk has quit IRC | 05:03 | |
*** matsuhashi has joined #openstack-ironic | 05:08 | |
*** matsuhashi has quit IRC | 05:27 | |
*** datajerk has joined #openstack-ironic | 05:29 | |
*** datajerk has quit IRC | 05:33 | |
*** matsuhashi has joined #openstack-ironic | 05:59 | |
openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/59627 | 06:00 |
*** prekarat has quit IRC | 06:08 | |
*** retr0h has quit IRC | 06:14 | |
*** retr0h has joined #openstack-ironic | 06:14 | |
*** arata has joined #openstack-ironic | 06:15 | |
*** datajerk has joined #openstack-ironic | 06:30 | |
*** yfujioka has joined #openstack-ironic | 06:31 | |
*** datajerk has quit IRC | 06:35 | |
*** vkozhukalov has quit IRC | 06:47 | |
*** datajerk has joined #openstack-ironic | 07:01 | |
*** datajerk has quit IRC | 07:07 | |
*** datajerk has joined #openstack-ironic | 07:32 | |
*** jcooley_ has quit IRC | 07:35 | |
*** jcooley_ has joined #openstack-ironic | 07:36 | |
*** datajerk has quit IRC | 07:37 | |
GheRivero | morning Ironic! | 07:54 |
*** datajerk has joined #openstack-ironic | 08:03 | |
Haomeng | morning:) | 08:06 |
*** datajerk has quit IRC | 08:08 | |
GheRivero | morning Haomeng | 08:13 |
Haomeng | GheRivero: :) | 08:13 |
*** vkozhukalov has joined #openstack-ironic | 08:27 | |
*** datajerk has joined #openstack-ironic | 08:34 | |
*** jcooley_ has quit IRC | 08:36 | |
*** jcooley_ has joined #openstack-ironic | 08:38 | |
*** datajerk has quit IRC | 08:38 | |
*** jistr has joined #openstack-ironic | 08:48 | |
*** sjing has quit IRC | 08:49 | |
*** derekh has joined #openstack-ironic | 08:57 | |
*** romcheg has joined #openstack-ironic | 09:00 | |
*** lynxman has quit IRC | 09:02 | |
*** datajerk has joined #openstack-ironic | 09:05 | |
*** romcheg has quit IRC | 09:05 | |
*** ndipanov has joined #openstack-ironic | 09:08 | |
*** datajerk has quit IRC | 09:10 | |
*** romcheg has joined #openstack-ironic | 09:12 | |
*** max_lobur_afk is now known as max_lobur | 09:14 | |
*** martyntaylor has joined #openstack-ironic | 09:19 | |
openstackgerrit | Max Lobur proposed a change to openstack/ironic: Add missing link for drivers resource https://review.openstack.org/59081 | 09:24 |
*** lucasagomes has joined #openstack-ironic | 09:32 | |
*** jcooley_ has quit IRC | 09:34 | |
*** datajerk has joined #openstack-ironic | 09:36 | |
*** agordeev has joined #openstack-ironic | 09:38 | |
*** datajerk has quit IRC | 09:40 | |
*** lynxman has joined #openstack-ironic | 09:43 | |
*** kushi_ has joined #openstack-ironic | 09:50 | |
*** matsuhashi has quit IRC | 09:51 | |
*** martyntaylor1 has joined #openstack-ironic | 09:52 | |
*** kui has quit IRC | 09:52 | |
*** martyntaylor has quit IRC | 09:53 | |
*** matsuhashi has joined #openstack-ironic | 09:54 | |
*** kushi_ has quit IRC | 09:55 | |
*** jcooley_ has joined #openstack-ironic | 10:05 | |
*** datajerk has joined #openstack-ironic | 10:06 | |
*** datajerk has quit IRC | 10:11 | |
*** jcooley_ has quit IRC | 10:12 | |
*** yfujioka has quit IRC | 10:18 | |
openstackgerrit | Dmitry Shulyak proposed a change to openstack/ironic: [PROOF OF CONCEPT]Allow usage of alembic tool for migrations https://review.openstack.org/59668 | 10:25 |
*** rongze has joined #openstack-ironic | 10:28 | |
openstackgerrit | Dmitry Shulyak proposed a change to openstack/ironic: [PROOF OF CONCEPT]Allow usage of alembic tool for migrations https://review.openstack.org/59206 | 10:32 |
openstackgerrit | Dmitry Shulyak proposed a change to openstack/ironic: Usefull tools for alembic usage https://review.openstack.org/59672 | 10:42 |
*** rongze has quit IRC | 10:46 | |
*** prekarat has joined #openstack-ironic | 10:46 | |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: If no block devices abort deployment https://review.openstack.org/59493 | 11:02 |
*** rongze has joined #openstack-ironic | 11:03 | |
*** jcooley_ has joined #openstack-ironic | 11:05 | |
*** jcooley_ has quit IRC | 11:10 | |
*** kui has joined #openstack-ironic | 11:17 | |
*** prekarat has quit IRC | 11:22 | |
*** rongze has quit IRC | 11:39 | |
*** arata has left #openstack-ironic | 11:44 | |
*** jcooley_ has joined #openstack-ironic | 11:51 | |
*** rongze_ has joined #openstack-ironic | 11:56 | |
*** romcheg has quit IRC | 12:03 | |
openstackgerrit | Dmitry Shulyak proposed a change to openstack/ironic: Usefull tools for alembic usage https://review.openstack.org/59672 | 12:03 |
*** romcheg has joined #openstack-ironic | 12:04 | |
*** max_lobur has quit IRC | 12:05 | |
*** max_lobur has joined #openstack-ironic | 12:05 | |
openstackgerrit | Dmitry Shulyak proposed a change to openstack/ironic: Useful tools for alembic usage https://review.openstack.org/59672 | 12:10 |
*** jcooley_ has quit IRC | 12:25 | |
Haomeng | good morning/evening Ironic:) | 12:26 |
*** datajerk has joined #openstack-ironic | 12:41 | |
*** lucasagomes is now known as lucas-hungry | 12:44 | |
*** datajerk has quit IRC | 12:46 | |
*** prekarat has joined #openstack-ironic | 12:57 | |
*** SpamapS has quit IRC | 12:59 | |
*** jcooley_ has joined #openstack-ironic | 13:08 | |
openstackgerrit | Max Lobur proposed a change to openstack/python-ironicclient: Let CLI print exception traceback from 'debuginfo' https://review.openstack.org/59038 | 13:09 |
*** datajerk has joined #openstack-ironic | 13:12 | |
*** jcooley_ has quit IRC | 13:13 | |
*** datajerk has quit IRC | 13:17 | |
*** jdob has joined #openstack-ironic | 13:23 | |
*** jcooley_ has joined #openstack-ironic | 13:23 | |
openstackgerrit | Max Lobur proposed a change to openstack/python-ironicclient: Let CLI print exception traceback from 'debuginfo' https://review.openstack.org/59038 | 13:26 |
*** jcooley_ has quit IRC | 13:28 | |
*** jdob has quit IRC | 13:35 | |
*** datajerk has joined #openstack-ironic | 13:36 | |
*** jdob has joined #openstack-ironic | 13:37 | |
*** kui has quit IRC | 13:37 | |
*** matsuhashi has quit IRC | 13:40 | |
*** jdob has quit IRC | 13:51 | |
*** jdob has joined #openstack-ironic | 13:52 | |
*** yuriyz has joined #openstack-ironic | 14:00 | |
*** lucas-hungry is now known as lucasagomes | 14:04 | |
*** jcooley_ has joined #openstack-ironic | 14:17 | |
*** nosnos has quit IRC | 14:20 | |
*** matty_dubs|gone is now known as matty_dubs | 14:21 | |
*** jcooley_ has quit IRC | 14:21 | |
*** rloo has joined #openstack-ironic | 14:25 | |
*** rongze_ has quit IRC | 14:27 | |
*** linggao has joined #openstack-ironic | 14:29 | |
*** linggao has quit IRC | 14:34 | |
*** linggao has joined #openstack-ironic | 14:34 | |
*** jbjohnso has joined #openstack-ironic | 14:35 | |
*** rongze has joined #openstack-ironic | 14:38 | |
*** romcheg has quit IRC | 14:44 | |
*** romcheg has joined #openstack-ironic | 14:46 | |
*** linggao_ has joined #openstack-ironic | 14:47 | |
*** linggao_ has quit IRC | 14:47 | |
*** datajerk has quit IRC | 14:56 | |
*** romcheg has quit IRC | 15:06 | |
*** jbjohnso has quit IRC | 15:06 | |
*** romcheg has joined #openstack-ironic | 15:13 | |
*** agordeev has quit IRC | 15:16 | |
*** datajerk has joined #openstack-ironic | 15:16 | |
*** jcooley_ has joined #openstack-ironic | 15:17 | |
*** jbjohnso has joined #openstack-ironic | 15:20 | |
*** jcooley_ has quit IRC | 15:21 | |
*** viktors has joined #openstack-ironic | 15:23 | |
NobodyCam | Good Morning iRonic, says the man waiting for coffee | 15:24 |
GheRivero | uhm... coffee, I need it. Morning NobodyCam | 15:25 |
NobodyCam | Morning GheRivero :) | 15:25 |
yuriyz | Morning Ironic | 15:25 |
NobodyCam | morning yuriyz | 15:25 |
*** datajerk1 has joined #openstack-ironic | 15:26 | |
romcheg | Morning all! | 15:27 |
NobodyCam | morning romcheg :) | 15:27 |
*** datajerk has quit IRC | 15:29 | |
lucasagomes | morning NobodyCam | 15:30 |
lucasagomes | romcheg yuriyz and GheRivero | 15:30 |
NobodyCam | hey hey good afternoon lucasagomes :-p | 15:30 |
lucasagomes | NobodyCam, hey ya | 15:30 |
lucasagomes | gotta talk to u | 15:31 |
NobodyCam | sure | 15:31 |
NobodyCam | gtalk | 15:31 |
lucasagomes | cool | 15:31 |
NobodyCam | I have no coffee yet | 15:31 |
lucasagomes | np go grab it :) | 15:32 |
*** romcheg has quit IRC | 15:33 | |
*** romcheg has joined #openstack-ironic | 15:34 | |
NobodyCam | :) ok got some coffee | 15:35 |
*** romcheg has quit IRC | 15:43 | |
*** vkozhukalov has quit IRC | 15:45 | |
*** SpamapS has joined #openstack-ironic | 15:51 | |
NobodyCam | lucasagomes: I think you may be right, Pxe deploy should prob be doing the power control | 15:52 |
NobodyCam | GheRivero: ^^^ do you have any thoughts on ^^^ | 15:52 |
lucasagomes | NobodyCam, yup, def there's a mismatch between that diagram and the current code | 15:52 |
NobodyCam | GheRivero: diagram is ( https://docs.google.com/drawings/d/1azAWh0ZfhDqEUsC14ZEBawbnAmdQ2_yl3CfOdDbPvOk ) | 15:53 |
* GheRivero looking at the diagram | 15:56 | |
GheRivero | where does this diagram come from? | 15:58 |
NobodyCam | deva | 15:58 |
GheRivero | yeah, the pxe driver doesn-t do the power control phase now | 16:00 |
*** romcheg has joined #openstack-ironic | 16:00 | |
GheRivero | but it's easy to adpat to the diagram | 16:02 |
GheRivero | do we have an idea about how to cancel deploys? | 16:02 |
devananda | mornin, all | 16:03 |
*** jcooley_ has joined #openstack-ironic | 16:04 | |
*** tatyana has joined #openstack-ironic | 16:05 | |
lucasagomes | GheRivero, not yet | 16:07 |
lucasagomes | devananda, morning | 16:07 |
max_lobur | morning Ironic | 16:11 |
NobodyCam | morning max_lobur and devananda | 16:12 |
lucasagomes | hey max_lobur | 16:12 |
NobodyCam | I need to update the dib walk thru today | 16:16 |
NobodyCam | :-p | 16:16 |
dkehn | NobodyCam: in the merge.py portion of the wike its referencing a $TRIPLEO_ROOT/tripleo-heat-templates/undercloud-vm.yaml should that be a /tripleo-heat-templates/undercloud-bm-source.yaml | 16:18 |
NobodyCam | are you using real hardware? | 16:19 |
*** jcooley_ has quit IRC | 16:19 | |
*** jcooley_ has joined #openstack-ironic | 16:19 | |
*** jcooley_ has quit IRC | 16:20 | |
romcheg | Finally I rebased the changes to infra config. | 16:20 |
NobodyCam | romcheg: w00t | 16:20 |
romcheg | Now I think they can be merged | 16:20 |
yuriyz | people, workinprogress patch for Ironic deploy ramdisk https://review.openstack.org/59770 | 16:21 |
dkehn | NobodyCam: got it | 16:21 |
dkehn | NobodyCam: nope should layout into VMNs | 16:21 |
dkehn | VMs | 16:21 |
NobodyCam | romcheg: yuriyz how is the revolution going? | 16:22 |
NobodyCam | so undercloud-vm.yaml :) | 16:22 |
romcheg | NobodyCam: not as good as we want :) | 16:22 |
NobodyCam | :( | 16:22 |
romcheg | Today there was a chance to overthrow the government peacefully but that didn't work | 16:22 |
* romcheg gets hit crowbar ready | 16:23 | |
NobodyCam | lol | 16:23 |
* NobodyCam makes note not ot upset romcheg | 16:24 | |
*** tatyana has quit IRC | 16:24 | |
dkehn | NobodyCam: undercloud-vm.yaml == undercloud-vm-source.yaml? | 16:26 |
NobodyCam | ya looks like it was renamed | 16:26 |
lucasagomes | devananda, ping | 16:29 |
devananda | lucasagomes: pong | 16:29 |
NobodyCam | morning devananda :) | 16:29 |
lucasagomes | devananda, I saw that u've a patch that implements a periodic task to check the power state of the nodes and update the db | 16:30 |
lucasagomes | right now we have something like GET /nodes/<uuid>/state/ that would issue an RPC call and get the state | 16:30 |
lucasagomes | but as we already have this periodic task makes it make more sense to just return what is in the db | 16:31 |
lucasagomes | what u think? | 16:31 |
devananda | lucasagomes: i think there will be some optimization we can do :) | 16:32 |
lucasagomes | right | 16:32 |
devananda | lucasagomes: i'm nearly done refactoring that patch chain | 16:32 |
devananda | should be able to push once i finish a few unit tests | 16:32 |
lucasagomes | because I our current states are kinda over engineered | 16:32 |
lucasagomes | devananda, ok | 16:32 |
lucasagomes | for e.g... right now we have: | 16:33 |
lucasagomes | GET /nodes/<uuid>/state => show both power and provision current states | 16:33 |
lucasagomes | GET /nodes/<uuid>/state/power => show the current and target power state | 16:33 |
lucasagomes | GET /nodes/<uuid>state/provision => show the current and target provision state | 16:33 |
lucasagomes | why not just have /nodes/<uuid>/state to show all the current and target states + last_error | 16:34 |
devananda | lucasagomes: because we need POST for state/power and state/provision ? *shrug* | 16:34 |
*** jcooley_ has joined #openstack-ironic | 16:35 | |
lucasagomes | we still can do that | 16:35 |
lucasagomes | but not having all GETs separated | 16:35 |
*** datajerk1 has quit IRC | 16:35 | |
devananda | i see your point | 16:35 |
*** datajerk has joined #openstack-ironic | 16:43 | |
*** datajerk1 has joined #openstack-ironic | 16:45 | |
*** datajerk has quit IRC | 16:47 | |
*** jcooley_ has quit IRC | 16:47 | |
*** datajerk1 has quit IRC | 16:51 | |
*** SpamapS has quit IRC | 16:52 | |
*** datajerk has joined #openstack-ironic | 16:52 | |
*** hemna has joined #openstack-ironic | 16:54 | |
NobodyCam | morning walkies time... brb | 17:00 |
*** SpamapS has joined #openstack-ironic | 17:05 | |
*** SpamapS has quit IRC | 17:05 | |
*** SpamapS has joined #openstack-ironic | 17:05 | |
*** jistr has quit IRC | 17:10 | |
*** jcooley_ has joined #openstack-ironic | 17:12 | |
*** jbjohnso has quit IRC | 17:15 | |
*** jcooley_ has quit IRC | 17:15 | |
*** jcooley_ has joined #openstack-ironic | 17:16 | |
*** jcooley_ has quit IRC | 17:20 | |
dkehn | NobodyCam: post merge.py runing into an error were the AdminPassword is not defined in template, I see it in the ripleo-heat-templates/undercloud-source.yaml, but I don't think its in the undercloud-vm-source.yaml, thoughts | 17:22 |
*** jcooley_ has joined #openstack-ironic | 17:23 | |
NobodyCam | dkehn: I'm going thru and updating the walk thru now. Let me see what has changed in the past week or so | 17:23 |
dkehn | NobodyCam: k | 17:23 |
NobodyCam | ahh ok | 17:26 |
*** ndipanov has quit IRC | 17:26 | |
NobodyCam | dkehn: yes things have changed a bit | 17:26 |
NobodyCam | but looks like I got the ironic yaml creation in the make file | 17:27 |
NobodyCam | https://github.com/openstack/tripleo-heat-templates/blob/master/Makefile | 17:27 |
NobodyCam | so | 17:27 |
*** ndipanov has joined #openstack-ironic | 17:27 | |
NobodyCam | the dev test walk thru does make -C $TRIPLEO_ROOT/tripleo-heat-templates undercloud-vm.yaml | 17:27 |
NobodyCam | which should also creat the ironic yaml | 17:28 |
*** ndipanov has quit IRC | 17:28 | |
NobodyCam | *create even | 17:28 |
dkehn | NobodyCam: so would it be better to do the devtest.sh than the route I'm going? | 17:28 |
*** ndipanov has joined #openstack-ironic | 17:28 | |
dkehn | NobodyCam: and just use the underclound for dev purposes,? | 17:29 |
*** datajerk has quit IRC | 17:29 | |
*** jbjohnso has joined #openstack-ironic | 17:32 | |
*** datajerk has joined #openstack-ironic | 17:32 | |
*** viktors has left #openstack-ironic | 17:32 | |
*** matty_dubs is now known as matty_dubs|lunch | 17:35 | |
*** datajerk has quit IRC | 17:36 | |
devananda | lucasagomes: patches INC. should be easier to review in small bites, I hope :) | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add missing hash_partition_exponent config option https://review.openstack.org/59793 | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add config option for # of conductor replicas https://review.openstack.org/59794 | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: list_active_conductor_drivers returns dict of sets https://review.openstack.org/59795 | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add heartbeat_interval to new 'conductor' cfg group https://review.openstack.org/58851 | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement sync_power_state periodic task https://review.openstack.org/56530 | 17:37 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add prepare and cleanup methods to deploy interface https://review.openstack.org/58894 | 17:37 |
lucasagomes | devananda, wow great | 17:37 |
NobodyCam | dkehn: I have run the devtest script. I have been doing the steps manually. I do that because I do not need the over cloud | 17:37 |
*** max_lobur is now known as max_lobur_afk | 17:37 | |
*** derekh has quit IRC | 17:41 | |
NobodyCam | doh... I have "NOT" run the devtest script it self..... | 17:42 |
*** tatyana has joined #openstack-ironic | 17:43 | |
*** datajerk has joined #openstack-ironic | 17:45 | |
*** martyntaylor1 has quit IRC | 17:52 | |
*** datajerk has quit IRC | 17:55 | |
*** vkozhukalov has joined #openstack-ironic | 17:55 | |
*** rongze has quit IRC | 17:58 | |
dkehn | NobodyCam: the question is should I be running the devtest.sh, I can always put an exit in before the overcloud part starts | 17:59 |
dkehn | NobodyCam: for me de3vtest.sh works | 18:00 |
dkehn | NobodyCam: hmm, the 3 key no where near the v key, how did that happen | 18:00 |
*** tatyana has quit IRC | 18:01 | |
*** datajerk has joined #openstack-ironic | 18:05 | |
NobodyCam | :) | 18:09 |
*** datajerk has quit IRC | 18:09 | |
NobodyCam | I just submitted https://review.openstack.org/59800 | 18:10 |
NobodyCam | dkehn: ^^ | 18:10 |
dkehn | NobodyCam: saw that | 18:10 |
NobodyCam | I am going to try and get things to a point that you can run the devtest script. | 18:10 |
dkehn | NobodyCam: ^^^^ should I just run devtest.sh and kill before undercloud to get where I want to go? | 18:11 |
NobodyCam | I have not done that yet. so I not sure what your outcome will be. | 18:11 |
NobodyCam | I do it manually following http://docs.openstack.org/developer/tripleo-incubator/devtest.html | 18:11 |
dkehn | NobodyCam: in othereword will it have the ironic stuf in it | 18:12 |
NobodyCam | ya. | 18:13 |
dkehn | NobodyCam: I've built tripleo many time doing the devtest.sh route, with sucess, just wondering is it a suitable dev env for ironic | 18:13 |
NobodyCam | yes it is | 18:13 |
dkehn | NobodyCam: ok, then thats the plan, thx | 18:14 |
jbjohnso | hello all | 18:15 |
jbjohnso | just fyi, I'll be vanishing from the internet after tomorrow til next year... vacation time | 18:16 |
dkehn | now thats a VAC | 18:16 |
NobodyCam | wow jbjohnso | 18:17 |
NobodyCam | thats awesome | 18:17 |
devananda | jbjohnso: enjoy! | 18:18 |
jbjohnso | Love some long vacation... spending at least some of it being something of a luddite | 18:21 |
*** rloo has quit IRC | 18:21 | |
devananda | NobodyCam: not sure if you saw -- i posted a whole bunch of patches ^^ for the hash ring work. | 18:29 |
*** matty_dubs|lunch is now known as matty_dubs | 18:30 | |
NobodyCam | I did see. have not look at them yet. try go go thru the dib walk thru so I can update it to actually work :-p | 18:30 |
devananda | NobodyCam: awesome :) | 18:30 |
*** openstackgerrit has quit IRC | 18:33 | |
*** openstackgerrit has joined #openstack-ironic | 18:33 | |
devananda | lucasagomes: i think i did not explain what replicas means in this context | 18:44 |
lucasagomes | devananda, oh | 18:45 |
lucasagomes | is not the same as the replicas on the other consistent hashing algorithms? | 18:45 |
devananda | lucasagomes: you may be thinking of buckets | 18:47 |
devananda | where # of buckets affects the smoothness of distribution of keys across the ring | 18:48 |
devananda | here taht is the hash_partition_exponent | 18:48 |
devananda | # of buckets == 2 ** hash_part_exponent | 18:48 |
devananda | replicas == # of conductors which prepare the deployment environment for each node | 18:49 |
devananda | replicas=1 --> no backup. | 18:49 |
lucasagomes | ahhh | 18:49 |
lucasagomes | righttt | 18:49 |
*** rongze has joined #openstack-ironic | 18:49 | |
devananda | replicas=2 --> one backup. any conductor goes down, all the nodes which were mapped to it can be serviced by other conductors (without downloading new images from glance) | 18:49 |
* devananda needs to write docs :) | 18:49 | |
lucasagomes | got it | 18:50 |
NobodyCam | devananda: if replicas is == 2 will one of the nodes always be primary and the other always secondary? | 18:50 |
lucasagomes | yea it's hard to think about a name for it | 18:50 |
lucasagomes | so replicas == 2 is like having a standby node | 18:51 |
devananda | yep | 18:51 |
devananda | and yep | 18:51 |
lucasagomes | if one goes down the other one will take over | 18:51 |
devananda | i've been fighting with names | 18:51 |
lucasagomes | got | 18:51 |
NobodyCam | :) | 18:51 |
lucasagomes | yea, its confusing and yet more confusing when mixed with consistent hashing because people use replicas as bucket | 18:51 |
devananda | there will be a periodic task that does something like this | 18:51 |
*** hemna has quit IRC | 18:52 | |
devananda | for node in all_associated_nodes: | 18:52 |
devananda | if self.host == ring.get_host(node): | 18:52 |
lucasagomes | maybe replication would be a better name for that? | 18:52 |
devananda | driver.deploy.prepare() | 18:52 |
devananda | gah. need etherpad. | 18:53 |
devananda | lucasagomes: that's why i used "replicas" | 18:53 |
devananda | lucasagomes: maybe if the help text says more? | 18:53 |
lucasagomes | devananda, that would help | 18:54 |
lucasagomes | I also spotted a nit right now | 18:54 |
lucasagomes | in the DocString it says "Default: 1" | 18:54 |
lucasagomes | but the default is actually the CONF option | 18:54 |
lucasagomes | so it would be 2 | 18:54 |
devananda | oh! yep | 18:54 |
*** rongze has quit IRC | 18:54 | |
lucasagomes | devananda, but yea if the help text says more that would be much better | 18:55 |
*** hemna has joined #openstack-ironic | 18:55 | |
devananda | Number of conductors which will be prepared for ... | 18:56 |
devananda | gah. i'm having a hard time explaining this in one or two sentences | 18:56 |
devananda | lucasagomes: have you seen https://etherpad.openstack.org/p/IronicConsistentHashingForInstances | 18:57 |
lucasagomes | maybe if we have a concept of active/standby | 18:57 |
lucasagomes | number of standby conductors | 18:57 |
devananda | that explains the logic behind this | 18:57 |
lucasagomes | no it's even worst | 18:57 |
devananda | yea | 18:57 |
devananda | it's not a standby conductor at all | 18:57 |
lucasagomes | yea all them are active | 18:58 |
devananda | this is also why i split out deploy.prepare() from deploy.deploy() | 18:58 |
devananda | so all ?mapped? conductors can call prepare() but only one calls deploy() | 18:58 |
devananda | i was chatting with NobodyCam yesterday about how we might coordinate such preparation | 18:59 |
devananda | right now i can do it asycn // via periodic task | 18:59 |
lucasagomes | devananda, I've seem the diagram today | 18:59 |
devananda | this handles the case of a running instance with the PXE driver, where the conductor which set it up died, and another has already prepared the PXE environment (for user-image) | 18:59 |
lucasagomes | there's one missing part there, because it's assuming that the PXE driver will control the power | 18:59 |
devananda | it just needs to update neutron | 19:00 |
devananda | ahh | 19:00 |
devananda | that diagram doesn't explain any of the hash ring things | 19:00 |
devananda | my brain melted when i got to trying to draw that | 19:00 |
*** jistr has joined #openstack-ironic | 19:00 | |
lucasagomes | haha I can imagine | 19:00 |
devananda | if i cant explain it, it means i dont fully undertand it yet | 19:00 |
devananda | :( | 19:00 |
lucasagomes | :) | 19:00 |
jbjohnso | broader question, I know I rattled off a few theoretical approaches to secure credential exchange in bare metal deploy | 19:02 |
*** datajerk has joined #openstack-ironic | 19:02 | |
*** datajerk has quit IRC | 19:02 | |
lucasagomes | devananda, maybe we could always replicate the configs across all conductors | 19:02 |
jbjohnso | did people decide what to do/not do on the front of that? | 19:02 |
devananda | jbjohnso: no. there are still seeral critical pieces that we are working on just to make Ironic function | 19:02 |
jbjohnso | ok | 19:03 |
lucasagomes | but we use the hashring to determine what conductor will be responsable for that deploy at that time | 19:03 |
devananda | lucasagomes: if you set # replicas > # conductors, you'll get that behavior | 19:03 |
lucasagomes | so every conductor could potentially take care of every node (this implies that conductors my have the same drivers) | 19:03 |
devananda | actually it doesn't imply that :) | 19:03 |
devananda | there are discrete hash rings for each driver | 19:04 |
devananda | (again, i really need to doc what i'm coding ...) | 19:04 |
lucasagomes | hehe right I will read that etherpad | 19:04 |
lucasagomes | my brain is also not fully working now it's late here | 19:04 |
devananda | lucasagomes: np. have a good night! | 19:05 |
lucasagomes | devananda, so just consider that -1 to fix that nit and (if you can) improve that help text | 19:05 |
devananda | ack | 19:05 |
NobodyCam | have a good night lucasagomes | 19:05 |
lucasagomes | devananda, thanks g'night! | 19:05 |
devananda | i also realized that the current approach to list active conductors doesn't mesh well with the intent of the hash ring | 19:05 |
lucasagomes | NobodyCam, good night sir! | 19:05 |
devananda | :-/ | 19:05 |
lucasagomes | devananda, hmm | 19:06 |
lucasagomes | what about a hangout tomorrow? | 19:06 |
lucasagomes | I mean google hangout | 19:06 |
devananda | sounds good | 19:06 |
NobodyCam | :) | 19:06 |
lucasagomes | so I can read the etherpad and then we can discuss the ideas | 19:06 |
lucasagomes | NobodyCam, u too | 19:06 |
lucasagomes | cool so we talk more about it tomorrow, g'night | 19:07 |
*** lucasagomes is now known as lucas-dinner | 19:07 | |
*** Haomeng has quit IRC | 19:08 | |
*** Haomeng has joined #openstack-ironic | 19:09 | |
devananda | right, wifi is beign flaky. I'm goign to step away for a few anyway... bbiab | 19:19 |
NobodyCam | ok :) | 19:21 |
*** rloo_ has joined #openstack-ironic | 19:23 | |
*** rloo_ has quit IRC | 19:26 | |
*** datajerk has joined #openstack-ironic | 19:26 | |
*** rloo has joined #openstack-ironic | 19:26 | |
*** Haomeng has quit IRC | 19:26 | |
*** Haomeng has joined #openstack-ironic | 19:29 | |
*** rloo has quit IRC | 19:34 | |
*** rloo has joined #openstack-ironic | 19:35 | |
*** jcooley_ has quit IRC | 19:37 | |
*** jcooley_ has joined #openstack-ironic | 19:37 | |
*** jcooley_ has quit IRC | 19:41 | |
*** jcooley_ has joined #openstack-ironic | 19:42 | |
*** datajerk has quit IRC | 19:46 | |
*** jcooley_ has quit IRC | 19:47 | |
*** jcooley_ has joined #openstack-ironic | 19:47 | |
NobodyCam | brb | 19:48 |
*** rongze has joined #openstack-ironic | 19:51 | |
*** jcooley_ has quit IRC | 19:51 | |
*** jcooley_ has joined #openstack-ironic | 19:52 | |
*** rongze has quit IRC | 19:56 | |
*** lynxman has quit IRC | 19:58 | |
*** lynxman has joined #openstack-ironic | 19:59 | |
*** datajerk has joined #openstack-ironic | 20:01 | |
*** epim has joined #openstack-ironic | 20:03 | |
devananda | back | 20:04 |
devananda | and about to patch bomb again | 20:04 |
NobodyCam | lol | 20:05 |
* NobodyCam ducks | 20:05 | |
* devananda waits for unit tests | 20:06 | |
*** jcooley_ has quit IRC | 20:06 | |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add config option for # of conductor replicas https://review.openstack.org/59794 | 20:06 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: list_active_conductor_drivers returns dict of sets https://review.openstack.org/59795 | 20:06 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add heartbeat_interval to new 'conductor' cfg group https://review.openstack.org/58851 | 20:06 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement sync_power_state periodic task https://review.openstack.org/56530 | 20:06 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add prepare and cleanup methods to deploy interface https://review.openstack.org/58894 | 20:06 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement consistent hashing of nodes <-> conductors https://review.openstack.org/58895 | 20:06 |
devananda | aaaaand again. this time, just rebasing them all | 20:08 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add missing hash_partition_exponent config option https://review.openstack.org/59793 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add config option for # of conductor replicas https://review.openstack.org/59794 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: list_active_conductor_drivers returns dict of sets https://review.openstack.org/59795 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add heartbeat_interval to new 'conductor' cfg group https://review.openstack.org/58851 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement sync_power_state periodic task https://review.openstack.org/56530 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add prepare and cleanup methods to deploy interface https://review.openstack.org/58894 | 20:09 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement consistent hashing of nodes <-> conductors https://review.openstack.org/58895 | 20:09 |
devananda | done :) | 20:09 |
NobodyCam | :) | 20:09 |
* devananda makes tea | 20:13 | |
*** vkozhukalov has quit IRC | 20:19 | |
NobodyCam | devananda: can you restore https://review.openstack.org/#/c/51328/ | 20:24 |
NobodyCam | and keep it wip? | 20:24 |
*** datajerk has quit IRC | 20:27 | |
devananda | sure | 20:33 |
NobodyCam | TY | 20:34 |
NobodyCam | :-) | 20:34 |
devananda | restored | 20:34 |
devananda | are you posting an update? | 20:34 |
NobodyCam | I'll get a cleaned up version up later. one that will pass mr J's eye | 20:34 |
devananda | :) | 20:34 |
devananda | also, it looks like https://review.openstack.org/#/c/51328/3/nova/scheduler/filters/compute_capabilities_filter.py doesn't need to be in the patch | 20:35 |
devananda | useful for your debugging as it surely is :) | 20:35 |
NobodyCam | hehehe ya I'll pull it out | 20:35 |
devananda | NobodyCam: you should probably also add yourself as a co-author in the commit message, and stuff like that | 20:36 |
NobodyCam | :) humm ok :) | 20:37 |
*** kui has joined #openstack-ironic | 20:39 | |
*** datajerk has joined #openstack-ironic | 20:40 | |
*** jistr has quit IRC | 20:48 | |
NobodyCam | brb | 20:50 |
*** jcooley_ has joined #openstack-ironic | 20:51 | |
*** jcooley_ has quit IRC | 20:53 | |
*** rongze has joined #openstack-ironic | 20:53 | |
NobodyCam | flake8 nova/virt/ironic/driver.py | 20:55 |
*** datajerk has quit IRC | 20:55 | |
NobodyCam | doh | 20:55 |
NobodyCam | ww | 20:55 |
*** rongze has quit IRC | 20:58 | |
*** datajerk has joined #openstack-ironic | 20:59 | |
*** datajerk has quit IRC | 21:03 | |
*** jcooley_ has joined #openstack-ironic | 21:24 | |
*** jdob has quit IRC | 21:41 | |
*** ndipanov has quit IRC | 21:43 | |
NobodyCam | brb making toast | 21:45 |
*** rloo has quit IRC | 21:48 | |
*** rloo has joined #openstack-ironic | 21:49 | |
devananda | mmm, toast | 21:54 |
*** rongze has joined #openstack-ironic | 21:55 | |
NobodyCam | :) cinnamon raisin toast | 21:55 |
devananda | NobodyCam: when you get to a point where you think it's ready for review, let me know. We can trade. I'd like your eyes on some of my patches too :) | 21:55 |
NobodyCam | :) will start look at them :) (while I eat the toast) lol | 21:56 |
devananda | NobodyCam: first patch in the series is https://review.openstack.org/#/c/59793/ | 21:57 |
devananda | it's a fairly long series .... but hopefully each patch is small | 21:57 |
devananda | *enough to be quick | 21:57 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add config option for # of conductor replicas https://review.openstack.org/59794 | 22:00 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: list_active_conductor_drivers returns dict of sets https://review.openstack.org/59795 | 22:00 |
openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement consistent hashing of nodes <-> conductors https://review.openstack.org/58895 | 22:00 |
NobodyCam | devananda: are landing these or just +2 for now?.. ie. until others also have eyes on them | 22:00 |
*** rongze has quit IRC | 22:00 | |
NobodyCam | ie 59793 | 22:00 |
NobodyCam | is good to land | 22:01 |
devananda | at your discretion. some are probably good to land | 22:03 |
devananda | i marked https://review.openstack.org/#/c/58895/ as WIP just now -- it is not finished. | 22:03 |
devananda | the others are probably OK to land | 22:03 |
*** jbjohnso has quit IRC | 22:04 | |
*** linggao has quit IRC | 22:04 | |
devananda | https://review.openstack.org/#/c/56682/ is also blocking several changes to the API. I've already +2'd | 22:06 |
NobodyCam | 59794 looks like it should also depend on 59793 | 22:09 |
NobodyCam | but we'll land 93 now | 22:10 |
NobodyCam | 93 +a'd | 22:11 |
NobodyCam | brb.. | 22:13 |
*** datajerk has joined #openstack-ironic | 22:18 | |
openstackgerrit | A change was merged to openstack/ironic: Add missing hash_partition_exponent config option https://review.openstack.org/59793 | 22:20 |
*** datajerk has quit IRC | 22:24 | |
*** datajerk has joined #openstack-ironic | 22:28 | |
rloo | devananda, yt? have a few minutes? | 22:37 |
NobodyCam | devananda: you still here? | 22:38 |
devananda | rloo: hi! | 22:38 |
devananda | NobodyCam: yes | 22:38 |
rloo | NobodyCam - get in line ;) | 22:38 |
NobodyCam | just a quick question | 22:38 |
NobodyCam | lol | 22:38 |
rloo | ok, NobodyCam, go ahead :-) | 22:38 |
NobodyCam | lol TY rloo :) | 22:38 |
devananda | hehe | 22:39 |
NobodyCam | devananda: on @periodic_task.periodic_task(spacing=CONF.conductor.heartbeat_interval) why spacing vs something like interval | 22:39 |
*** epim has quit IRC | 22:39 | |
devananda | NobodyCam: that parameter name is part of the definition of periodic_task decorator,w hich is coming from Oslo | 22:40 |
devananda | NobodyCam: I didn't choose to name it "spacing" :) | 22:40 |
NobodyCam | :) | 22:40 |
NobodyCam | ack | 22:40 |
NobodyCam | see rloo was quick | 22:40 |
NobodyCam | lol | 22:40 |
NobodyCam | :-p | 22:40 |
rloo | NobodyCam. ha ha, yup, you're quick! | 22:40 |
rloo | so devananda, wrt IcehouseIronicNextSteps -- "API to break a task_manager lock". | 22:41 |
devananda | ya? | 22:41 |
rloo | devananda, does that have anything to do with this blueprint: https://blueprints.launchpad.net/ironic/+spec/breaking-resource-locks | 22:41 |
devananda | yes | 22:42 |
devananda | one and the same | 22:42 |
rloo | ok good. that's what I thought. | 22:42 |
rloo | do you think things/code/etc are in a state where it might be worth starting on this, or to wait a bit? | 22:43 |
devananda | rloo: well.... | 22:46 |
rloo | devananda: ok, I'll wait a bit then :-) | 22:47 |
devananda | rloo: jsuta second | 22:47 |
devananda | rloo: see https://review.openstack.org/#/c/55549/ | 22:49 |
devananda | it looks like yuriyz already started | 22:49 |
devananda | rloo: so please feel free to work on it | 22:49 |
devananda | brb | 22:49 |
rloo | devananda: if yuriyz already started, shouldn't he just finish? | 22:49 |
devananda | rloo: his patch was to fix this bug: https://bugs.launchpad.net/ironic/+bug/1250348 | 22:51 |
devananda | rloo: which overlaps with the BP | 22:51 |
*** epim has joined #openstack-ironic | 22:52 | |
rloo | devananda: ok, i'll ping yuriyz tomorrow to make sure he wasn't going to do more. | 22:53 |
devananda | rloo: great, thanks :) | 22:53 |
rloo | just call me tom sawyer... ;) | 22:53 |
*** michchap has quit IRC | 22:56 | |
*** rongze has joined #openstack-ironic | 22:57 | |
NobodyCam | devananda: on line 392 of manger.py in patch 56530 you use log.warning any thought on using audit like https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L411-L412 | 23:00 |
NobodyCam | thats info I would like to always see? but I'm strange | 23:01 |
*** michchap has joined #openstack-ironic | 23:02 | |
*** rongze has quit IRC | 23:02 | |
NobodyCam | quick walkies.. brb | 23:15 |
*** jcooley_ has quit IRC | 23:18 | |
NobodyCam | back | 23:21 |
*** matty_dubs is now known as matty_dubs|gone | 23:30 | |
*** yfujioka has joined #openstack-ironic | 23:43 | |
NobodyCam | deva you here? | 23:43 |
NobodyCam | :-p | 23:43 |
*** hemna has quit IRC | 23:50 | |
*** datajerk has quit IRC | 23:53 | |
*** hemna has joined #openstack-ironic | 23:54 | |
openstackgerrit | A change was merged to openstack/ironic: Add heartbeat_interval to new 'conductor' cfg group https://review.openstack.org/58851 | 23:55 |
openstackgerrit | A change was merged to openstack/ironic: Add missing link for drivers resource https://review.openstack.org/59081 | 23:55 |
*** openstackgerrit has quit IRC | 23:56 | |
*** openstackgerrit has joined #openstack-ironic | 23:56 | |
*** kui has quit IRC | 23:56 | |
*** datajerk has joined #openstack-ironic | 23:56 | |
*** rongze has joined #openstack-ironic | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!