rloo | JayF: that's why I'm asking. Don't know if you guys didn't want it, or if it was an omission. | 00:00 |
---|---|---|
jroll | JayF: it's a separate interface | 00:00 |
jroll | it was an omission, rloo | 00:00 |
jroll | not intentional | 00:00 |
jroll | just... not something we noticed | 00:00 |
rloo | jroll: ok, then. i can fix it. i have the powah. | 00:00 |
jroll | \o/ | 00:00 |
jroll | hooray for open source. | 00:00 |
rloo | :D | 00:01 |
openstackgerrit | Ruby Loo proposed a change to openstack/ironic: Add console interface to agent_ipmitool driver https://review.openstack.org/115826 | 00:11 |
jroll | rloo: no tests? :P | 00:14 |
*** harlowja has quit IRC | 00:15 | |
*** harlowja_ has joined #openstack-ironic | 00:15 | |
openstackgerrit | Gregory Haynes proposed a change to openstack/ironic: More helpful failure for tests on noexec /tmp https://review.openstack.org/115827 | 00:17 |
*** christop1eraedo is now known as christopheraedo | 00:18 | |
rloo | jroll: nope, cuz there weren't any tests for the other non-agent stuff :-) | 00:22 |
jroll | huh, ok | 00:24 |
rloo | jroll: do you want me to ask the folks that wrote the agent stuff, to add tests? | 00:27 |
jroll | yeah | 00:28 |
jroll | those guys suck | 00:28 |
rloo | jroll: ok, will do. Nah, they're nice guys. It is the reviewers that are the problelm. | 00:28 |
rloo | hey jroll: would you please add unit tests for the agent, that test all the interfaces? It only tests the agent-related interfaces now. Some jroll guy wants them. | 00:29 |
jroll | lol | 00:30 |
jroll | file a bug :P | 00:30 |
jroll | (mostly kidding) | 00:30 |
rloo | jroll: huh, 'mostly', not 'totally'? ;) | 00:31 |
jroll | :P | 00:31 |
greghaynes | aye, I need to learn how to db migration | 00:35 |
*** penick has quit IRC | 00:37 | |
*** eghobo has quit IRC | 00:39 | |
*** penick has joined #openstack-ironic | 00:39 | |
*** eghobo has joined #openstack-ironic | 00:39 | |
openstackgerrit | Gregory Haynes proposed a change to openstack/ironic: WIP: Add conductor_hostname property to nodes https://review.openstack.org/109688 | 00:41 |
*** penick has quit IRC | 00:51 | |
*** penick has joined #openstack-ironic | 00:54 | |
*** jcoufal has quit IRC | 00:55 | |
*** dlaube has quit IRC | 01:00 | |
*** penick has quit IRC | 01:01 | |
*** shakamunyi has joined #openstack-ironic | 01:17 | |
*** nosnos has joined #openstack-ironic | 01:44 | |
*** rloo has quit IRC | 01:51 | |
*** lazy_prince has quit IRC | 01:53 | |
*** killer_prince has joined #openstack-ironic | 01:55 | |
*** killer_prince is now known as lazy_prince | 01:55 | |
greghaynes | Any reason why my migration isnt being run in CI on https://review.openstack.org/#/c/109688/5 ? | 01:59 |
greghaynes | oh, it is running I just fail at writing them ;) | 02:01 |
openstackgerrit | Gregory Haynes proposed a change to openstack/ironic: WIP: Add conductor_hostname property to nodes https://review.openstack.org/109688 | 02:02 |
*** christopheraedo has quit IRC | 02:03 | |
*** christopheraedo has joined #openstack-ironic | 02:04 | |
*** christopheraedo has left #openstack-ironic | 02:05 | |
openstackgerrit | Ellen Hui proposed a change to openstack/ironic: Make DHCP provider pluggable https://review.openstack.org/112351 | 02:08 |
*** lazy_prince has quit IRC | 02:15 | |
*** killer_prince has joined #openstack-ironic | 02:16 | |
*** killer_prince is now known as lazy_prince | 02:16 | |
*** rushiagr_away is now known as rushiagr | 02:23 | |
*** shakayumi has joined #openstack-ironic | 02:25 | |
*** shakamunyi has quit IRC | 02:28 | |
*** alexiz has joined #openstack-ironic | 02:31 | |
*** shakayumi has quit IRC | 02:34 | |
*** harlowja_ is now known as harlowja_away | 02:42 | |
*** eghobo has quit IRC | 02:53 | |
*** yuanying has quit IRC | 03:00 | |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Add send-data-to-ceilometer support for pxe_ipminative driver https://review.openstack.org/112486 | 03:04 |
*** Poornima has joined #openstack-ironic | 03:08 | |
*** vinbs has joined #openstack-ironic | 03:09 | |
*** ramineni has joined #openstack-ironic | 03:15 | |
*** yuanying has joined #openstack-ironic | 03:16 | |
*** yuanying has quit IRC | 03:21 | |
*** shakamunyi has joined #openstack-ironic | 03:54 | |
*** alexiz has quit IRC | 03:57 | |
*** yuanying has joined #openstack-ironic | 04:34 | |
*** rameshg87 has joined #openstack-ironic | 04:38 | |
*** rushiagr is now known as rushiagr_away | 04:58 | |
*** rushiagr_away is now known as rushiagr | 05:01 | |
*** rushiagr is now known as rushiagr_away | 05:01 | |
*** Poornima has quit IRC | 05:09 | |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/115600 | 05:16 |
*** pcrews has quit IRC | 05:16 | |
openstackgerrit | Vinay B S proposed a change to openstack/ironic: Adds network creation instructions https://review.openstack.org/114886 | 05:17 |
Haomeng | vinbs: morning:) | 05:21 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/115600 | 05:25 |
*** shakamunyi has quit IRC | 05:37 | |
*** Poornima has joined #openstack-ironic | 05:43 | |
*** killer_prince has joined #openstack-ironic | 05:45 | |
*** pradipta_away is now known as pradipta | 05:45 | |
vinbs | Haomeng: morning! :) | 05:45 |
Haomeng | vinbs: :) | 05:45 |
*** killer_prince has quit IRC | 05:46 | |
*** teju has joined #openstack-ironic | 05:50 | |
*** bmahalakshmi has joined #openstack-ironic | 05:53 | |
*** dtantsur|afk is now known as dtantsur | 06:03 | |
dtantsur | Morning Ironic | 06:04 |
mrda | hi dtantsur | 06:05 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/115878 | 06:08 |
dtantsur | mrda, g'evening | 06:09 |
*** rushiagr_away is now known as rushiagr | 06:20 | |
*** k4n0 has joined #openstack-ironic | 06:25 | |
openstackgerrit | Vinay B S proposed a change to openstack/ironic: Adds network creation instructions https://review.openstack.org/114886 | 06:29 |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Implement hardware discovery setup in PXE driver https://review.openstack.org/110031 | 06:31 |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: Add newly_discovered column to Node object https://review.openstack.org/107389 | 06:31 |
*** tatyana has joined #openstack-ironic | 06:40 | |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: Add support for creating vfat and iso images https://review.openstack.org/105413 | 06:46 |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: Add support for interacting with swift https://review.openstack.org/105795 | 06:50 |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: IloVirtualMediaIscsi deploy driver https://review.openstack.org/113865 | 06:54 |
*** jcoufal has joined #openstack-ironic | 07:02 | |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: IloVirtualMediaIscsi deploy driver https://review.openstack.org/113865 | 07:03 |
*** shakamunyi has joined #openstack-ironic | 07:04 | |
*** yongli has joined #openstack-ironic | 07:06 | |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: IloVirtualMediaIscsi deploy driver https://review.openstack.org/113865 | 07:08 |
*** shakamunyi has quit IRC | 07:08 | |
*** athomas has joined #openstack-ironic | 07:10 | |
*** ndipanov has joined #openstack-ironic | 07:13 | |
openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: IloVirtualMediaAgent deploy driver https://review.openstack.org/115885 | 07:30 |
yuriyz | morning Ironic | 07:32 |
dtantsur | yuriyz, morning | 07:58 |
*** derekh has joined #openstack-ironic | 08:16 | |
*** lucasagomes has joined #openstack-ironic | 08:16 | |
*** viktors|afk is now known as viktors | 08:28 | |
*** igordcard has joined #openstack-ironic | 08:48 | |
*** pelix has joined #openstack-ironic | 08:51 | |
*** Haomeng has quit IRC | 08:55 | |
openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Add driver name on driver load exception https://review.openstack.org/112049 | 09:04 |
*** igordcard has quit IRC | 09:20 | |
*** Haomeng has joined #openstack-ironic | 09:22 | |
*** igordcard has joined #openstack-ironic | 09:25 | |
*** yuanying has quit IRC | 09:31 | |
*** yuanying has joined #openstack-ironic | 09:35 | |
*** pradipta is now known as pradipta_away | 09:41 | |
*** yuanying has quit IRC | 10:10 | |
*** aswadr has quit IRC | 10:12 | |
*** dtantsur is now known as dtantsur|lunch | 10:26 | |
*** rameshg87 has quit IRC | 10:28 | |
*** yuanying has joined #openstack-ironic | 10:41 | |
*** lazy_prince is now known as killer_prince | 10:48 | |
*** killer_prince is now known as lazy_prince | 10:53 | |
*** teju has quit IRC | 11:00 | |
*** yuanying has quit IRC | 11:15 | |
*** mikedillion has joined #openstack-ironic | 11:18 | |
*** Poornima has quit IRC | 11:29 | |
*** mikedillion has quit IRC | 11:35 | |
*** mkerrin1 has quit IRC | 11:38 | |
marios | lucasagomes: 'ironic node-set-power-state e99fe1b3-4cca-49e5-9774-7cb164aeece9 on' (or should that be 'power_on' or something else?) | 11:46 |
marios | lucasagomes: nm sorry got the help message | 11:47 |
marios | lucasagomes: apologies for the noise | 11:47 |
lucasagomes | marios, heh no worries | 11:47 |
lucasagomes | it's 'on' 'off' 'reboot' afair | 11:47 |
marios | lucasagomes: \o/ it asks for dhcp service and dnsmasq replies from the 'discovery range' (i.e. the distinct range from the 'normal' undercloud one) | 11:54 |
lucasagomes | marios, w00t! | 11:54 |
marios | we can play with the tags and see if we can serve up different pxe options as well (should just be getting config right) | 11:55 |
lucasagomes | marios, a nice one, yeah... for iPXE I have a test where I check if the pxe request comes from iPXE or standard PXE and if it's standard I send the iPXE boot image so machine boot can chainload it | 11:58 |
lucasagomes | the machine booting* | 11:58 |
marios | lucasagomes: cool, lets talk later | 11:58 |
lucasagomes | marios, aight | 11:58 |
*** vinbs has quit IRC | 12:02 | |
*** yuanying has joined #openstack-ironic | 12:11 | |
*** nosnos has quit IRC | 12:13 | |
*** octavian1001 has joined #openstack-ironic | 12:13 | |
*** ramineni1 has joined #openstack-ironic | 12:14 | |
*** ramineni has quit IRC | 12:16 | |
dtantsur|lunch | marios, \o/ | 12:30 |
*** dtantsur|lunch is now known as dtantsur | 12:30 | |
marios | :) | 12:31 |
dtantsur | now to something completely different: after we increased default image cache values, I'm seeing complains from people with hard drive overflow | 12:31 |
dtantsur | I wonder if we should revisit caching logic or revert to lower values | 12:32 |
dtantsur | lifeless, lucasagomes ^^^ | 12:32 |
lucasagomes | dtantsur, they were running it in production or just testing? | 12:35 |
dtantsur | lucasagomes, testing. That's why they have little disc space. | 12:35 |
dtantsur | And seems like that *2 logic to account for converting does not always work either :( | 12:36 |
lucasagomes | dtantsur, right so they have to tune down the config because in OpenStack the defaults should be production defaults | 12:36 |
lucasagomes | afaiui | 12:36 |
dtantsur | well, yes... I wonder if we should set them down in devstack | 12:37 |
lucasagomes | right... so it's buggy yeah we gotta try to fix it | 12:37 |
dtantsur | lucasagomes, also see https://bugs.launchpad.net/ironic/+bug/1356362 | 12:37 |
lucasagomes | dtantsur, yeah, makes sense in devstack | 12:37 |
lucasagomes | because it's mostly used for tests | 12:37 |
*** bmahalakshmi has quit IRC | 12:43 | |
*** yuanying has quit IRC | 12:44 | |
*** octavian1001 has left #openstack-ironic | 12:45 | |
dtantsur | hmmm... who are we waiting for approval of https://review.openstack.org/#/c/115493/ ? I doubt we should wait for devananda :) | 12:45 |
agordeev | good morning/evening folks! is anyone here familiar with `dracut`? | 12:57 |
openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Control extra space for images conversion in image_cache https://review.openstack.org/115974 | 13:00 |
yuriyz | dtantsur, https://review.openstack.org/#/c/115974/ | 13:08 |
*** rloo has joined #openstack-ironic | 13:08 | |
dtantsur | yuriyz, already looking, thanks a lot! | 13:10 |
*** jasondotstar has joined #openstack-ironic | 13:19 | |
*** pcrews has joined #openstack-ironic | 13:20 | |
*** openstackgerrit has quit IRC | 13:21 | |
lucasagomes | agordeev, hey... hmm not super familiar but whats up? | 13:32 |
*** linggao has joined #openstack-ironic | 13:34 | |
rloo | morning ironickers | 13:37 |
lucasagomes | rloo, morning | 13:37 |
rloo | I see quite a few jenkins failures. Does anyone know why? | 13:37 |
*** lazy_prince is now known as killer_prince | 13:38 | |
*** krtaylor has quit IRC | 13:41 | |
*** yuanying has joined #openstack-ironic | 13:41 | |
linggao | morning rloo lucasagomes | 13:45 |
lucasagomes | linggao, good morning | 13:46 |
linggao | rloo, I saw your review comments on 97331. Do you really need a bug for it? | 13:47 |
linggao | (console for ipminative) | 13:47 |
rloo | morning linggao. | 13:47 |
agordeev | lucasagomes: it may not too related with ironic at all, but accoring to the last news it becomes capable of loading root live-images via torrents. I'm just wondering how that can possibly help us in a case of provisioning. | 13:47 |
rloo | linggao: that is adding a new feature. it is at least a bug. | 13:48 |
lucasagomes | agordeev, nice... actually dracut is pretty cool we could do many things using it | 13:48 |
lucasagomes | even generating our ramdisk instead of using DIB | 13:48 |
lucasagomes | you can use only dracut | 13:48 |
rloo | linggao: it isn't difficult to add a bug. be glad no one has asked you to write a spec for it ;) | 13:49 |
lucasagomes | def something to keep an eye at | 13:49 |
dtantsur | rloo, linggao, agordeev, morning! | 13:49 |
linggao | rloo, I do not mind create a bug for it as long as you guys/girls can approve. I am just afraid that one bug is open, someone will say it needs a spec. This patch has been there for a long time. I just want to finish it up. | 13:49 |
rloo | hi dtantsur! | 13:49 |
agordeev | dtantsur: morning :) | 13:49 |
linggao | good moring dtantsur agordeev | 13:50 |
agordeev | linggao: morning too! | 13:50 |
rloo | linggao: I cannot speak for others. Someone may ask for a spec. that would be regardless if you opened a bug or not though. | 13:50 |
agordeev | rloo: morning also :) | 13:50 |
rloo | linggao: I don't think it needs a spec though ;) | 13:50 |
jroll | morning ironic :) | 13:50 |
rloo | hi agordeev (not sure what time zone you're in). | 13:50 |
jroll | agordeev! long time no see | 13:50 |
rloo | morning jroll | 13:50 |
jroll | morning rloo lucasagomes linggao dtantsur | 13:51 |
linggao | rloo, thanks. I'll open a bug then. | 13:51 |
lucasagomes | jroll, morning | 13:51 |
agordeev | jroll: morning, yup! i'm still being too busy with our product. Not sure what to do with my work for IPA | 13:51 |
linggao | good morning jroll. | 13:51 |
jroll | agordeev: all inventory sorts of things have been bumped to kilo, so it will sit some more :P | 13:52 |
jroll | dtantsur: we could probably approve https://review.openstack.org/#/c/115493/ | 13:52 |
jroll | dtantsur: the only person I would want to check with is mrda, but I'm going to assume he wants it landed | 13:53 |
*** datajerk has quit IRC | 13:53 | |
* jroll jfdi | 13:53 | |
* NobodyCam wakes up slowly | 13:54 | |
NobodyCam | good morning Ironic | 13:54 |
jroll | morning NobodyCam :) | 13:55 |
*** k4n0 has quit IRC | 13:55 | |
rloo | morning NobodyCam | 13:56 |
linggao | good morning NobodyCam | 13:56 |
NobodyCam | morning jroll rloo and linggao :) | 13:57 |
*** openstackgerrit has joined #openstack-ironic | 13:59 | |
*** datajerk has joined #openstack-ironic | 14:00 | |
dtantsur | NobodyCam, jroll, morning | 14:01 |
NobodyCam | good morning dtantsur :) | 14:02 |
rloo | anyone seen gate_tempest_dsvm-virtual-ironic failing in tearDownClass? Looks like it is trying to delete the chassis but it has nodes. | 14:07 |
rloo | I guess our tempest experts are on vacation this week? | 14:07 |
jroll | rloo: got a link? | 14:07 |
rloo | jroll: http://logs.openstack.org/92/115692/1/gate/gate-tempest-dsvm-virtual-ironic/0dfc429/logs/testr_results.html.gz | 14:08 |
rloo | jroll: under tearDownClass | 14:08 |
rloo | jroll: this is the first time I'm looking at this tempest page. | 14:08 |
jroll | that's a lot of failures | 14:10 |
jroll | might be unrelated | 14:10 |
rloo | jroll: I just looked at another patch where that gate fails, but it is failing in tempest in different place. | 14:10 |
jroll | yeah | 14:10 |
jroll | tempest fails everywhere :) | 14:10 |
jroll | http://logs.openstack.org/92/115692/1/gate/gate-tempest-dsvm-virtual-ironic/0dfc429/console.html#_2014-08-21_13_25_44_239 | 14:10 |
jroll | failed 78 tests | 14:11 |
rloo | jroll: do I ignore it? I see quite a few folks doing 'recheck no bug'. | 14:11 |
jroll | which means it's likely something didn't get set up right | 14:11 |
jroll | well | 14:11 |
jroll | we should try to track down the actual issue | 14:11 |
jroll | and see if there's a bug, if so do 'recheck bug xxx' | 14:11 |
rloo | jroll: yeah, i know. we should try to track it down :-( | 14:11 |
jroll | rloo: I'll take a look :) | 14:11 |
rloo | jroll: well, this is good experience. ha ha. | 14:11 |
*** chuckC_ has quit IRC | 14:11 | |
rloo | jroll: I also saw another gate-docs fail, haven't yet looked into that one. wanted to see how frequent it was failing first. | 14:12 |
jroll | heh | 14:12 |
jroll | so with this first one | 14:12 |
jroll | I usually like to go to the logs, console first (which is what I linked to you above) | 14:12 |
jroll | since there's so many failures, something was probably set up wrong | 14:12 |
jroll | so starting from here: http://logs.openstack.org/92/115692/1/gate/gate-tempest-dsvm-virtual-ironic/0dfc429/ | 14:13 |
jroll | I go to logs, then to devstack-gate-setup-host.txt.gz or devstack-gate-setup-workspace-new.txt.gz | 14:13 |
jroll | first one looks a bit short to me :| | 14:14 |
*** yuanying has quit IRC | 14:14 | |
rloo | jroll: so do you get the logs from a good session and compare? | 14:15 |
jroll | rloo: I have in the past, turns out that first log is actually normal | 14:16 |
jroll | (checked against a good one) | 14:16 |
rloo | jroll: ok. I just got a good one to compare ;) | 14:17 |
jroll | k :) | 14:17 |
*** krtaylor has joined #openstack-ironic | 14:17 | |
* rloo thinks a tool to smartly-diff the logs between two gate runs might be nice. | 14:18 | |
jroll | hmm, could be | 14:19 |
jroll | so looking at the tempest logs | 14:19 |
jroll | there's a ton of | 14:19 |
jroll | 2014-08-21 13:16:06.648 22733 TRACE tempest.test IdentityError: Got identity error | 14:19 |
jroll | 2014-08-21 13:16:06.648 22733 TRACE tempest.test Details: Unexpected status code 400 | 14:19 |
*** ChuckC has quit IRC | 14:20 | |
rloo | jroll: I was thinking those failures were due to the first failed cleanup. | 14:23 |
jroll | rloo: I think it's keystone being dumb | 14:23 |
*** pradipta_away is now known as pradipta | 14:26 | |
jroll | rloo: I'm having trouble actually finding anything, though, I might just 'recheck no bug' for now | 14:28 |
NobodyCam | I have a question, what would be the best way of getting our client into https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/contrib/baremetal_nodes.py | 14:28 |
jroll | NobodyCam: go to sleep and dream about it? | 14:29 |
rloo | jroll: ok thx. I might look a bit more. or not ;) | 14:29 |
jroll | :P | 14:29 |
jroll | NobodyCam: you mean, python-ironicclient? | 14:29 |
NobodyCam | lol. yeppers | 14:29 |
rloo | NobodyCam: why? | 14:30 |
jroll | why | 14:30 |
*** pradipta is now known as pradipta_away | 14:30 | |
jroll | I don't think that's valuable, then we have two places to update | 14:30 |
*** rushiagr is now known as rushiagr_away | 14:30 | |
NobodyCam | so we can proxy nova baremetal-node-list, and nova baremetal-show commands | 14:30 |
NobodyCam | I'm trying not to import https://github.com/openstack/ironic/blob/master/ironic/nova/virt/ironic/driver.py#L51-L77 | 14:31 |
jroll | just... import the client? | 14:31 |
jroll | mmmmmmmm | 14:31 |
jroll | oh | 14:32 |
jroll | so you're saying best way to import the client | 14:32 |
NobodyCam | I'm if I import the conf values I end up with a ValueError: Circular reference detected | 14:32 |
NobodyCam | yes! | 14:32 |
jroll | grr circular imports | 14:32 |
jroll | NobodyCam: can I see the code? | 14:34 |
rloo | NobodyCam: do we need this for graduation? | 14:34 |
jroll | also | 14:34 |
jroll | do we import baremetal_nodes in the virt driver? | 14:34 |
jroll | rloo: yeah :| | 14:34 |
lucasagomes | someone knows Stig Telfer IRC name? | 14:36 |
rloo | what exactly does 'proxy' mean? That someone has openstack with baremetal, or openstack with ironic, and issuing nova baremetal* commands? | 14:36 |
jroll | rloo: something like that, it's for upgrade testing iirc | 14:36 |
*** jgrimm has joined #openstack-ironic | 14:37 | |
NobodyCam | yes, we only need to handle simple command node-list and node-show | 14:37 |
NobodyCam | add node/port can error | 14:37 |
rloo | jroll: would be good to nail down the requirements. Is it just for our testing, or is it something our users will use? | 14:37 |
jroll | idk | 14:38 |
jroll | ask NobodyCam :P | 14:38 |
rloo | jroll: sorry, i thought maybe you all knew ;) | 14:38 |
jroll | I don't wanna know :P | 14:38 |
NobodyCam | nova would like the node-list and node-show command to be proxied, As I understand it, its for testing and users | 14:39 |
jroll | oh, goody | 14:39 |
rloo | NobodyCam: so what's the scenario. I haven't actually thought about it. Someone has openstack w baremetal w ironic? Then 1. they use the db-migrate-scripts to get bm node info into ironic db. | 14:40 |
NobodyCam | jroll: just some initial hacking attempts: http://paste.openstack.org/show/X7F8rz7SsZs4DII4wS5F/ | 14:41 |
rloo | NobodyCam: then 2. they issue nova bm* commands and we proxy to ironic API? | 14:41 |
NobodyCam | yep, We just need to handle the baremetal-node-list & baremetal-node-show commands | 14:42 |
rloo | NobodyCam: so its possible to have openstack+bm+ironic at the same time? | 14:43 |
NobodyCam | no | 14:44 |
NobodyCam | this will be for once NBM is removed | 14:44 |
lucasagomes | NobodyCam, u def have to convert the values coming from ironic to a nova bm node style | 14:44 |
lucasagomes | like that commented dict you have there before the return | 14:44 |
NobodyCam | oh ya! | 14:45 |
NobodyCam | lucasagomes: my first hack http://paste.openstack.org/show/hNLkJ4Yjb33Z11H9LHJn/ | 14:45 |
NobodyCam | lol | 14:45 |
lucasagomes | NobodyCam, ah awesome, almost there then ;) | 14:45 |
lucasagomes | (surprised that there's no tests checking the type of the fields there...) | 14:46 |
NobodyCam | lol shhhh | 14:46 |
rloo | NobodyCam: ha ha. | 14:46 |
jroll | NobodyCam: have the traceback from the circular import? | 14:46 |
NobodyCam | sure one sec | 14:46 |
NobodyCam | jroll: http://paste.openstack.org/show/bMfaM5MMKswQLoZKIHpV/ | 14:47 |
jroll | OH | 14:48 |
jroll | that's not a circular import | 14:48 |
jroll | that's like, bad json or something | 14:48 |
NobodyCam | I haven't dug into it a whole lot. I hit that at like 5 lastnight | 14:49 |
lucasagomes | +1 | 14:49 |
lucasagomes | NobodyCam, when you do a node = icli.node.list() and then nodes.append(node) | 14:51 |
lucasagomes | I bet ur appending a list of objects in a list | 14:52 |
lucasagomes | and that may not be serializable | 14:52 |
lucasagomes | hence the error ur getting | 14:52 |
lucasagomes | AFAIUI, you have to get the node.list() and then convert each of the items to the nova bm format | 14:52 |
lucasagomes | and not append with another list, just return the list of converted items | 14:52 |
NobodyCam | oh let me test that. I was thinking the errors where comming from the CONF.import_opt stuff | 14:53 |
jroll | lucasagomes: yeah, sounds right | 14:54 |
NobodyCam | but first a conf call :-p | 14:55 |
lucasagomes | :) | 14:55 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/115600 | 14:55 |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/114058 | 14:55 |
*** coolsvap has quit IRC | 14:58 | |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Endpoint for accepting discovered data https://review.openstack.org/116003 | 14:59 |
jroll | dtantsur really wants this discovery stuff | 14:59 |
*** ChuckC has joined #openstack-ironic | 14:59 | |
dtantsur | jroll, not really _me_, but well yeah :) | 14:59 |
jroll | heh | 15:00 |
*** chuckC_ has joined #openstack-ironic | 15:00 | |
dtantsur | it proved to be faaaaar more complex, than I expected ~ month ago :( | 15:01 |
jroll | right, we punted it for a reason :P | 15:01 |
*** coolsvap has joined #openstack-ironic | 15:04 | |
*** mikedillion has joined #openstack-ironic | 15:05 | |
*** yuanying has joined #openstack-ironic | 15:11 | |
*** rwsu has quit IRC | 15:23 | |
openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Control extra space for images conversion in image_cache https://review.openstack.org/115974 | 15:24 |
*** dtantsur is now known as dtantsur|afk | 15:31 | |
*** rwsu has joined #openstack-ironic | 15:39 | |
*** yuanying has quit IRC | 15:44 | |
*** mikedillion has quit IRC | 15:58 | |
*** mikedillion has joined #openstack-ironic | 16:00 | |
*** krtaylor has quit IRC | 16:00 | |
NobodyCam | gah locked up my mac book :-p | 16:01 |
lifeless | dtantsur|afk: tuning down in devstack would be fine IMO | 16:02 |
lifeless | dtantsur|afk: the 2x thing we knew was wrong, not surprising ;) | 16:02 |
JayF | A question for the folks here; I know when we started IPA the idea was that we'd do releases periodically similar to how the clients do | 16:15 |
JayF | but I'm starting to wonder if we should cut a release for Juno, so we have an agent that's 'blessed' for use in installs | 16:16 |
JayF | I personally would not use this, as we obviously run master/master+unmerged patches, but I'm curious what other users of the agent think | 16:16 |
jroll | I'd like to hear from lifeless and lucasagomes on this ^^ | 16:16 |
jroll | and NobodyCam | 16:16 |
NobodyCam | humm | 16:21 |
NobodyCam | ipa has landed so I could see a juno release | 16:21 |
JayF | Like I continue to think that master of ironic should work with any agent master (well, from merge onwards, at least) due to versioning and such | 16:22 |
JayF | plus most big or multitenant installs will want to build their own agent with their own hardware manager | 16:22 |
NobodyCam | JayF: with a release, wouldn't that be more like the refference version? | 16:24 |
JayF | I don't know, honestly | 16:25 |
JayF | that's why I'm asking :) | 16:25 |
JayF | I still don't understand why someone would want to run a release from an arbitrary date because it has a name instead of running the latest and greatest ;) | 16:26 |
lucasagomes | jroll, NobodyCam JayF does TripleO goes through the release process... if so having a release of IPA for Juno is the same of having a release for Juno of DIB (and it's elements, which is == our default ramdisk) | 16:26 |
lucasagomes | so I would say yes, it makes sense to have a release for IPA | 16:27 |
JayF | the second question that follows is what form is the release? | 16:27 |
JayF | reference images? pypi (I don't love this but understand why it could be useful)? | 16:27 |
lucasagomes | yeah... well that's the thing we would probably have to follow the whole release process | 16:28 |
lucasagomes | https://wiki.openstack.org/wiki/Release_Cycle | 16:28 |
lucasagomes | which we haven't done (I think)... we didn't have any milestone release for IPA etc... | 16:28 |
jroll | well | 16:28 |
jroll | the goal is to release like the client | 16:29 |
jroll | the reason I see a juno release as valuable: | 16:29 |
jroll | 1) stable/juno CI | 16:29 |
jroll | 2) to say "this agent works with juno" | 16:29 |
rloo | jroll: question about ironic.drivers.modules.agent wrt driver_info properties | 16:29 |
jroll | pypi makes sense e.g. to pip install in a ramdisk / docker container | 16:30 |
jroll | deb/rpm makes absolutely no sense | 16:30 |
JayF | like if a juno bug is reported, step 1 is to test with the 'reference' version for juno, to see if it's a new regression in the agent or something in ironic | 16:30 |
jroll | rloo: shoot | 16:30 |
JayF | right? | 16:30 |
*** dlaube has joined #openstack-ironic | 16:30 | |
rloo | jroll: the agent code returns {} for get_properties(). but... | 16:30 |
*** viktors is now known as viktors|afk | 16:30 | |
jroll | rloo: yeah, so deva and I chatted about this, since the deployer doesn't enter anything in driver_info it makes sense for get_properties to return {} | 16:31 |
rloo | jroll: I see that 'agent_url' should be in driver_info? | 16:31 |
rloo | jroll: who puts in agent_url? | 16:31 |
JayF | the agent does. | 16:31 |
JayF | on it's lookup/heartbeat | 16:31 |
jroll | rloo: computers :) | 16:31 |
rloo | jroll: ahhh. ok, got it. | 16:32 |
kylestev | JayF: what do you think about the agent being available in pypi? | 16:32 |
JayF | I think I personally wouldn't use it, but I can see the value for folks wanting to build their own agents to have a 'stable' base | 16:34 |
JayF | but honestly IDK if I'd want to encourage "build a hardware manager on this arbitrary SHA1" as a pattern vs having those users follow master or pick a spot to integrate themselves | 16:34 |
jroll | brb | 16:34 |
*** ramineni1 has quit IRC | 16:39 | |
*** yuanying has joined #openstack-ironic | 16:41 | |
*** vdrok_ has joined #openstack-ironic | 16:43 | |
*** krtaylor has joined #openstack-ironic | 16:45 | |
*** athomas has quit IRC | 16:45 | |
vdrok_ | good evening Ironic! | 16:48 |
NobodyCam | morning vdrok_ :) | 16:49 |
NobodyCam | :-p | 16:49 |
vdrok_ | morning NobodyCam :) | 16:49 |
vdrok_ | There is a question about how to import translation functions from i18n. rloo suggested to add i18n module to tox.ini as import_exception and then import functions like _, _LW directly from there. What do you guys think about that? | 16:52 |
rloo | vdrok_: I think it is a grand idea ;) | 16:53 |
vdrok_ | :) | 16:53 |
*** derekh has quit IRC | 16:56 | |
lucasagomes | aight I will call it a day | 16:57 |
lucasagomes | I've to take the train back home yet | 16:57 |
lucasagomes | have a good night everybody! | 16:57 |
rloo | bye lucasagomes | 16:57 |
lucasagomes | bye bye | 16:58 |
*** lucasagomes has quit IRC | 16:58 | |
vdrok_ | rloo: seems that it's done this way in nova | 17:06 |
*** bandicot has joined #openstack-ironic | 17:06 | |
rloo | vdrok_: yeah. not that we always want to copy nova, but :-). It is also similar to how we handled it with the old gettextutils. | 17:06 |
rloo | vdrok_: of course, if others think we ought to do it differently, we can discuss. | 17:07 |
vdrok_ | rloo: sure | 17:07 |
*** tonycoffman has joined #openstack-ironic | 17:07 | |
rloo | vdrok_: it would be good to get agreement before you make changes. I don't want you to get frustrated by different opinions. | 17:08 |
*** bandicot is now known as bensonmiller | 17:09 | |
*** tatyana has quit IRC | 17:10 | |
lifeless | jroll: I'm a huge CD fan | 17:10 |
lifeless | jroll: so I totally support treating IPA as always-stable always-working-with-all-supported-Ironics | 17:11 |
vdrok_ | rloo: ok :) | 17:11 |
*** ndipanov has quit IRC | 17:11 | |
lifeless | jroll: and cutting releases out of trunk to keep inventory under control | 17:11 |
russell_h | +1 | 17:12 |
lifeless | jroll: assuming thats what you meant :) | 17:12 |
*** yuanying has quit IRC | 17:14 | |
*** harlowja_away is now known as harlowja_ | 17:25 | |
jroll | lifeless: +1 :) | 17:26 |
lifeless | jroll: I suggest starting a -dev thread on this | 17:27 |
jroll | lifeless: yeah, today is crazy but will try | 17:27 |
jroll | maybe I can con JayF into doing so | 17:27 |
lifeless | jroll: as ttx may have thoughts/concerns and it will ensure we've had wide visibility | 17:27 |
lifeless | jroll: JayF: also -operators should be given a chance to scream noooooo | 17:27 |
jroll | indeed | 17:28 |
*** jcoufal has quit IRC | 17:33 | |
*** pelix has quit IRC | 17:42 | |
aweeks | greghaynes: I've got a question for you about https://review.openstack.org/#/c/109688/ if you have a moment | 18:00 |
*** mikedillion has quit IRC | 18:05 | |
*** igordcard has quit IRC | 18:05 | |
* JayF dodges the bus | 18:07 | |
greghaynes | aweeks: Hey | 18:09 |
greghaynes | aweeks: At the airport, so I have a little time while I wait | 18:09 |
aweeks | greghaynes: I was just wondering if the idea is to make it more efficient for API nodes to decide which conductor requests need to be forwarded to | 18:10 |
greghaynes | aweeks: No, its for rebalancing when a conductor fails (its in the commit message) | 18:10 |
greghaynes | if we dont have this index we have to walk all the nodes when a failure happens | 18:11 |
*** yuanying has joined #openstack-ironic | 18:11 | |
*** penick has joined #openstack-ironic | 18:14 | |
aweeks | greghaynes: ok, I'm probably just missing something, or am not familiar enough with the code, but how is the failure detected? | 18:14 |
JayF | I think it's more that | 18:15 |
JayF | he wants to know what nodes are impacted via a mysql query | 18:15 |
greghaynes | I dont believe its implemented yet but the plan is periodic task. | 18:15 |
greghaynes | yes | 18:15 |
JayF | if he knows shiny-conductor17.mydomain.tld goes down | 18:15 |
aweeks | ok, that makes more sense | 18:16 |
greghaynes | The only way to get the list of nodes that belong to a conductor right now is to compile a list by evaluating the conductor every node should belong to | 18:16 |
aweeks | yeah | 18:16 |
lifeless | its also for ensuring tha tfailures are handled | 18:27 |
*** mikedillion has joined #openstack-ironic | 18:28 | |
*** mikedill_ has joined #openstack-ironic | 18:32 | |
*** mikedillion has quit IRC | 18:33 | |
*** mikedill_ has quit IRC | 18:34 | |
*** mikedillion has joined #openstack-ironic | 18:36 | |
*** yuanying has quit IRC | 18:44 | |
NobodyCam | brb | 18:51 |
*** mikedillion has quit IRC | 19:05 | |
* jroll shakes a fist at JayF | 19:07 | |
*** mikedillion has joined #openstack-ironic | 19:08 | |
*** harlowja_ has quit IRC | 19:08 | |
*** harlowja has joined #openstack-ironic | 19:15 | |
*** vdrok_ has quit IRC | 19:17 | |
*** harlowja has quit IRC | 19:19 | |
*** harlowja has joined #openstack-ironic | 19:28 | |
*** harlowja has quit IRC | 19:34 | |
*** harlowja has joined #openstack-ironic | 19:40 | |
openstackgerrit | Kyle Stevenson proposed a change to openstack/ironic: Add implementation of pluggable distributed sync https://review.openstack.org/115148 | 19:40 |
*** yuanying has joined #openstack-ironic | 19:41 | |
adam_g | NobodyCam, FYI WRT soft poweroff, not sure if you've seen it.. just landed in nova https://review.openstack.org/#/c/68942/32 | 19:58 |
* NobodyCam looks | 19:58 | |
NobodyCam | oh happy happy joy jpy | 19:59 |
NobodyCam | adam_g: TY | 19:59 |
NobodyCam | :) | 19:59 |
adam_g | incidentally, it breaks our driver and im getting lots of new tracebacks :| | 19:59 |
NobodyCam | :( | 20:06 |
NobodyCam | humm. teh non-voting job did pass | 20:07 |
NobodyCam | on that review | 20:07 |
adam_g | hmm | 20:08 |
adam_g | our gate looks mostly broken ATM | 20:08 |
adam_g | http://no-carrier.net/~adam/openstack/ironic_gate_status.html | 20:08 |
NobodyCam | are thouse our error? http://logs.openstack.org/39/115339/1/check/check-tempest-dsvm-virtual-ironic-nv/54e6dd5/console.html#_2014-08-21_19_56_55_713 | 20:10 |
adam_g | thats what im seeing | 20:10 |
NobodyCam | s/error/errors/ | 20:10 |
adam_g | (unrelated to that nova driver change) | 20:10 |
JayF | did a requirement bump? like perhaps a new release of keystroneclient getting installed now? </guess> | 20:12 |
*** yuanying has quit IRC | 20:15 | |
openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Update nova driver's power_off() parameters https://review.openstack.org/116093 | 20:15 |
adam_g | mrda, ^^^ | 20:16 |
adam_g | sort of looks like a tempest issue | 20:16 |
NobodyCam | adam_g: Ty for ^^ and the bug too :) | 20:16 |
rloo | someone filed a bug about the keystone thing in tempest: https://bugs.launchpad.net/ironic/+bug/1359929 | 20:18 |
adam_g | yeah | 20:20 |
adam_g | just found that | 20:20 |
openstackgerrit | Kyle Stevenson proposed a change to openstack/ironic: Add implementation of pluggable distributed sync https://review.openstack.org/115148 | 20:21 |
*** penick has quit IRC | 20:45 | |
*** stevebaker has quit IRC | 20:49 | |
*** stevebaker has joined #openstack-ironic | 20:49 | |
openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Update nova driver's power_off() parameters https://review.openstack.org/116093 | 20:55 |
*** stevebaker has quit IRC | 21:01 | |
*** stevebaker has joined #openstack-ironic | 21:02 | |
rloo | adam_g: wrt 116093, sorry to be a PITA, the other docstrings use 'Ignored by this driver.' Should we do the same here? | 21:02 |
adam_g | rloo, sure. one min | 21:03 |
rloo | adam_g: am also wondering about opening a bug to handle them in the future? | 21:03 |
rloo | adam_g: I looked in nova code. the other drivers were updated with this: "# TODO(PhilDay): Add support for timeout (clean shutdown)" | 21:06 |
NobodyCam | its all true I tell you: http://www.caninedistractions.com/when-danes-destroy-30-photos-that-capture-the-wreckage/ | 21:07 |
*** yuanying has joined #openstack-ironic | 21:11 | |
rloo | I updated the gate status in our IronicWhiteBoard. | 21:23 |
adam_g | i think ive found the issue | 21:27 |
*** jasondotstar has quit IRC | 21:27 | |
rloo | adam_g: you da man | 21:27 |
adam_g | https://review.openstack.org/#/c/116103/ | 21:34 |
adam_g | ^ that should hopefully pass the virtual-ironic job and fix us | 21:36 |
rloo | adam_g: that doesn't just affect ironic jobs? | 21:38 |
adam_g | rloo, it sort of does | 21:39 |
adam_g | rloo, devstack-vm-gate.sh calls different tox envs differently depending on the job. i think ironic is the only one that uses the code path to call tox -eall | 21:40 |
rloo | adam_g: ahh. ok. and what changed yesterday that caused this to break? | 21:40 |
*** krtaylor has quit IRC | 21:41 | |
adam_g | rloo, tox was unpinned and upgraded to 1.7.2 across slaves. that now sets the a random hashseed, and apparently lots of test code can't handle that :) | 21:41 |
rloo | adam_g: apparently :-) | 21:42 |
adam_g | rloo, http://lists.openstack.org/pipermail/openstack-dev/2014-July/041283.html | 21:42 |
*** linggao has quit IRC | 21:44 | |
rloo | adam_g: glad you can keep track of these changes. Although I don't see how dict-entry-ordering caused the keystone-xml error but hopefully it did. | 21:44 |
*** yuanying has quit IRC | 21:44 | |
openstackgerrit | Kyle Stevenson proposed a change to openstack/ironic-specs: Add spec for pluggable distributed synchronization https://review.openstack.org/114632 | 21:50 |
NobodyCam | gah check-tripleo-ironic-undercloud-precise-nonha failed | 21:53 |
NobodyCam | brb | 21:59 |
openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Update nova driver's power_off() parameters https://review.openstack.org/116093 | 22:02 |
*** rloo is now known as rloo_afk | 22:12 | |
* NobodyCam sobs just a little: http://paste.openstack.org/show/RyDz3eHQ2ydnkxZaAzCX/ | 22:15 | |
jroll | adam_g: nice find | 22:15 |
jroll | NobodyCam: please tell me that's local and not the gate | 22:15 |
NobodyCam | yes thats local | 22:15 |
NobodyCam | lol | 22:15 |
adam_g | NobodyCam, tripleo? | 22:15 |
NobodyCam | si | 22:15 |
NobodyCam | adam_g: it's my seed vm. oh :-p | 22:16 |
NobodyCam | oot@ubuntu:~# uptime 22:15:56 up 7 days, 1:53, 1 user, load average: 0.07, 0.10, 0.13 | 22:16 |
*** scubacuda has joined #openstack-ironic | 22:17 | |
adam_g | NobodyCam, wonder how big the keystone token table is | 22:17 |
NobodyCam | adam_g: you know the command to check? | 22:17 |
adam_g | NobodyCam, i usually just mysql -c keystone -e 'select count(*) from token' | 22:19 |
NobodyCam | lol | 22:22 |
NobodyCam | +----------+ | 22:22 |
NobodyCam | | count(*) | | 22:22 |
NobodyCam | +----------+ | 22:22 |
NobodyCam | | 2 | | 22:22 |
NobodyCam | +----------+ | 22:22 |
*** jgrimm has quit IRC | 22:22 | |
* NobodyCam dose happy dance | 22:25 | |
NobodyCam | http://paste.openstack.org/show/vodkbUatKRI5siIrJJnE/ | 22:25 |
jroll | NobodyCam: http://www.downloadmoreram.com/ | 22:25 |
NobodyCam | lol | 22:25 |
jroll | what's sad, is when I searched for that, google's autocomplete wanted to add "legit" and "virus" | 22:26 |
jroll | oh ha https://answers.yahoo.com/question/index?qid=20111123175542AA3CoOH | 22:26 |
NobodyCam | lol You make me weep for humanity. | 22:27 |
NobodyCam | lov that one | 22:27 |
jroll | yeahhhhhh | 22:28 |
NobodyCam | tho with gmailfs (http://en.wikipedia.org/wiki/GmailFS) and a swap file it could look like downloaded ram... lol | 22:29 |
jroll | hahaha | 22:30 |
jroll | raas - ram as a service | 22:30 |
JayF | I bet you could really break a machine | 22:30 |
JayF | with a flakey fuse driver | 22:30 |
NobodyCam | lol.. that would actually be really cool | 22:30 |
JayF | and a swapfile on that filesystem | 22:30 |
NobodyCam | oh ya | 22:30 |
jroll | oh god | 22:30 |
NobodyCam | and quick it is not lol | 22:31 |
*** yuanying has joined #openstack-ironic | 22:41 | |
mrda | adam_g: is this something we MUST have in the ironci driver for J? | 22:50 |
adam_g | mrda, yeah | 22:50 |
adam_g | mrda, tempest wont pass without it | 22:51 |
adam_g | :( | 22:51 |
mrda | ok, I'm on a plane today, but I should get that done somewhere where's there's internets :) | 22:51 |
* mrda makes a note to himself | 22:51 | |
*** shakamunyi has joined #openstack-ironic | 22:51 | |
adam_g | mrda, coo.. thanks and sorry. nova updated the driver yesterday | 22:51 |
mrda | if this comes up again, add it to the Whiteboard and ping me where too. That way it won't get skipped :) | 22:52 |
NobodyCam | question: do we need to fill out all the fields? RE: http://paste.openstack.org/show/jNQpecPEYVEvxAeFx5B1 | 22:59 |
NobodyCam | jroll: JayF: JoshNang: TY for the list(detail=true) :) +1 | 23:02 |
NobodyCam | lol anyone have to have a nova baremetal up. I need a sample of nova baremetal-node-show <blah> | 23:06 |
NobodyCam | s/have to/happen to/ | 23:07 |
NobodyCam | nm: http://docs.openstack.org/user-guide/content/cli_nova_baremetal-details.html | 23:08 |
*** yuanying has quit IRC | 23:14 | |
adam_g | hmm. looks like we need a patch to land in stable/havana tempest before https://review.openstack.org/#/c/116103/ can merge and fix us | 23:20 |
*** yuanying has joined #openstack-ironic | 23:30 | |
NobodyCam | :( | 23:30 |
NobodyCam | anyone have a minute to take a quick look at http://paste.openstack.org/show/ijF44Q6nzoJwYAWOGakP | 23:30 |
NobodyCam | mrda: ^^^^^^ | 23:31 |
*** chuckC_ has quit IRC | 23:31 | |
NobodyCam | brb | 23:31 |
NobodyCam | back :) | 23:47 |
adam_g | NobodyCam, could the interfaces field of the nova node show the result of a node-port-list call? | 23:53 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!