| openstackgerrit | Josh Gachnang proposed openstack/ironic: Improve IPA client library https://review.openstack.org/111118 | 00:02 |
|---|---|---|
| *** GheRivero has quit IRC | 00:03 | |
| *** hemna has joined #openstack-ironic | 00:11 | |
| *** ryanpetrello has quit IRC | 00:15 | |
| *** GheRivero has joined #openstack-ironic | 00:15 | |
| *** andreykurilin has quit IRC | 00:20 | |
| openstackgerrit | Steven Dake proposed openstack/ironic-specs: Override boot options via glance property https://review.openstack.org/144235 | 00:24 |
| *** hemna has quit IRC | 00:27 | |
| *** ijw has joined #openstack-ironic | 00:29 | |
| *** hemna has joined #openstack-ironic | 00:30 | |
| *** ijw has quit IRC | 00:30 | |
| *** ijw has joined #openstack-ironic | 00:31 | |
| *** ChuckC has quit IRC | 00:32 | |
| *** ijw has quit IRC | 00:33 | |
| NobodyCam | ok gots to drop from the internets for a few... | 00:33 |
| *** Masahiro has joined #openstack-ironic | 00:33 | |
| *** hemna has quit IRC | 00:34 | |
| *** david-ly_ has joined #openstack-ironic | 00:36 | |
| *** Masahiro has quit IRC | 00:38 | |
| *** david-ly_ has quit IRC | 00:41 | |
| *** hemna has joined #openstack-ironic | 00:42 | |
| *** hemna_ has joined #openstack-ironic | 00:47 | |
| *** hemna has quit IRC | 00:48 | |
| *** zhidong has joined #openstack-ironic | 00:48 | |
| *** ryanpetrello has joined #openstack-ironic | 00:49 | |
| *** achanda has joined #openstack-ironic | 00:50 | |
| *** achanda has quit IRC | 00:55 | |
| *** achanda has joined #openstack-ironic | 00:56 | |
| *** Masahiro has joined #openstack-ironic | 00:56 | |
| *** achanda has quit IRC | 01:01 | |
| *** ChuckC has joined #openstack-ironic | 01:03 | |
| *** hemna_ has quit IRC | 01:06 | |
| *** hemna_ has joined #openstack-ironic | 01:07 | |
| *** GheRivero has quit IRC | 01:10 | |
| *** GheRivero has joined #openstack-ironic | 01:11 | |
| *** GheRivero has quit IRC | 01:20 | |
| *** hemna_ has quit IRC | 01:24 | |
| *** hemna_ has joined #openstack-ironic | 01:25 | |
| *** GheRivero has joined #openstack-ironic | 01:27 | |
| *** chenglch has joined #openstack-ironic | 01:29 | |
| *** nosnos has joined #openstack-ironic | 01:50 | |
| *** penick has joined #openstack-ironic | 01:53 | |
| *** penick_ has joined #openstack-ironic | 01:54 | |
| *** hemna_ has quit IRC | 01:54 | |
| *** hemna_ has joined #openstack-ironic | 01:55 | |
| *** penick has quit IRC | 01:57 | |
| *** penick_ is now known as penick | 01:57 | |
| *** rwsu has quit IRC | 02:03 | |
| *** hemna_ has quit IRC | 02:06 | |
| *** ChuckC has quit IRC | 02:07 | |
| *** hemna_ has joined #openstack-ironic | 02:07 | |
| *** GheRivero has quit IRC | 02:09 | |
| *** ryanpetrello has quit IRC | 02:14 | |
| *** GheRivero has joined #openstack-ironic | 02:15 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-specs: Make sure the filename is the same as the blueprint name https://review.openstack.org/145088 | 02:15 |
| *** ChuckC has joined #openstack-ironic | 02:18 | |
| *** eghobo has quit IRC | 02:27 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Improve IPA client library https://review.openstack.org/111118 | 02:30 |
| *** yjiang5 has left #openstack-ironic | 02:46 | |
| *** chenglch|2 has joined #openstack-ironic | 02:58 | |
| *** vinbs has joined #openstack-ironic | 03:01 | |
| *** chenglch has quit IRC | 03:01 | |
| *** dlaube has quit IRC | 03:11 | |
| *** hemna_ has quit IRC | 03:15 | |
| *** pensu has joined #openstack-ironic | 03:21 | |
| *** jerryz has joined #openstack-ironic | 03:32 | |
| *** nosnos has quit IRC | 03:34 | |
| *** Haomeng has joined #openstack-ironic | 03:34 | |
| *** Haomeng|2 has quit IRC | 03:34 | |
| *** ryanpetrello has joined #openstack-ironic | 03:38 | |
| *** pensu has quit IRC | 03:40 | |
| *** chenglch|2 has quit IRC | 03:42 | |
| *** david-lyle has joined #openstack-ironic | 03:48 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Allow use of multiple simultaneous HW managers https://review.openstack.org/143193 | 03:49 |
| *** penick has quit IRC | 03:51 | |
| *** Nisha has joined #openstack-ironic | 03:52 | |
| *** penick has joined #openstack-ironic | 03:54 | |
| *** ryanpetrello has quit IRC | 03:55 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Allow use of multiple simultaneous HW managers https://review.openstack.org/143193 | 03:58 |
| *** ramineni has joined #openstack-ironic | 03:59 | |
| *** naohirot has joined #openstack-ironic | 04:04 | |
| naohirot | GA ironic! | 04:05 |
| *** rloo has quit IRC | 04:07 | |
| *** pensu has joined #openstack-ironic | 04:14 | |
| *** david-lyle is now known as david-lyle_afk | 04:25 | |
| *** nosnos has joined #openstack-ironic | 04:29 | |
| *** rameshg87 has joined #openstack-ironic | 04:30 | |
| Nisha | #join openstack-meeting-3 | 04:36 |
| *** chenglch has joined #openstack-ironic | 04:36 | |
| *** dlaube has joined #openstack-ironic | 04:36 | |
| *** pensu has quit IRC | 04:40 | |
| *** hshiina has joined #openstack-ironic | 04:43 | |
| *** penick has quit IRC | 04:43 | |
| *** lsmola has quit IRC | 04:45 | |
| *** pcrews has quit IRC | 04:50 | |
| devananda | hi folks! | 04:55 |
| devananda | almost meeting time | 04:55 |
| Haomeng | devananda: ok | 04:59 |
| *** sdake has quit IRC | 04:59 | |
| *** lsmola has joined #openstack-ironic | 05:00 | |
| *** lintan has joined #openstack-ironic | 05:01 | |
| *** eghobo has joined #openstack-ironic | 05:07 | |
| *** chenglch has quit IRC | 05:22 | |
| *** k4n0 has joined #openstack-ironic | 05:24 | |
| *** k4n0 has quit IRC | 05:25 | |
| *** rakesh_hs has joined #openstack-ironic | 05:26 | |
| *** eghobo has quit IRC | 05:26 | |
| *** pensu has joined #openstack-ironic | 05:27 | |
| *** pradipta_away is now known as pradipta | 05:29 | |
| *** eghobo has joined #openstack-ironic | 05:30 | |
| jroll | NobodyCam: imagine how you might suspend an ironic instance for non-payment, you don't want to destroy data in case they do end up paying, right? | 05:55 |
| NobodyCam | jroll: https://review.openstack.org/#/c/107778/ | 05:56 |
| jroll | NobodyCam: can you link me to somewhere I can download extra time in the day, though? | 05:57 |
| jroll | :P | 05:57 |
| jroll | I'm aware of this one | 05:57 |
| jroll | but ty :) | 05:57 |
| rameshg87 | NobodyCam, why was it dropped anyway ? | 06:04 |
| devananda | jroll: snapshot :) | 06:05 |
| jroll | devananda: I may or may not get that joke(?) | 06:05 |
| jroll | or do you mean, snapshot an ironic instance | 06:06 |
| jroll | because wat | 06:06 |
| devananda | jroll: yes. reboot into IPA, compress data, archive to separate storage | 06:06 |
| jroll | hmm, that may be doable. | 06:06 |
| jroll | yeah | 06:06 |
| jroll | how do you do that in a sane fashion with 3TB of data? :) | 06:06 |
| russell_h | swift | 06:07 |
| devananda | also, 107778 was abandoned because it used a conf option to ALWAYS soft-off | 06:07 |
| devananda | rather than per-request | 06:07 |
| jroll | russell_h: I bet our swift group would love for us to push/pull 3tb at once | 06:07 |
| russell_h | eh, they'd survive | 06:07 |
| devananda | jroll: imagine if someone snapshotted a bare metal swift node into swift ... | 06:07 |
| jroll | heh | 06:07 |
| russell_h | jroll: the real problem would be when we do it 30x at once | 06:08 |
| devananda | there are clearly some things you shouldn't do | 06:08 |
| jroll | right. | 06:08 |
| devananda | but, cold-migrate an instance like taht? way cool | 06:08 |
| jroll | super cool | 06:08 |
| lintan | jroll: I like the idea to take a snapshot of instance | 06:08 |
| devananda | boot IPA on two nodes, stream the disk from one to the other, update ports and so on, reboot | 06:08 |
| russell_h | snapshots are very much on our internal wish list | 06:08 |
| russell_h | at least of boot disks | 06:08 |
| * jroll learned something today ^ | 06:09 | |
| lintan | We also need this in our environment, yes boot disks | 06:09 |
| jroll | :P | 06:09 |
| jroll | yeah, this is super interesting | 06:10 |
| * jroll wants 'nova suspend' first | 06:10 | |
| russell_h | jroll: I wrote some thing that was archiving all the data cloud monitoring collected to swift | 06:10 |
| jroll | or a 'nova stop' that doesn't unplug a computer | 06:10 |
| devananda | jroll: nova pause ? | 06:10 |
| russell_h | jroll: by the time they added compression it was uploading a 500MB JSON file every few seconds | 06:11 |
| jroll | devananda: maybe | 06:11 |
| jroll | devananda: 'a way to cut off instance access without a hard power off' | 06:11 |
| russell_h | unfortunately for my street cred it turns out JSON compresses really well | 06:11 |
| jroll | we have a hack for this today but it isn't awesome | 06:11 |
| jroll | (kill the ports) | 06:11 |
| devananda | jroll: hm. that's just "dom.suspend()" | 06:12 |
| devananda | probably not the right thing | 06:12 |
| jroll | devananda: eh? ironic doesn't support anything like this | 06:12 |
| devananda | also, i have no idea how to do that generically for hardware | 06:12 |
| jroll | well, soft power off would be a start | 06:12 |
| devananda | nope | 06:12 |
| devananda | that's TOTALLY different | 06:12 |
| jroll | dunno how generic that would be, AMT probably doesn't support it | 06:13 |
| jroll | why | 06:13 |
| devananda | you get all kinds of local state changes when you ask an OS to shut down | 06:13 |
| russell_h | nutty question, can an iLO dump the contents of RAM? | 06:13 |
| jroll | nova stop && nova lock | 06:13 |
| jroll | sure | 06:13 |
| jroll | but like | 06:13 |
| jroll | you can't suspend hardware. | 06:13 |
| russell_h | (or any other high-end-ish BMC) | 06:13 |
| jroll | it's the closest you can get imo | 06:13 |
| jroll | (with any hardware I'm aware of, my knowledge is not vast) | 06:14 |
| devananda | russell_h: sure, but you can feel safe because it doesn't have a documented API for that | 06:14 |
| russell_h | lmao | 06:14 |
| lintan | maybe we should not think about the snapshot of memory at first? | 06:14 |
| devananda | russell_h: sorry, it's late, my humor is showing :) | 06:14 |
| jroll | read: drinks are poured | 06:14 |
| russell_h | devananda: its cool, I expect the same is true of Intel AMT crap | 06:14 |
| russell_h | which is even safer because its barely documented at all | 06:15 |
| jroll | because it's for... desktops. | 06:15 |
| jroll | at any rate | 06:15 |
| devananda | laptops too | 06:15 |
| jroll | I think soft power off is a good first step | 06:15 |
| jroll | devananda: !servers, I should say | 06:16 |
| devananda | jroll: I agree. nothing wrong with supporting soft power off in a generic (and optional) way | 06:16 |
| devananda | keeping in mind that some power managers may not implement it | 06:16 |
| devananda | eg, a PDU | 06:16 |
| jroll | yeah | 06:16 |
| devananda | so calls to eg. "ironic node-set-power-state --soft off" need to error appropriately in that case | 06:17 |
| jroll | I'd also love 'nova stop' to default to stop and handle that error appropriately | 06:17 |
| jroll | default to soft* | 06:18 |
| devananda | also: ssh $host "sudo shutdown now" | 06:18 |
| devananda | jroll: with appropriate timeouts and falling back to --hard, yes, we should do that | 06:18 |
| jroll | oh, have you heard our funny story about "sudo shutdown"? | 06:18 |
| devananda | nope | 06:18 |
| jroll | so nova has a loop that checks power status for instances, right? | 06:19 |
| jroll | (keep in mind we do force_power_state_sync=false | 06:19 |
| jroll | so someone shuts an instance down, ironic notices and records power_state=off | 06:19 |
| jroll | nova sees that and decides it needs to call the stop() api | 06:19 |
| jroll | which calls virt_driver.stop() | 06:20 |
| jroll | so then it's stopped in nova, powered off in ironic | 06:20 |
| jroll | nova start should work fine in this case, but rackspace doesn't expose that | 06:21 |
| jroll | so theoretically you're stuck, but turns out nova reboot will turn it back on | 06:21 |
| jroll | but it took some time to figure that out | 06:21 |
| jroll | I guess it's funnier if you dealt with the wtf :P | 06:22 |
| devananda | oh, why don't you expose nova start? :p | 06:23 |
| russell_h | still trying to figure that out | 06:23 |
| jroll | don't ask me | 06:23 |
| russell_h | I'm not convinced anyone knows | 06:23 |
| devananda | that I can evan ask that question makes me really sad | 06:23 |
| jroll | I live in a cell, I don't own any nova-api nodes :P | 06:23 |
| russell_h | my guess: since stop isn't exposed it never occured to anyone to expose start | 06:23 |
| jroll | russell_h: that just transforms the question :P | 06:24 |
| *** killer_prince is now known as lazy_prince | 06:24 | |
| russell_h | yeah but thats an easier question | 06:24 |
| russell_h | my guess on that one: people thought stopping a server meant they don't get billed | 06:24 |
| jroll | true | 06:25 |
| russell_h | 5 hours until SpaceX launch | 06:25 |
| jroll | api_response = 'you will still be billed' + api_response | 06:25 |
| russell_h | who is pumped | 06:25 |
| jroll | heh, staying up for that? | 06:25 |
| russell_h | not sure yet | 06:25 |
| jroll | or waking up early? | 06:25 |
| russell_h | I plan to be awake somehow | 06:25 |
| jroll | gotta be ready to trade | 06:25 |
| *** rameshg87 has quit IRC | 06:26 | |
| russell_h | SpaceX is private :( | 06:26 |
| jroll | oh. | 06:26 |
| jroll | but elon musk confidence changes must affect tesla amirit | 06:26 |
| *** rameshg87 has joined #openstack-ironic | 06:26 | |
| russell_h | probably | 06:26 |
| russell_h | Elon lost me a lot of money today | 06:26 |
| russell_h | I mean it wasn't his fault | 06:27 |
| russell_h | but between SCTY and TSLA | 06:27 |
| russell_h | and everything else | 06:27 |
| russell_h | rough day | 06:27 |
| jroll | that's what you get for going all in on TSLA | 06:27 |
| jroll | :P | 06:27 |
| *** chenglch has joined #openstack-ironic | 06:28 | |
| * jroll signs off for the night, see everyone tomorrow | 06:29 | |
| *** jiangfei has quit IRC | 06:41 | |
| *** jiangfei has joined #openstack-ironic | 06:42 | |
| jiangfei | can ironic deploy in docker, someone kown about it ? | 06:56 |
| Haomeng | jiangfei: good idea, I think the points is networking | 06:57 |
| jiangfei | Haomeng: hm, do you mean neutron-dhcp-server | 06:59 |
| Haomeng | jiangfei: yes, not sure if our ovs and neutron agents can be run in docker | 06:59 |
| Haomeng | jiangfei: I mean the docker networking env, if the os running in docker can receive the physical traffic, such as the baremetal dhcp discovery request | 07:00 |
| jiangfei | Haomeng: i see some news about: 'The docker driver does not work with neutron' | 07:01 |
| Haomeng | jiangfei: yes | 07:01 |
| Haomeng | jiangfei: so we just put ironic api and conductor into docker, that is not very useful:) | 07:02 |
| Haomeng | jiangfei: but it is easy for installation | 07:02 |
| jiangfei | Haomeng:yes | 07:03 |
| *** dlaube has quit IRC | 07:25 | |
| *** ifarkas has joined #openstack-ironic | 07:26 | |
| *** chlong has quit IRC | 07:49 | |
| *** chlong has joined #openstack-ironic | 07:49 | |
| *** chlong has quit IRC | 07:51 | |
| *** chlong has joined #openstack-ironic | 07:53 | |
| *** teju has joined #openstack-ironic | 07:58 | |
| *** vinbs_ has joined #openstack-ironic | 08:01 | |
| *** vinbs has quit IRC | 08:02 | |
| *** vinbs_ is now known as vinbs | 08:02 | |
| *** Nisha has quit IRC | 08:31 | |
| *** Nisha has joined #openstack-ironic | 08:38 | |
| *** eghobo has quit IRC | 08:43 | |
| *** hshiina has left #openstack-ironic | 08:48 | |
| *** jistr has joined #openstack-ironic | 08:58 | |
| *** jiangfei has quit IRC | 08:59 | |
| *** jiangfei has joined #openstack-ironic | 08:59 | |
| *** dtantsur|afk is now known as dtantsur | 09:00 | |
| *** jcoufal has joined #openstack-ironic | 09:03 | |
| *** jcoufal has quit IRC | 09:03 | |
| dtantsur | Morning | 09:04 |
| *** derekh has joined #openstack-ironic | 09:08 | |
| *** erwan_taf has joined #openstack-ironic | 09:09 | |
| *** athomas has joined #openstack-ironic | 09:12 | |
| *** zhidong has quit IRC | 09:21 | |
| *** Masahiro has quit IRC | 09:27 | |
| *** Masahiro has joined #openstack-ironic | 09:27 | |
| *** Masahiro has quit IRC | 09:31 | |
| *** Masahiro has joined #openstack-ironic | 09:37 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic-specs: New driver interface for RAID configuration https://review.openstack.org/135899 | 09:38 |
| *** Masahiro has quit IRC | 09:39 | |
| Nisha | dtantsur, morning | 09:39 |
| Nisha | Happy new year | 09:39 |
| dtantsur | Nisha, hi! Happy new year (if applies :) | 09:40 |
| Nisha | dtantsur, :) | 09:41 |
| Nisha | dtantsur, hope you had wonderful vacations | 09:41 |
| erwan_taf | hey dtantsur | 09:41 |
| dtantsur | yeah, pretty awesome :) | 09:42 |
| dtantsur | erwan_taf, o/ | 09:42 |
| dtantsur | Nisha, and you? | 09:42 |
| Nisha | dtantsur, yes me too | 09:42 |
| Nisha | dtantsur, i saw your +2 on the discovery spec but it is not yet approved.... any reason? | 09:43 |
| Nisha | now it has 2 +2 | 09:43 |
| Nisha | and 2 +1 | 09:43 |
| Nisha | dtantsur, and no comments so far | 09:43 |
| dtantsur | Nisha, just wanted other folks to have a look. I'm afraid I've influenced it too much :) | 09:43 |
| Nisha | hehe | 09:43 |
| dtantsur | but as nobody jumped in, I think it's fine for me to approve it now | 09:44 |
| Nisha | :) | 09:44 |
| dtantsur | +A | 09:46 |
| Nisha | dtantsur, thanks | 09:46 |
| dtantsur | Nisha, thank you for all your efforts :) | 09:47 |
| *** GheRivero has quit IRC | 09:48 | |
| Nisha | dtantsur, thanks for all the help on this | 09:48 |
| Nisha | just hope code also goes in as soon as i post it | 09:48 |
| Nisha | i think i need to wait for states code to get merged frst? | 09:48 |
| *** GheRivero has joined #openstack-ironic | 09:48 | |
| *** Masahiro has joined #openstack-ironic | 09:50 | |
| openstackgerrit | Merged openstack/ironic-specs: Discover node properties using new target inspect with node-set-provision-state https://review.openstack.org/100951 | 09:52 |
| *** Masahiro has quit IRC | 09:53 | |
| *** andreykurilin has joined #openstack-ironic | 09:53 | |
| *** GheRivero has quit IRC | 09:53 | |
| *** GheRivero has joined #openstack-ironic | 09:53 | |
| dtantsur | Nisha, yeah, we should. anyway I think you get propose some peaces already. | 09:57 |
| *** andreykurilin has quit IRC | 09:57 | |
| *** Masahiro has joined #openstack-ironic | 09:57 | |
| dtantsur | * get = can | 09:57 |
| *** andreykurilin has joined #openstack-ironic | 09:58 | |
| dtantsur | * peace = piece | 09:58 |
| * dtantsur gets more coffee | 09:58 | |
| Nisha | dtantsur, :) | 09:58 |
| Nisha | will do | 09:58 |
| Nisha | thanks | 09:58 |
| Nisha | its merged | 09:58 |
| *** Masahiro has quit IRC | 10:00 | |
| *** alexpilotti has joined #openstack-ironic | 10:02 | |
| *** vinbs has quit IRC | 10:04 | |
| *** jcoufal has joined #openstack-ironic | 10:04 | |
| *** Masahiro has joined #openstack-ironic | 10:09 | |
| *** vinbs has joined #openstack-ironic | 10:10 | |
| *** Masahiro has quit IRC | 10:14 | |
| *** vdrok has joined #openstack-ironic | 10:15 | |
| *** pradipta is now known as pradipta_away | 10:15 | |
| *** Nisha has quit IRC | 10:19 | |
| *** GheRivero has quit IRC | 10:20 | |
| *** naohirot has quit IRC | 10:26 | |
| *** igordcard has joined #openstack-ironic | 10:29 | |
| *** igordcard has quit IRC | 10:29 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Store introspection result in the local database https://review.openstack.org/145008 | 10:30 |
| dtantsur | ifarkas, ^^^ | 10:30 |
| *** GheRivero has joined #openstack-ironic | 10:31 | |
| *** vinbs has quit IRC | 10:31 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Introduce __version__ and __version_info__ https://review.openstack.org/145192 | 10:34 |
| gilliard | Hi Can someone help me understand what I ought to do about https://review.openstack.org/#/c/142782/ ? | 10:42 |
| ifarkas | dtantsur, will take a look in that | 10:43 |
| gilliard | I'm planning on splitting up the get_image_info method into two, for prepare() and clean_up() - but that won't fix the related issue raised at https://bugs.launchpad.net/ironic/+bug/1405448 | 10:44 |
| *** chenglch has quit IRC | 10:46 | |
| gilliard | To fix that case, I think we need to know whether we're doing a deploy or a rebuild. But both cases are handled by the same code path. | 10:46 |
| *** GheRivero has quit IRC | 10:46 | |
| dtantsur | gilliard, I think these 2 bugs/cases can be handled separately... | 10:47 |
| *** GheRivero has joined #openstack-ironic | 10:47 | |
| gilliard | dtantsur: The bugs are marked as being duplicates, which I sort-of agree with. Although they describe separate problems. | 10:48 |
| gilliard | 2 patches with a partial-bug each? | 10:48 |
| dtantsur | yep | 10:48 |
| gilliard | OK. Thanks. | 10:49 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: In-band hardware properties inspection via ironic-discoverd https://review.openstack.org/135605 | 10:50 |
| gilliard | I'm still not sure what approach to take with the 2nd one though. The image cache needs to be told whether it needs to flush the kernel/ramdisk (rebuild) or if it can reuse them | 10:51 |
| *** Masahiro has joined #openstack-ironic | 10:54 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Introduce __version__ and __version_info__ https://review.openstack.org/145192 | 10:56 |
| *** Masahiro has quit IRC | 10:58 | |
| *** ramineni has quit IRC | 11:00 | |
| *** Masahiro has joined #openstack-ironic | 11:00 | |
| *** pelix has joined #openstack-ironic | 11:04 | |
| *** lucasagomes has joined #openstack-ironic | 11:05 | |
| lucasagomes | morning | 11:18 |
| lucasagomes | just to have an idea, who is willing to go to the mid-cycle meetup in grenoble? | 11:18 |
| *** rameshg87 has quit IRC | 11:22 | |
| dtantsur | lucasagomes, I'd like to, but I doubt I'll be sent :) | 11:23 |
| *** GheRivero has quit IRC | 11:24 | |
| *** GheRivero has joined #openstack-ironic | 11:24 | |
| lucasagomes | dtantsur, ack, I'm asking to ask devananda to try to find out the number of participants | 11:25 |
| *** naohirot has joined #openstack-ironic | 11:27 | |
| *** bradjones has quit IRC | 11:28 | |
| *** bradjones has joined #openstack-ironic | 11:30 | |
| *** andreykurilin has quit IRC | 11:31 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add iRMC Driver and its iRMC Power module https://review.openstack.org/144901 | 11:43 |
| *** pensu has quit IRC | 11:45 | |
| *** Masahiro has quit IRC | 11:48 | |
| * dtantsur short walk, brb | 11:53 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Store introspection result in the local database https://review.openstack.org/145008 | 11:57 |
| *** pradipta_away is now known as pradipta | 12:03 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: iRMC Virtual Media Deploy Driver for Ironic https://review.openstack.org/134865 | 12:26 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: iRMC Management Driver for Ironic https://review.openstack.org/136020 | 12:35 |
| *** ryanpetrello has joined #openstack-ironic | 12:45 | |
| *** Masahiro has joined #openstack-ironic | 12:49 | |
| *** dprince has joined #openstack-ironic | 12:51 | |
| *** Masahiro has quit IRC | 12:54 | |
| *** teju has left #openstack-ironic | 12:59 | |
| *** pensu has joined #openstack-ironic | 13:04 | |
| *** rakesh_hs has quit IRC | 13:06 | |
| dtantsur | lucasagomes, heh, didn't notice that I have to be back on Feb 6th for DevConf here in Brno to talk about Ironic :) midcycle is Feb 3-5, right? | 13:10 |
| lucasagomes | dtantsur, yes, mid-cycle is 3-5 | 13:11 |
| lucasagomes | devconf cool! | 13:11 |
| dtantsur | lucasagomes, ask mgmt to bring you for devconf, let's have midcycle afterparty ;) | 13:12 |
| *** pradipta is now known as pradipta_away | 13:16 | |
| lucasagomes | dtantsur, that would be awesome! I would love to | 13:16 |
| lucasagomes | I will see if I can go, it's pretty close :) | 13:17 |
| *** vdrok has quit IRC | 13:21 | |
| *** athomas has quit IRC | 13:22 | |
| openstackgerrit | Matthew Gilliard proposed openstack/ironic: Check glance for new kernel and ramdisk on rebuild https://review.openstack.org/142782 | 13:24 |
| openstackgerrit | Matthew Gilliard proposed openstack/ironic: Check glance for new kernel and ramdisk on prepare https://review.openstack.org/142782 | 13:24 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: PXE/ISCSI: Call partprobe after partioning the disk https://review.openstack.org/142838 | 13:38 |
| *** athomas has joined #openstack-ironic | 13:39 | |
| *** jjohnson2 has joined #openstack-ironic | 13:39 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Discover node properties for iLO drivers https://review.openstack.org/103007 | 13:52 |
| *** Nisha has joined #openstack-ironic | 13:54 | |
| *** rloo has joined #openstack-ironic | 13:56 | |
| *** Masahiro has joined #openstack-ironic | 14:00 | |
| *** pensu has quit IRC | 14:03 | |
| *** nosnos has quit IRC | 14:04 | |
| *** Masahiro has quit IRC | 14:05 | |
| *** naohirot has quit IRC | 14:11 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Make database a required configuration option https://review.openstack.org/145231 | 14:20 |
| dtantsur | lucasagomes, do you want me to update https://review.openstack.org/#/c/144769 for you (if you agree with comments)? | 14:22 |
| *** kragniz has quit IRC | 14:22 | |
| * lucasagomes checks | 14:22 | |
| lucasagomes | dtantsur, ah wait, because we might not need it | 14:22 |
| dtantsur | ack | 14:22 |
| *** kragniz has joined #openstack-ironic | 14:22 | |
| *** lazy_prince is now known as killer_prince | 14:28 | |
| *** vdrok has joined #openstack-ironic | 14:30 | |
| *** GheRivero has quit IRC | 14:42 | |
| *** GheRivero has joined #openstack-ironic | 14:49 | |
| *** vipul has quit IRC | 14:52 | |
| *** vipul has joined #openstack-ironic | 14:52 | |
| *** jistr has quit IRC | 14:56 | |
| *** jistr has joined #openstack-ironic | 14:57 | |
| *** hypnosb has quit IRC | 14:59 | |
| *** hypnosb has joined #openstack-ironic | 14:59 | |
| *** krtaylor has quit IRC | 15:03 | |
| NobodyCam | morning Ironic | 15:04 |
| *** zz_jgrimm is now known as jgrimm | 15:05 | |
| *** krtaylor has joined #openstack-ironic | 15:11 | |
| NobodyCam | do we have a way to force a images download. I has some folks that are replacing the image with a new image but seems like the orginal image is cached and not downloading the new image | 15:20 |
| *** rameshg87 has joined #openstack-ironic | 15:21 | |
| lucasagomes | NobodyCam, morning | 15:29 |
| lucasagomes | NobodyCam, the final user image? or ramdisks? | 15:30 |
| lucasagomes | not sure if we actually look whether the image have been updated or not | 15:30 |
| lucasagomes | maybe worth looking at the image_cache.py module | 15:30 |
| * lucasagomes 1 sec, going for lunch with my fam | 15:30 | |
| lucasagomes | bbl | 15:30 |
| *** lucasagomes is now known as lucas-hungry | 15:30 | |
| *** MattMan has quit IRC | 15:31 | |
| *** MattMan has joined #openstack-ironic | 15:31 | |
| NobodyCam | morning lucas-hungry | 15:32 |
| jroll | morning all :) | 15:34 |
| jroll | NobodyCam: don't we cache the image by glance uuid? | 15:35 |
| jroll | afaik you can't change an image and keep the same uuid | 15:35 |
| jroll | via glance api anyway | 15:35 |
| gilliard | We cache by <node-uuid>/<label> where label is 'disk' or 'kernel' etc | 15:40 |
| gilliard | https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/image_cache.py#L78 | 15:40 |
| gilliard | So if the node uuid doesn't change, the cache doesn't seem to download new images. EG during nova rebuild. | 15:41 |
| NobodyCam | gilliard: morning jroll | 15:42 |
| NobodyCam | gah | 15:42 |
| gilliard | morning :) | 15:42 |
| NobodyCam | jroll: morning | 15:42 |
| NobodyCam | gilliard: morning too | 15:42 |
| jroll | gilliard: it seems to use the glance id | 15:42 |
| jroll | https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/image_cache.py#L101-102 | 15:42 |
| jroll | master_file_name is the glance uuid there | 15:43 |
| gilliard | yes, but we return early if the dest_path exists a few lines later. | 15:43 |
| jroll | or at least it caches images by uuid, then links as you say | 15:43 |
| *** jcoufal_ has joined #openstack-ironic | 15:44 | |
| jroll | mmm. | 15:44 |
| jroll | nice one | 15:44 |
| jroll | is there a bug filed for this? | 15:44 |
| NobodyCam | ahh so unlinking would allow it wee the new uuid? | 15:44 |
| jroll | we should check if that path exists *and* links to the correct image | 15:44 |
| Nisha | hi, i am facing the issue spoken of in http://lists.openstack.org/pipermail/openstack-dev/2014-September/046094.html on 14.04 Ubuntu also now with latest devstack | 15:45 |
| Nisha | does anyone know the workaround? | 15:45 |
| NobodyCam | morning Nisha | 15:45 |
| jroll | if someone writes up a bug, I'm going to throw up a quick patch | 15:45 |
| Nisha | NobodyCam, morning | 15:45 |
| NobodyCam | ++ sudo ovs-vsctl -- --if-exists del-port | 15:46 |
| NobodyCam | ovs-tap1 -- add-port br-int ovs-tap1 tag= | 15:46 |
| Nisha | yes | 15:46 |
| NobodyCam | hummm | 15:46 |
| *** jcoufal has quit IRC | 15:47 | |
| Nisha | i just updated the devstack | 15:47 |
| Nisha | and it is failing at this point | 15:47 |
| Shrews | Nisha: 12.04? | 15:48 |
| Nisha | no 14.04 | 15:48 |
| Shrews | oh, hrm | 15:48 |
| *** rwsu has joined #openstack-ironic | 15:48 | |
| Shrews | Nisha: did you try recloning everything? | 15:48 |
| Nisha | yes | 15:48 |
| Nisha | i removed old stack | 15:48 |
| Nisha | i cloned devstack freshly and ran stack | 15:48 |
| *** Masahiro has joined #openstack-ironic | 15:49 | |
| rameshg87 | JayF, hi | 15:50 |
| Nisha | actually issue is happening here | 15:50 |
| openstackgerrit | Matthew Gilliard proposed openstack/ironic: Check glance for new kernel and ramdisk on prepare https://review.openstack.org/142782 | 15:50 |
| Nisha | 2015-01-06 14:09:46.265 | +++ sudo ip netns exec qdhcp-6a65f545-1d41-4dab-8a12-1702965c0d72 ip link list | 15:50 |
| Nisha | 2015-01-06 14:09:46.297 | Cannot open network namespace "qdhcp-6a65f545-1d41-4dab-8a12-1702965c0d72": No such file or directory | 15:50 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Check if image link points to the correct image https://review.openstack.org/145254 | 15:50 |
| gilliard | jroll: yes there's a bug https://bugs.launchpad.net/ironic/+bug/1401501 | 15:51 |
| *** subscope has quit IRC | 15:51 | |
| jroll | gilliard: I would think that patch I just made should fix this, but I haven't tested | 15:51 |
| jroll | though I see you're doing it, and very differently | 15:52 |
| *** BertieFulton has joined #openstack-ironic | 15:52 | |
| rloo | NobodyCam, gilliard, jroll: is this bug related? https://bugs.launchpad.net/bugs/1405448 | 15:52 |
| jroll | rloo: "This bug report is a duplicate of: Bug #1401501: rebuild with new image does not update kernel and ramdisk images." | 15:52 |
| jroll | I'd say yes ;) | 15:53 |
| rloo | jroll: ha ha. I didn't actually read it all. | 15:53 |
| NobodyCam | Nisha: looks like some thing is not set: that error is comming from https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L469 | 15:53 |
| NobodyCam | that should be set up on line 459 | 15:54 |
| *** Masahiro has quit IRC | 15:54 | |
| jroll | NobodyCam: now I'm wondering if this isn't sleeping long enough or something https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L455 | 15:55 |
| jroll | either that or ovs-vsctl changed the output format | 15:55 |
| jroll | gilliard: at any rate, are you willing to give my patch a test? | 15:56 |
| jroll | https://review.openstack.org/145254 | 15:56 |
| openstackgerrit | Merged openstack/ironic-python-agent: Skip failing execute tests when using /tmp noexec https://review.openstack.org/144294 | 15:57 |
| NobodyCam | brb | 15:57 |
| gilliard | jroll: Oh certainly. You mean https://review.openstack.org/145254 | 15:58 |
| jroll | yes | 15:58 |
| gilliard | will do | 15:58 |
| jroll | without going too deep, I *think* that should fix it | 15:58 |
| jroll | I can also test it later on sometime | 15:58 |
| gilliard | My patch should be a partial-fix as well. | 16:00 |
| gilliard | I don't think they conflict though. I'm not 100% sure that they're duplicates really. | 16:00 |
| jroll | the bugs or the patches? | 16:01 |
| gilliard | I don't think the 2 patches conflict, and I don't think the bugs are duplicates. | 16:02 |
| gilliard | 2 different problems, we have to fix both before nove rebuild will work. | 16:02 |
| gilliard | my patch is https://review.openstack.org/142782 | 16:03 |
| gilliard | BTW | 16:03 |
| jroll | they look like the same bug | 16:03 |
| jroll | right, I see it | 16:03 |
| gilliard | Firstly, we didn't check if the kernel or ramdisj had changed for an image. Secondly, we didn't update the cache, even if we did check. | 16:04 |
| gilliard | 2 sides of the same bug :) | 16:04 |
| jroll | ohhhhhhhh | 16:04 |
| jroll | I see :) | 16:05 |
| gilliard | was I talking nonsense? I am having coffee-flow problems. | 16:05 |
| jroll | same :P | 16:05 |
| jroll | gilliard: left a quick note there | 16:07 |
| jroll | didn't do a full review, though | 16:07 |
| NobodyCam | back | 16:09 |
| Nisha | NobodyCam, i re-checked everything. IMO, i have set everything | 16:10 |
| Nisha | i have tried it 6-7 times already and it is failing at same point | 16:10 |
| gilliard | jroll: saw, thanks | 16:10 |
| jroll | Nisha: it must be something with your environment as the gate is still working and uses the same code :) | 16:11 |
| NobodyCam | wow lag | 16:11 |
| jroll | Nisha: I'd recommend just grabbing a fresh install of ubuntu and starting again | 16:11 |
| NobodyCam | Nisha: can you run "sudo ovs-vsctl show |grep ${tapdev} -A1 -m1 | grep tag | cut -d':' -f2 | cut -b2-" from th command line? | 16:11 |
| rameshg87 | gilliard, jroll, i have a question regarding https://review.openstack.org/#/c/142782/4 | 16:12 |
| Nisha | NobodyCam, that line is after the failing command line | 16:12 |
| jroll | NobodyCam: | 16:12 |
| jroll | 15:50:22 Nisha | 2015-01-06 14:09:46.265 | +++ sudo ip netns exec qdhcp-6a65f545-1d41-4dab-8a12-1702965c0d72 ip link list | 16:12 |
| jroll | 15:50:22 Nisha | 2015-01-06 14:09:46.297 | Cannot open network namespace "qdhcp-6a65f545-1d41-4dab-8a12-1702965c0d72": No such file or directory | 16:12 |
| rameshg87 | gilliard, jroll, why don't we just clear off instance_info['kernel'] and instance_info['ramdisk'] before doing a rebuild ? | 16:12 |
| rameshg87 | gilliard, jroll, looks to me that should solve the problem | 16:12 |
| Nisha | the command line failing needs to be passed to make next line work | 16:13 |
| NobodyCam | ahh | 16:13 |
| jroll | rameshg87: that's a great point | 16:13 |
| *** mikedillion has joined #openstack-ironic | 16:13 | |
| gilliard | rameshg87: how do we tell if we're doing a rebuild? | 16:13 |
| rameshg87 | jroll, and if we are lucky, we should still hit the image_cache even after clearing the instance_info | 16:14 |
| *** mikedillion has quit IRC | 16:14 | |
| jroll | gilliard: just always clear it when building | 16:14 |
| rameshg87 | gilliard, i guess here https://github.com/openstack/nova/blob/master/nova/virt/ironic/driver.py#L954 | 16:14 |
| jroll | gilliard: (if it's there) | 16:14 |
| jroll | if it's a new build, it won't be there and thus won't hurt | 16:14 |
| *** jistr has quit IRC | 16:15 | |
| *** jistr has joined #openstack-ironic | 16:16 | |
| Nisha | NobodyCam, at gate is the devstack updated> | 16:16 |
| *** jistr has quit IRC | 16:17 | |
| *** jistr has joined #openstack-ironic | 16:17 | |
| jroll | Nisha: yes | 16:17 |
| NobodyCam | Nisha: ya | 16:17 |
| *** erwan_taf has quit IRC | 16:18 | |
| Nisha | :( | 16:18 |
| jroll | Nisha: really, wipe the machine and run devstack with a fresh install of ubuntu | 16:19 |
| jroll | sometimes weird things get in there that break things | 16:19 |
| jroll | the gate runs everything on a fresh install partially for this reason | 16:19 |
| Nisha | jroll, ok | 16:19 |
| Nisha | trying one last time :( | 16:19 |
| jroll | Nisha: are you using RECLONE=true? | 16:20 |
| Nisha | yes i set it because it was not happening | 16:20 |
| jroll | ok | 16:20 |
| Nisha | jroll, should we use it always or sometimes? | 16:21 |
| Nisha | means set it to true always or occasionally | 16:21 |
| jroll | you should use it when you want to update dependencies like neutron | 16:21 |
| jroll | so, periodically should be fine | 16:21 |
| Nisha | k | 16:21 |
| vdrok | rameshg87, kernel and ramdisk in instance_info are used now here https://review.openstack.org/#/c/136741/ | 16:24 |
| vdrok | hi all | 16:24 |
| gilliard | hi vdrok | 16:24 |
| vdrok | hi gilliard ! | 16:24 |
| rameshg87 | vdrok, hi | 16:25 |
| vdrok | rameshg87, so if they're cleared on rebuild i'll have save them to some other place | 16:25 |
| jroll | vdrok: where do they come from there? the API? | 16:26 |
| rameshg87 | vdrok, hmm .. that's true .. | 16:26 |
| NobodyCam | morning vdrok :) | 16:26 |
| vdrok | jroll, hi, they should be specified in instance_info when using not glance image | 16:26 |
| jroll | right, ok | 16:26 |
| vdrok | morning NobodyCam :) | 16:26 |
| jroll | maybe they should have glance:// prepended if from glance | 16:26 |
| jroll | and then if they start with glance://, delete and re-fetch | 16:27 |
| gilliard | Wherever else they come from might have the same issue. | 16:27 |
| vdrok | yup, glance:// or no prefix if its just uuid | 16:27 |
| rameshg87 | vdrok, but how do you relate your use-case with rebuild ? | 16:27 |
| rameshg87 | vdrok, your use-case of non-glance image reference doesn't consider nova | 16:28 |
| jroll | rameshg87: on rebuild, the user would change instance_info then issue rebuild | 16:28 |
| vdrok | rameshg87, yes, as i understood its part of the effort to make ironic more standalone | 16:28 |
| rameshg87 | vdrok, i mean today you cannot give any thing other than glance image from nova | 16:28 |
| vdrok | rameshg87, yup | 16:29 |
| rameshg87 | jroll, yeah i agree | 16:29 |
| rameshg87 | jroll, vdrok, but my point was change nova rebuild to clear off instance_info['kernel | 16:29 |
| jroll | the point is to (help) not use nova :P | 16:29 |
| jroll | oh | 16:29 |
| jroll | ohhh. | 16:29 |
| vdrok | ah, right | 16:29 |
| rameshg87 | jroll, vdrok, it doesn't affect your use-case | 16:30 |
| vdrok | rameshg87, sorry, didn't get it :) | 16:30 |
| jroll | rameshg87: but you can rebuild without nova | 16:30 |
| jroll | but yeah that's a good start | 16:30 |
| vdrok | rameshg87, understand now | 16:30 |
| rameshg87 | jroll, but why would you rebuild without nova ? :) | 16:30 |
| jroll | rameshg87: why not, maybe somebody wrote a lightweight nova sort of thing | 16:30 |
| jroll | and uses the api for build/rebuild | 16:30 |
| rameshg87 | jroll, hmm okay | 16:31 |
| jroll | it is possible :) | 16:31 |
| rameshg87 | jroll, :) | 16:31 |
| rameshg87 | vdrok, and i did give a try of your code with ilo driver | 16:32 |
| rameshg87 | vdrok, i faced one issue which i have commented | 16:33 |
| rameshg87 | vdrok, fixing that works perfectly fine :) | 16:33 |
| vdrok | rameshg87, yup, saw your comments | 16:34 |
| rameshg87 | vdrok, and i had one question | 16:34 |
| vdrok | rameshg87, i'm on vacation now, will update next monday :) | 16:34 |
| vdrok | rameshg87, sure | 16:34 |
| rameshg87 | vdrok, sure :) have a happy vacation then .. | 16:34 |
| rameshg87 | vdrok, how do you see file://xxx working with multiple conductors ? | 16:34 |
| vdrok | rameshg87, thanks :) | 16:34 |
| vdrok | rameshg87, i thought about that | 16:35 |
| rameshg87 | vdrok, are we expecting that the same file be present on all the conductors ? | 16:35 |
| vdrok | rameshg87, but it seems to be out of ironic scope | 16:35 |
| vdrok | rameshg87, you may set up some share | 16:35 |
| vdrok | rameshg87, if you don't then it should be on every conductor | 16:35 |
| rameshg87 | vdrok, hmm okay .. | 16:36 |
| *** BertieFulton has left #openstack-ironic | 16:36 | |
| rameshg87 | jroll, JayF talked about hardware manager today in the meeting | 16:36 |
| rameshg87 | jroll, wanted to talk to you or JayF regarding that | 16:36 |
| NobodyCam | brb | 16:37 |
| jroll | rameshg87: go ahead, JayF isn't online yet but can read back | 16:37 |
| rameshg87 | jroll, how do we actually put another hardware manager into agent | 16:38 |
| rameshg87 | jroll, is there some mechanism available for that ? | 16:38 |
| *** dlaube has joined #openstack-ironic | 16:38 | |
| rameshg87 | jroll, looking at code i see we recognise all hardware manager using stevedore | 16:38 |
| jroll | rameshg87: use entry points | 16:39 |
| * jroll finds an example | 16:39 | |
| jroll | rameshg87: https://github.com/rackerlabs/onmetal-ironic-hardware-manager/blob/master/setup.cfg#L18-20 | 16:39 |
| jroll | rameshg87: and then just pip install that when building the agent | 16:39 |
| *** pcrews has joined #openstack-ironic | 16:40 | |
| rameshg87 | jroll, ah okay .. i get it now | 16:41 |
| Nisha | reviewers, https://review.openstack.org/#/c/103007/26 | 16:41 |
| jroll | rameshg87: you could even use that repo as a template for your manager | 16:41 |
| rameshg87 | jroll, i can build a new hardware manager and add an entry point in the same namespace | 16:41 |
| jroll | yep :) | 16:42 |
| jroll | rameshg87: and you could write your own dockerfile that inherits from the upstream dockerfile | 16:42 |
| jroll | and then just build that | 16:42 |
| rameshg87 | jroll, would be much more easy if i use dib to build :) | 16:43 |
| rameshg87 | jroll, just add one more element and build it with ironic-agent | 16:43 |
| jroll | rameshg87: I guess | 16:43 |
| jroll | that will also require at least 4gb of ram | 16:44 |
| rameshg87 | jroll, yeah 3gb i guess | 16:44 |
| gilliard | Thanks for the discussion about image caches. I'm out now but will put up new patchset tomorrow. | 16:44 |
| jroll | rameshg87: right... docker is straightforward too | 16:45 |
| jroll | rameshg87: here's a simplified version of our downstream dockerfile: https://gist.github.com/jimrollenhagen/5f0d0369e06acd5f3aa8 | 16:45 |
| rameshg87 | jroll, looks even simpler infact :) | 16:46 |
| jroll | the "FROM upstream" inherits from the dockerfile in tree | 16:46 |
| rameshg87 | jroll, okay | 16:47 |
| rameshg87 | jroll, but what if someone has a hetergenous environment | 16:47 |
| rameshg87 | jroll, so they will have one ramdisk built with all the hardware managers for his environment | 16:47 |
| jroll | rameshg87: it's designed to allow that | 16:48 |
| rameshg87 | jroll, but i see it looking for next hardware manager only if it raises NotImplementedError | 16:48 |
| rameshg87 | jroll, lines 391-399 https://review.openstack.org/#/c/143193/13/ironic_python_agent/hardware.py | 16:49 |
| jroll | rameshg87: right, per function | 16:49 |
| jroll | so let's take erase_device for example | 16:49 |
| jroll | when you call erase_device, it will go through each hw manager until one is found that does not raise NotImplementedError | 16:49 |
| jroll | so if you had a "fancy pci-e flash drive" hardware manager | 16:50 |
| jroll | it might check if the device to be erased is a fancy pci-e flash drive | 16:50 |
| jroll | if not, raise NotImplementedError | 16:50 |
| jroll | if so, erase it and return | 16:50 |
| rameshg87 | jroll, yeah i got your point | 16:50 |
| rameshg87 | jroll, but may be NotImplementedError doesn't seem to be right word :) | 16:50 |
| jroll | rameshg87: put that feedback on the review then :) | 16:51 |
| rameshg87 | jroll, sure :) | 16:51 |
| jroll | perhaps another exception we can catch as well | 16:51 |
| rameshg87 | jroll, yeah | 16:51 |
| *** jerryz has quit IRC | 16:51 | |
| rameshg87 | jroll, so if i had to add a raid configuration method there | 16:51 |
| rameshg87 | jroll, i would add it to HardwareManager class | 16:52 |
| jroll | rameshg87: it would make sense for that hw manager to raise NotImplementedError for something like 'get_cpus' | 16:52 |
| jroll | right | 16:52 |
| rameshg87 | jroll, have an abstract metho raise NotImplementedError | 16:52 |
| rameshg87 | jroll, and add my own hardware manager which implements it and then bundle it along and create a new agent ramdisk | 16:52 |
| rameshg87 | jroll, correct ? | 16:53 |
| jroll | yep :) | 16:53 |
| rameshg87 | jroll, great :) | 16:53 |
| rameshg87 | jroll, that serves my use-case then | 16:53 |
| jroll | great :) | 16:53 |
| rameshg87 | jroll, thanks i will go through the code tomorrow and check .. | 16:53 |
| jroll | rameshg87: cool, good luck :) | 16:54 |
| * rameshg87 thinks it is time to go to sleep | 16:54 | |
| rameshg87 | goodnight ironic :) | 16:54 |
| *** rameshg87 has quit IRC | 16:54 | |
| jroll | night | 16:54 |
| NobodyCam | brb | 16:55 |
| *** lucas-hungry is now known as lucasagomes | 16:59 | |
| JayF | jroll: thanks for fielding that ? for me :) | 17:00 |
| jroll | np | 17:00 |
| *** ifarkas has quit IRC | 17:04 | |
| *** eghobo has joined #openstack-ironic | 17:06 | |
| *** vdrok is now known as vdrok_afk | 17:09 | |
| *** mikedillion has joined #openstack-ironic | 17:13 | |
| *** pensu has joined #openstack-ironic | 17:15 | |
| NobodyCam | morning JayF :) | 17:15 |
| JayF | morning | 17:15 |
| *** jcoufal_ has quit IRC | 17:18 | |
| devananda | morning, all | 17:20 |
| jroll | hiya devananda | 17:21 |
| *** jistr has quit IRC | 17:22 | |
| NobodyCam | morning devananda | 17:23 |
| lucasagomes | devananda, jroll, JayF morning | 17:26 |
| JayF | morning | 17:26 |
| jroll | \o lucas | 17:26 |
| *** mikedillion has quit IRC | 17:29 | |
| NobodyCam | devananda: have you had a chance to look over rloo's comment(s) on https://review.openstack.org/#/c/139217? | 17:30 |
| *** MattMan has left #openstack-ironic | 17:38 | |
| *** Masahiro has joined #openstack-ironic | 17:38 | |
| *** Masahiro has quit IRC | 17:43 | |
| *** Nisha has quit IRC | 17:46 | |
| *** athomas has quit IRC | 17:51 | |
| *** ijw has joined #openstack-ironic | 17:53 | |
| * NobodyCam makes a fresh pot of coffee | 17:57 | |
| *** penick has joined #openstack-ironic | 17:58 | |
| *** eghobo has quit IRC | 18:01 | |
| *** derekh has quit IRC | 18:04 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic-specs: Check that filename is same as blueprint name https://review.openstack.org/145088 | 18:09 |
| openstackgerrit | Alex Weeks proposed openstack/ironic-specs: Add pluggable metrics backend for Ironic and IPA https://review.openstack.org/137171 | 18:12 |
| aweeks | NobodyCam: ^ I think I addressed your comments | 18:13 |
| aweeks | And, if any non-Rackspace people could take a look at https://review.openstack.org/#/c/137171/ it would be very much appreciated | 18:13 |
| NobodyCam | aweeks: :) will take a look in a few | 18:14 |
| aweeks | thanks | 18:14 |
| *** ijw has quit IRC | 18:23 | |
| NobodyCam | aweeks: have you already started on the shared metrics reporting library? | 18:24 |
| aweeks | NobodyCam: yes, currently its here: https://github.com/rackerlabs/metricslogger/pull/1/files | 18:25 |
| JayF | NobodyCam: It's running in our production environment, although not split out fully yet :) | 18:25 |
| aweeks | I'm hoping to push another big update to the PR today | 18:25 |
| JayF | oh wow you got further splitting that out than I realized | 18:25 |
| JayF | nice | 18:26 |
| aweeks | JayF: yeah, refactoring the config functions to be more sane, and the tests | 18:26 |
| NobodyCam | gah whats up with the lag today :( | 18:26 |
| *** jistr has joined #openstack-ironic | 18:30 | |
| *** dtantsur is now known as dtantsur|afk | 18:31 | |
| *** jistr has quit IRC | 18:31 | |
| *** jistr has joined #openstack-ironic | 18:32 | |
| mjturek | hey does anyone know a good guide for provisioning physical hardware using devstack with ironic enabled? Been stuck on getting neutrondhcp/tftp working with my external node | 18:37 |
| mjturek | neutron's * | 18:37 |
| Shrews | mjturek: there was a devstack patch up to make that easier, but I don't have a link handy. Dealing with a dead laptop right now... :( | 18:40 |
| mjturek | ahhhh, thanks Shrews. I'll search launchpad a bit! | 18:41 |
| *** pelix has quit IRC | 18:47 | |
| *** killer_prince has quit IRC | 18:47 | |
| *** killer_prince has joined #openstack-ironic | 18:47 | |
| *** killer_prince is now known as lazy_prince | 18:48 | |
| Shrews | mjturek: i can't find it, but this recent review seems to indicate that is already merged: https://review.openstack.org/141966 That's all I know about it, though. | 18:48 |
| mjturek | Shrews, thanks! I'll see what I can get out of it :) | 18:49 |
| devananda | mjturek: sdake was in here asking the same thing yesterday. and several other folks were discussing it before the break | 18:49 |
| devananda | it's definitely an area where our docs need to be improved | 18:49 |
| *** datajerk has quit IRC | 18:50 | |
| mjturek | devananda gotcha, maybe digging through the logs would be useful then | 18:50 |
| NobodyCam | mjturek: if your going through the process now. I would love to review a doc update patch :) | 18:51 |
| NobodyCam | <shameless pulg> | 18:51 |
| mjturek | NobodyCam, I would be more than happy to help out with documenting the process :) | 18:51 |
| mjturek | just trying to get through it successfully first :-P | 18:51 |
| NobodyCam | ofc :) | 18:52 |
| devananda | perhaps an etherpad where everyone working on that shares notes -- as a starting point? | 18:52 |
| mjturek | devananda, sounds like a great idea | 18:52 |
| NobodyCam | wehave https://etherpad.openstack.org/p/IronicDocumentationTasks | 18:53 |
| mjturek | how about a section there then? | 18:53 |
| NobodyCam | that would work for /me | 18:53 |
| devananda | how about https://etherpad.openstack.org/p/ironic-devstack-real-hardware | 18:54 |
| NobodyCam | ahh another ether pad to keep track of ...lol | 18:55 |
| devananda | hah | 18:55 |
| devananda | never mind - doc is probably better :) | 18:55 |
| NobodyCam | heheheh | 18:55 |
| *** ijw has joined #openstack-ironic | 18:55 | |
| mjturek | devananda, NobodyCam, okay cool :) I'll drop notes there as I go | 18:55 |
| NobodyCam | mjturek: awesome Thank you :) | 18:56 |
| mjturek | NobodyCam, no problem! always looking to contribute | 18:56 |
| NobodyCam | (and i am okay with another etherpad) | 18:56 |
| *** ijw_ has joined #openstack-ironic | 18:58 | |
| *** datajerk has joined #openstack-ironic | 18:58 | |
| *** ijw has quit IRC | 19:01 | |
| *** penick has quit IRC | 19:02 | |
| NobodyCam | brb quick walkies time | 19:03 |
| *** ijw_ has quit IRC | 19:04 | |
| *** penick has joined #openstack-ironic | 19:06 | |
| *** eghobo has joined #openstack-ironic | 19:08 | |
| *** erwan_taf has joined #openstack-ironic | 19:09 | |
| *** dkingshott has joined #openstack-ironic | 19:11 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Allow use of multiple simultaneous HW managers https://review.openstack.org/143193 | 19:11 |
| *** vdrok_afk has quit IRC | 19:11 | |
| JayF | I updated https://review.openstack.org/#/c/143193/ in response to code review comments; would love additional eyes | 19:22 |
| *** eghobo has quit IRC | 19:23 | |
| *** ryanpetrello_ has joined #openstack-ironic | 19:25 | |
| *** Masahiro has joined #openstack-ironic | 19:27 | |
| *** ryanpetrello has quit IRC | 19:27 | |
| *** ryanpetrello_ is now known as ryanpetrello | 19:27 | |
| NobodyCam | ahh TY JayF you get the recheck on rloo's patch :) | 19:29 |
| * rloo looks up from reviewing. huh? ;) | 19:30 | |
| *** eghobo has joined #openstack-ironic | 19:30 | |
| *** Masahiro has quit IRC | 19:31 | |
| NobodyCam | lol your title check patch needed to be rechecked | 19:32 |
| rloo | NobodyCam: just noticed. is that error / time out happening a lot? | 19:32 |
| NobodyCam | I haven't seen a it... but I maybe behind the curve | 19:33 |
| rloo | NobodyCam: ok, will try to keep an eye out. I think I may have seen it happen last night. | 19:33 |
| lucasagomes | jroll, fyi https://review.openstack.org/145235 | 19:36 |
| lucasagomes | tested locally | 19:36 |
| NobodyCam | oh neat-oh | 19:36 |
| JayF | rloo: Might I request a world-famous Ruby-review on https://review.openstack.org/#/c/143193/ | 19:38 |
| JayF | :P | 19:38 |
| rloo | NobodyCam: that is Bug 1326813 - for the timeout thing | 19:38 |
| rloo | JayF: not sure. let me see... | 19:38 |
| *** david-lyle_afk is now known as david-lyle | 19:39 | |
| rloo | JayF: hmm. I haven't yet (I don't think) reviewed any agent code. Should I start now? | 19:39 |
| JayF | That would probably be a good place to start. | 19:42 |
| NobodyCam | JayF: you happen to have a free minute to take a look at https://review.openstack.org/#/c/139217 | 19:42 |
| NobodyCam | lucasagomes: ^^^^^^ also | 19:42 |
| JayF | It touches a little bit of everything | 19:42 |
| JayF | opening it in a tab NobodyCam | 19:42 |
| lucasagomes | ack | 19:43 |
| JayF | itsatrap! State machine code! Pull the ripcord! | 19:43 |
| JayF | :P | 19:43 |
| *** mriedem has joined #openstack-ironic | 19:43 | |
| NobodyCam | lol | 19:43 |
| devananda | adam_g: recent grenade-sideways failure, which seems to be the result of a kernel panic while starting the deploy ramdisk -- http://logs.openstack.org/60/144760/1/gate/gate-grenade-dsvm-ironic-sideways/5d08c02/logs/ironic-bm-logs/baremetalbrbm_2_console.txt.gz | 19:43 |
| mriedem | devananda: i'm unfamiliar with http://logs.openstack.org/60/144760/1/gate/gate-grenade-dsvm-ironic-sideways/5d08c02/logs/ironic-bm-logs/baremetalbrbm_2_console.txt.gz - can you point out what specifically you're looking at | 19:44 |
| devananda | mriedem: this is the kernel log of a VM used by devstack & ironic to "mock" bare metal behavior within the gate env | 19:44 |
| devananda | mriedem: scroll to the bottom - the problem should then be obvious | 19:44 |
| adam_g | yea, finding something to fingerprint for that failure is tough | 19:44 |
| mriedem | [ 422.148000] Kernel panic - not syncing: Fatal exception in interrupt | 19:44 |
| adam_g | those console logs aren't in logstash | 19:44 |
| devananda | adam_g: gah :( | 19:45 |
| mriedem | is there any way to get this back into the ironic logs as an error? | 19:45 |
| devananda | there's nothing in the screen logs to fingerprint -- because there's no error there, aside from a timeout | 19:45 |
| mriedem | right | 19:45 |
| devananda | mriedem: no, because it's outside of what Ironic can see | 19:45 |
| devananda | ironic knows the node is powered on -- it is monitoring that | 19:45 |
| devananda | and all the PXE env is correctly prepared -- we could watch for that, but it would pass, since the node has already PXE booted | 19:46 |
| devananda | ironic is left waiting for the OS and init scripts within the booted ramdisk to *do* something -- namely, call back to ir-api | 19:46 |
| devananda | we could adjust the timeout on this | 19:46 |
| devananda | so it fails faster | 19:46 |
| mriedem | what's the timeout today? | 19:46 |
| devananda | or add these kernel logs to logstash | 19:46 |
| JayF | a console log type functionality could possibly see this | 19:47 |
| devananda | 30 minutes | 19:47 |
| mriedem | i'm not sure how easy it is to add the kernel logs to logstash, would be a question for clarkb probably, i know it's easier if they have oslo log formatting | 19:47 |
| JayF | but not likely automatically detect it | 19:47 |
| *** hypnosb has quit IRC | 19:47 | |
| devananda | 537 # Timeout (seconds) for waiting callback from deploy ramdisk. | 19:47 |
| devananda | 538 # 0 - unlimited. (integer value) | 19:47 |
| devananda | 539 #deploy_callback_timeout=1800 | 19:47 |
| mriedem | but syslog is indexed so i don't see why this couldn't be | 19:47 |
| *** hypnosb has joined #openstack-ironic | 19:48 | |
| devananda | mriedem: it's not the devstack host's syslog or kernel log, though | 19:48 |
| mriedem | devananda: ok, the tempest build timeout is 5 minutes :) | 19:48 |
| mriedem | devananda: oh these are remote systems? | 19:48 |
| devananda | heh. so easy change there -- tune down the timeout for devstack-gate | 19:48 |
| devananda | mriedem: no. they're special VMs created by devstack and enrolled with Ironic | 19:49 |
| mriedem | ok | 19:49 |
| devananda | they behave like remote systems | 19:49 |
| mriedem | was figuring that since the logs are published we should be able to index them | 19:49 |
| NobodyCam | brb | 19:49 |
| devananda | for all intents and purposes, once tempest starts exercising Ironic and Nova, these are remote systems | 19:49 |
| adam_g | there is/was another issue with nodes provisioning, i thought we had a fingerprint to catch nodes failing to reach an ACTIVE state | 19:50 |
| mriedem | i think the node reaches active, it's the instance deploy that fails if i'm understanding the logs | 19:51 |
| mriedem | or is instance == node in ironic speak? | 19:51 |
| adam_g | let me look | 19:51 |
| devananda | mriedem: instance == what's on the node. node == the hardware | 19:51 |
| mriedem | k, that's what i thought | 19:52 |
| devananda | mriedem: the node reached POWER_ON. but it was still waiting-for-callback | 19:52 |
| mriedem | so not active | 19:52 |
| devananda | which means the deploy ramdisk did not POST back to ir-api | 19:52 |
| devananda | it hadn't even copied the instance image to the node yet | 19:52 |
| mriedem | b/c of said kernel panic | 19:52 |
| devananda | quite far from active | 19:52 |
| adam_g | ah | 19:52 |
| adam_g | https://git.openstack.org/cgit/openstack-infra/elastic-recheck/tree/queries/1393099.yaml | 19:52 |
| adam_g | we're failing in a nova test, so we're not tracking ironic state at all (on the client/tempest side at least) | 19:53 |
| devananda | that would catch it | 19:53 |
| adam_g | no | 19:53 |
| adam_g | Details: (ServerAddressesTestJSON:setUpClass) Server 4e4b8db2-9b2c-470b-a1d9-028696199413 failed to reach ACTIVE status and task state "None" within the required time (300 s). Current status: BUILD. Current task state: spawning. | 19:53 |
| devananda | ooh. there's no "Timed out waiting for node" in this log | 19:54 |
| adam_g | that existing query hits against scenario tests, which do some tracking of ironic states | 19:54 |
| mriedem | right, nothing really good to track on | 19:54 |
| devananda | right - tempest timeout hit first | 19:54 |
| mriedem | fwiw this is the bug i reported https://bugs.launchpad.net/ironic/+bug/1408067 | 19:55 |
| devananda | adam_g: what do you think of lowering deploy_callback_timeout within Ironic for this (and possibly many other) jobs? | 19:55 |
| devananda | possibly all of them? | 19:55 |
| devananda | 30 minutes is way longer than it should take, even on most hardware | 19:55 |
| JayF | +1 | 19:56 |
| adam_g | devananda, to fail faster? yeah, sounds doable | 19:56 |
| mriedem | devananda: adam_g: we could just set deploy_callback_timeout lower in devstack-gate for tempest runs | 19:56 |
| mriedem | but if 30 min isn't a sane default, i guess you can change it in the code | 19:57 |
| openstackgerrit | Merged openstack/ironic-specs: Check that filename is same as blueprint name https://review.openstack.org/145088 | 19:57 |
| devananda | I suspect a production-sane default here is still going to be longer than what we want in d-g | 19:57 |
| *** ijw has joined #openstack-ironic | 19:57 | |
| mriedem | i'd think so too | 19:57 |
| devananda | eg, 10 min is reasonable in production, but not in d-g | 19:57 |
| devananda | because some hardware POSTs slowly | 19:58 |
| NobodyCam | i was just about to say 10 min | 19:58 |
| adam_g | seems the callback timeout in devstack/tempest should be lower than the nova boot timeout | 19:58 |
| devananda | IIRC, we picked 30 as the default to be on the safe side | 19:58 |
| adam_g | so we get an ironic failure before the nova | 19:58 |
| devananda | adam_g: ++ | 19:58 |
| mriedem | so < 300, sure | 19:58 |
| JayF | devananda: I'd be a bigger fan of that timeout being more like 15m than 10m by default :) | 19:58 |
| mriedem | adam_g: devananda: could be like 4 min in d-g | 19:58 |
| JayF | devananda: although it'll obviously vary wildly by size of image written and such | 19:58 |
| adam_g | hmm | 19:59 |
| devananda | JayF: no. this is tiemout waiting for the first call back | 19:59 |
| devananda | JayF: not end of provisioning | 19:59 |
| adam_g | theres no active_timeout or similar, eh? | 19:59 |
| adam_g | i guess there can't be | 20:00 |
| devananda | it should be POST + deploy ramdisk transfer and boot time | 20:00 |
| JayF | oh wow | 20:00 |
| JayF | okay | 20:00 |
| devananda | adam_g: we dont, but should, have timeouts for each state transition | 20:00 |
| devananda | JayF: just double checked -- this timeout only applies to DEPLOYWAIT today | 20:01 |
| adam_g | devananda, but as JayF mentioned, that timeout depends largely on the size of images being booted | 20:01 |
| devananda | shouldn't | 20:02 |
| devananda | hardware POST time >> time to send a 1GB file over http | 20:02 |
| *** ijw has quit IRC | 20:03 | |
| devananda | also, IPA's image is not 1GB | 20:03 |
| devananda | and the iscsi/pxe image is even smaller | 20:03 |
| devananda | I need to run. bbi30 | 20:04 |
| adam_g | sorry, i mean a timeout on reaching ACTIVE | 20:04 |
| *** ijw has joined #openstack-ironic | 20:04 | |
| NobodyCam | the longest post i can recall traacking was like 7 or 8 minutes | 20:04 |
| NobodyCam | adam_g: thats nova's time out or ironics? | 20:04 |
| adam_g | NobodyCam, i dont know if nova has a timeout, but tempest has an active timeout it uses when booting instance. | 20:06 |
| adam_g | NobodyCam, but my concern is about having timeouts for ironic state transitions, in addition to exisitng callback timeout | 20:06 |
| *** dprince has quit IRC | 20:09 | |
| NobodyCam | humm | 20:09 |
| NobodyCam | how would that impace the long running agent folks | 20:10 |
| NobodyCam | impact even | 20:10 |
| devananda | NobodyCam: shouldnt | 20:13 |
| devananda | thats not a transition | 20:14 |
| NobodyCam | but wouldn't we see it as a node in deploy wait? | 20:14 |
| devananda | today? I dunno | 20:14 |
| devananda | with the new FSM? nope | 20:14 |
| devananda | there's soooo much stuff that is waiting for that :( | 20:15 |
| NobodyCam | :( | 20:15 |
| * devananda really goes AFK now | 20:15 | |
| * NobodyCam alsosteps afk | 20:18 | |
| rloo | JayF: there are a few specs that I want to look at. I think/try to look at 143193 by the end of Thurs. | 20:20 |
| *** erwan_taf has quit IRC | 20:21 | |
| adam_g | mriedem, https://review.openstack.org/#/q/status:open+branch:master+topic:callback_timeout,n,z | 20:24 |
| *** jistr has quit IRC | 20:25 | |
| mriedem | adam_g: thanks | 20:26 |
| mriedem | commented | 20:26 |
| openstackgerrit | Victor Howard proposed openstack/ironic: display error logging should be improved https://review.openstack.org/144771 | 20:27 |
| * NobodyCam is back | 20:30 | |
| JayF | rloo: it's not a spec; it's actual code, and tbh I'd love to have it landed before then :) | 20:32 |
| *** pensu has quit IRC | 20:32 | |
| rloo | JayF: yeah, I know it's code. Just that I'm behind on reviewing the specs. So this is a high priority thing for you? | 20:32 |
| JayF | rloo: it's blocking something I'd like to be able to do downstream (tear apart our big hardware manager into smaller pieces that possibly others could consume) | 20:33 |
| rloo | JayF: hmm. I'll try to get to it tomorrow then. Since it isn't code I'm familiar with, I need a bigger chunk of time to get up to speed/review. | 20:34 |
| *** sambetts has quit IRC | 20:36 | |
| adam_g | JayF, re IPA+devstack, is the time between a 'nova boot' and ironic powering the machine on typically <60 sec? with a cold cache | 20:36 |
| JayF | adam_g: I don't have a good feel for that; tbh | 20:37 |
| adam_g | der, i suppose i could just go look at gate logs | 20:37 |
| JayF | adam_g: I'd suspect so, though. It's only ~200mb transferred locally | 20:37 |
| *** ijw has quit IRC | 20:37 | |
| *** sambetts has joined #openstack-ironic | 20:37 | |
| adam_g | yeah, looks about 30 sec | 20:39 |
| JayF | This is a change with 2x +2 from Rackers on IPA, but we wanted an outside opinion before landing it: https://review.openstack.org/#/c/141957/ -- basically adds a really sweet standalone mode to IPA, whereby you can spin up an agent without an ironic to issue it commands for functional testing | 20:41 |
| *** andreykurilin has joined #openstack-ironic | 20:43 | |
| *** pcaruana is now known as pcaruana|afk| | 20:45 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Improve IPA client library https://review.openstack.org/111118 | 20:46 |
| jroll | 19:32:33 rloo | NobodyCam: just noticed. is that error / time out happening a lot? <- hp cloud is having networking issues, fyi | 21:01 |
| rloo | jroll: thx. there's a bug for it. 1326813. Guess it happens now and then :-( | 21:02 |
| jroll | yeah, networks are hard | 21:03 |
| JayF | ++ | 21:03 |
| *** ijw has joined #openstack-ironic | 21:08 | |
| *** ijw has quit IRC | 21:14 | |
| NobodyCam | doh | 21:15 |
| *** Masahiro has joined #openstack-ironic | 21:16 | |
| *** Masahiro has quit IRC | 21:20 | |
| *** Marga_ has joined #openstack-ironic | 21:21 | |
| *** russellb has joined #openstack-ironic | 21:27 | |
| *** eghobo has quit IRC | 21:41 | |
| * NobodyCam may need to run out for a bit.. | 21:43 | |
| *** russellb has left #openstack-ironic | 21:53 | |
| *** ryanpetrello_ has joined #openstack-ironic | 21:55 | |
| *** ryanpetrello has quit IRC | 21:56 | |
| *** ryanpetrello_ is now known as ryanpetrello | 21:56 | |
| *** penick has quit IRC | 21:56 | |
| *** eghobo has joined #openstack-ironic | 21:57 | |
| *** Marga_ has quit IRC | 21:57 | |
| *** Marga_ has joined #openstack-ironic | 21:57 | |
| Shrews | rloo: i gave https://review.openstack.org/139217 a second +2, but didn't approve. i want to give you time to make sure you're comfortable with it | 21:57 |
| rloo | Shrews: let me look. it was awhile ago. I think it was due to changing the logic a bit though? | 21:58 |
| *** Marga_ has quit IRC | 21:59 | |
| *** Marga_ has joined #openstack-ironic | 21:59 | |
| rloo | Shrews: wrt 139217, am I wrong about the > vs >= at line 1479? https://review.openstack.org/#/c/139217/9/ironic/conductor/manager.py | 22:08 |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Hide oslo.messaging DEBUG logs by default https://review.openstack.org/145361 | 22:17 |
| *** jjohnson2 has quit IRC | 22:18 | |
| JayF | devananda: ^ related: https://www.youtube.com/watch?v=9xLW-1c836k | 22:18 |
| *** penick has joined #openstack-ironic | 22:18 | |
| devananda | right | 22:19 |
| NobodyCam | devananda: nice catch | 22:21 |
| Shrews | rloo: sorry, went to the store real quick... | 22:29 |
| Shrews | rloo: yes, i believe devananda is correct here | 22:29 |
| rloo | Shrews: no worries | 22:29 |
| rloo | Shrews: ok, I'm not going to go through it to convince myself. | 22:30 |
| Shrews | rloo: the heart of it is, count is incremented before the power state check, but not checked until AFTER that check | 22:33 |
| Shrews | so it will get one more power state check before it goes over the config value | 22:33 |
| Shrews | but, yeah, a bit terse code there | 22:34 |
| Shrews | devananda: i *thought* there was a fix to oslo.messaging for that | 22:35 |
| Shrews | devananda: i'd rather just bump the version there, if so | 22:35 |
| * Shrews looks | 22:36 | |
| JayF | Shrews: we shouldn't bump our version independent of global reqs, right? | 22:36 |
| Shrews | JayF: yeah. i would expect someone to bump it in global reqs though | 22:37 |
| adam_g | Shrews, fix committed, but not released yet https://bugs.launchpad.net/designate/+bug/1399257 | 22:37 |
| JayF | for https://review.openstack.org/#/c/137926/ do we care that the "H" in github is not capitalized? | 22:37 |
| JayF | I'd like to land it. | 22:37 |
| Shrews | adam_g: thx | 22:37 |
| Shrews | i guess it won't hurt to go ahead and land it if the new oslo package isn't out | 22:39 |
| Shrews | devananda: that linked bug # is, erm, off? | 22:40 |
| rloo | JayF: wrt 137926, if that's the only reason? I don't care about the H. It can be updated afterwards. | 22:42 |
| Shrews | that seems like a silly reason to hold it back | 22:43 |
| rloo | Shrews: unless we could get sued or something? but even so, fix it later. | 22:44 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/142334 | 22:45 |
| NobodyCam | lol | 22:48 |
| rloo | Shrews: sorry, to be clear about 139217 wrt the counts. "devananda is correct" that there is a bug there, so this patch fixes that bug, right? And we're fine if it isn't mentioned that it was fixed in the commit? | 22:54 |
| NobodyCam | rloo: i'm okay with it just because I don't want to rebase the entire patch set :-p | 22:56 |
| *** ryanpetrello has quit IRC | 22:56 | |
| NobodyCam | wounder if just editing the commit message thru the web interface would work? | 22:56 |
| rloo | NobodyCam: so if the behaviour changes for someone because of this, and it isn't mentioned in the commit, and no bug is opened for it, is it OK? Sorry, just being devil's advocate here. | 22:57 |
| Shrews | rloo: you raise a valid point about it not being in the commit msg. but we can edit that thru the iface | 22:58 |
| Shrews | NobodyCam: are you trying that? | 22:58 |
| NobodyCam | lol was just about to... | 22:58 |
| rloo | Shrews, NobodyCam: thx :-) | 22:59 |
| NobodyCam | but have not yet | 22:59 |
| Shrews | if we can help deva push this along, we should | 22:59 |
| NobodyCam | yep | 22:59 |
| NobodyCam | this may require a full rebase but... | 22:59 |
| rloo | I still have a big problem with the code anyway; it doesn't handle the target-state properly or something. | 22:59 |
| rloo | I added a patch to the end of the series, with updated unit tests to show how it fails when it shouldn't fail. | 23:00 |
| Shrews | rloo: oh? | 23:00 |
| Shrews | ah, see it now | 23:01 |
| *** mriedem has left #openstack-ironic | 23:01 | |
| NobodyCam | rloo: - corrects minor issue with retry loop .. work? | 23:01 |
| rloo | NobodyCam: no. sec. | 23:02 |
| openstackgerrit | Merged openstack/ironic-specs: Add support for VirtualBox WebService. https://review.openstack.org/137926 | 23:02 |
| rloo | NobodyCam: devananda sez: 'the "max_retries" option is acting as "max_tries" -- in other words, set the option to 3, and it will attempt to sync the power state 3 times, not 4.' | 23:02 |
| rloo | NobodyCam: that's what he fixed. | 23:03 |
| Shrews | i.e., it retried one more time than it was supposed to | 23:03 |
| Shrews | rloo: i'll look at your fsm patch tomorrow. thx for pointing that out | 23:04 |
| rloo | devananda also fixed something else, but I am fine not mentioning it because you're going to ask me to word it for you and I'm too tired ;) | 23:04 |
| *** Masahiro has joined #openstack-ironic | 23:04 | |
| openstackgerrit | Jim Mankovich proposed openstack/ironic-specs: Support per Driver Sensor Meters https://review.openstack.org/130074 | 23:05 |
| NobodyCam | rloo: - corrects minor issue with max_retries running an extra attemp .. better? | 23:05 |
| NobodyCam | or - corrects issue with max_retries running an extra attemp .. better? | 23:05 |
| NobodyCam | or - corrects issue with max_retries running an extra attempt .. :-p | 23:06 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Refactor async helper methods in conductor/manager.py https://review.openstack.org/139217 | 23:07 |
| rloo | NobodyCam: yeah | 23:07 |
| NobodyCam | and now the rebasing starts | 23:07 |
| NobodyCam | lol I should have done it locally | 23:08 |
| *** Masahiro has quit IRC | 23:09 | |
| rloo | dinner time. ciao. | 23:10 |
| NobodyCam | night rloo | 23:10 |
| NobodyCam | :) | 23:10 |
| *** rloo is now known as rloo_afk | 23:11 | |
| *** alexpilotti has quit IRC | 23:12 | |
| *** eghobo has quit IRC | 23:13 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Enable async callbacks from task.process_event() https://review.openstack.org/140868 | 23:14 |
| openstackgerrit | Merged openstack/ironic-python-agent: Add standalone mode for IPA https://review.openstack.org/141957 | 23:17 |
| *** jgrimm is now known as zz_jgrimm | 23:18 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Convert check_deploy_timeout to use process_event https://review.openstack.org/140869 | 23:18 |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Convert drivers to use process_event() https://review.openstack.org/140883 | 23:20 |
| NobodyCam | rloo_afk: should I rebase 142303 | 23:20 |
| *** ijw has joined #openstack-ironic | 23:20 | |
| *** hypnosb has quit IRC | 23:20 | |
| *** ChuckC has quit IRC | 23:20 | |
| *** aweeks has quit IRC | 23:20 | |
| *** zer0c00l has quit IRC | 23:21 | |
| *** yuriyz has quit IRC | 23:21 | |
| *** lynxman has quit IRC | 23:21 | |
| *** NobodyCam has quit IRC | 23:21 | |
| *** soren has quit IRC | 23:21 | |
| *** TheJulia has quit IRC | 23:21 | |
| *** med_ has quit IRC | 23:21 | |
| *** aignatov has quit IRC | 23:21 | |
| *** kylestev has quit IRC | 23:21 | |
| *** ekarlso- has quit IRC | 23:21 | |
| *** marios has quit IRC | 23:21 | |
| *** TheJulia has joined #openstack-ironic | 23:21 | |
| *** NobodyCam has joined #openstack-ironic | 23:21 | |
| *** lynxman has joined #openstack-ironic | 23:21 | |
| *** yuriyz has joined #openstack-ironic | 23:21 | |
| *** aweeks has joined #openstack-ironic | 23:21 | |
| NobodyCam | hello? | 23:21 |
| NobodyCam | wow that was strange | 23:21 |
| *** zer0c00l has joined #openstack-ironic | 23:22 | |
| jroll | yay netsplits | 23:22 |
| *** ChuckC has joined #openstack-ironic | 23:22 | |
| *** med_ has joined #openstack-ironic | 23:22 | |
| *** aignatov has joined #openstack-ironic | 23:22 | |
| *** kylestev has joined #openstack-ironic | 23:22 | |
| *** ekarlso- has joined #openstack-ironic | 23:22 | |
| *** marios has joined #openstack-ironic | 23:22 | |
| NobodyCam | Irssi: Join to #openstack-ironic was synced in 64 secs | 23:23 |
| NobodyCam | wow supper slow | 23:23 |
| JayF | Net is splitting | 23:24 |
| NobodyCam | lol | 23:24 |
| NobodyCam | i'm not going to rebase the rloo last patch in the FSM set as the first line of the commit message is NOT TO BE MERGED | 23:24 |
| *** soren has joined #openstack-ironic | 23:26 | |
| *** zz_jgrimm has quit IRC | 23:26 | |
| NobodyCam | Shrews: commit message updated on the first of the fsm patches: https://review.openstack.org/#/c/139217 | 23:27 |
| Shrews | NobodyCam: re- +2'd | 23:28 |
| NobodyCam | :) I had a +2 | 23:28 |
| NobodyCam | but now I've edited it | 23:29 |
| *** zz_jgrimm has joined #openstack-ironic | 23:29 | |
| NobodyCam | lol | 23:29 |
| NobodyCam | devananda: gots a sec to toss a +1 on 139217 ... just to show your ok with the external (/me) edit... | 23:29 |
| *** penick has quit IRC | 23:34 | |
| *** ijw has quit IRC | 23:47 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Allow use of multiple simultaneous HW managers https://review.openstack.org/143193 | 23:51 |
| NobodyCam | Shrews: would you remove your -1 if I updated the bug on the commit message? | 23:54 |
| NobodyCam | for 145361 | 23:55 |
| devananda | NobodyCam: looks like rloo_afk removed her -1 but didn't + it ? | 23:55 |
| Shrews | NobodyCam: yup | 23:55 |
| *** hypnosb has joined #openstack-ironic | 23:55 | |
| NobodyCam | devananda: I'm just about to fix the bug number on 145361... | 23:56 |
| NobodyCam | devananda: on 139217? ya I just updated the commit message and rebased the set .. but rloo had dinner while I did that :-p | 23:56 |
| *** naohirot has joined #openstack-ironic | 23:56 | |
| devananda | hang on ... | 23:56 |
| Shrews | NobodyCam: what is the proper bug? | 23:57 |
| NobodyCam | https://bugs.launchpad.net/designate/+bug/1399257 | 23:57 |
| devananda | oh, hah. yes, wrong bug | 23:57 |
| devananda | somehow I copied the wrong ## | 23:57 |
| devananda | thanks | 23:57 |
| *** ryanpetrello has joined #openstack-ironic | 23:57 | |
| NobodyCam | :-p | 23:58 |
| devananda | I was wondering why it the bot didn't post a comment on the bug ... | 23:58 |
| *** dkingshott has quit IRC | 23:58 | |
| openstackgerrit | Chris Krelle proposed openstack/ironic: Hide oslo.messaging DEBUG logs by default https://review.openstack.org/145361 | 23:58 |
| NobodyCam | for edits like that the web gui thing isn't half bad I must say | 23:59 |
| openstackgerrit | Merged openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/142334 | 23:59 |
| Shrews | NobodyCam: heh, for some reason, i thought there was a diff bug. thx | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!