| *** ijw has quit IRC | 00:02 | |
| *** cemason has joined #openstack-ironic | 00:02 | |
| *** ijw has joined #openstack-ironic | 00:02 | |
| *** naohirot has joined #openstack-ironic | 00:07 | |
| *** cemason has quit IRC | 00:07 | |
| *** rloo has quit IRC | 00:11 | |
| *** aginwala has quit IRC | 00:16 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Follow up patch to correct code-contribute-guide https://review.openstack.org/254739 | 00:16 |
|---|---|---|
| *** sinval has joined #openstack-ironic | 00:19 | |
| openstackgerrit | Merged openstack/ironic: Explicitly depend on WebTest>=2.0 https://review.openstack.org/255793 | 00:24 |
| *** aginwala has joined #openstack-ironic | 00:28 | |
| *** Fdaisuke has quit IRC | 00:30 | |
| *** Sukhdev has joined #openstack-ironic | 00:31 | |
| *** jaybeale has quit IRC | 00:34 | |
| openstackgerrit | Merged openstack/ironic: Follow up patch to correct code-contribute-guide https://review.openstack.org/254739 | 00:35 |
| *** Fdaisuke has joined #openstack-ironic | 00:40 | |
| *** mjturek1 has quit IRC | 00:41 | |
| *** Sukhdev has quit IRC | 00:45 | |
| *** aginwala has quit IRC | 00:46 | |
| *** aginwala has joined #openstack-ironic | 00:49 | |
| *** jdk has joined #openstack-ironic | 00:51 | |
| *** chlong has joined #openstack-ironic | 00:52 | |
| *** maratoid has joined #openstack-ironic | 00:55 | |
| jroll | NobodyCam: that devstack review is +A if you want to land some stuff :) | 01:01 |
| *** praneshp_ has joined #openstack-ironic | 01:03 | |
| NobodyCam | 255879 bam. | 01:04 |
| jroll | <3 | 01:04 |
| NobodyCam | didn't make it all the way thru 255880 yet | 01:05 |
| jroll | no worries | 01:05 |
| NobodyCam | if not done by AM I will finish up | 01:05 |
| NobodyCam | :) | 01:05 |
| *** hoangcx has joined #openstack-ironic | 01:05 | |
| *** penick has quit IRC | 01:06 | |
| *** praneshp has quit IRC | 01:06 | |
| *** praneshp_ is now known as praneshp | 01:06 | |
| *** mtanino has quit IRC | 01:15 | |
| *** baoli has joined #openstack-ironic | 01:19 | |
| jlvillal | naohirot, ping | 01:19 |
| naohirot | jlvillal: Hi john, ge | 01:20 |
| jlvillal | naohirot, I think we are having a failure to communicate :) | 01:20 |
| *** cdearborn has joined #openstack-ironic | 01:20 | |
| jlvillal | Either I don't understand you or you don't understand me. | 01:20 |
| jlvillal | So it should be #host=localhost | 01:20 |
| jlvillal | It should NOT be #host=ironic | 01:20 |
| jlvillal | I think... | 01:21 |
| jlvillal | Now maybe I am wrong! | 01:21 |
| naohirot | jlvillal: really? why? | 01:21 |
| *** ijw has quit IRC | 01:21 | |
| *** rajathagasthya has quit IRC | 01:22 | |
| naohirot | jlvillal: #host=ironic is the original author's intention, I think. | 01:22 |
| jlvillal | #rabbit_host=localhost | 01:23 |
| jlvillal | That is what is there now | 01:23 |
| jlvillal | Ignore comment about #host | 01:23 |
| jroll | we will never be able to choose a sane production default here. so localhost is the closest we have. | 01:24 |
| naohirot | jlvillal: I couldn't figure out how generator.py can generate #rabbit_host=localhost. | 01:24 |
| jlvillal | jroll, Right. we are talking about https://review.openstack.org/#/c/255834/ | 01:25 |
| jroll | I know :) | 01:25 |
| jlvillal | jroll, So normally it gives 'localhost' | 01:25 |
| jlvillal | But if the system it is run on is named 'localhost' then the value goes to 'ironic' | 01:25 |
| jroll | jlvillal: yeah, I understand the patch | 01:25 |
| jlvillal | naohirot, On your system which I believe has a FQDN='localhost' | 01:26 |
| jlvillal | naohirot, So that causes it to become 'ironic' | 01:26 |
| jroll | naohirot: this is where the default comes from: https://github.com/openstack/oslo.messaging/blob/8e792a5f703628235710b4ee06d599b854e46111/oslo_messaging/_drivers/impl_rabbit.py#L89-L93 | 01:26 |
| naohirot | jlvillal: yes, my dev env is FQDN='localhost'. | 01:27 |
| jlvillal | naohirot, If you change it to 'example.com' | 01:27 |
| jlvillal | naohirot, Then it should change the generated value. | 01:27 |
| lintan | that's happen to me,too :) But I just wonder should we change the behavior of generator? | 01:28 |
| jroll | lintan: that's what's happening here :P | 01:28 |
| jlvillal | naohirot, My patch is trying to prevent cases where a system called 'localhost' causes values to change from default. | 01:28 |
| *** Marga_ has quit IRC | 01:29 | |
| jroll | well, what your patch is doing is: if the default is localhost, leave it at localhost | 01:29 |
| jroll | which I think is the right thing to do here | 01:29 |
| jlvillal | jroll, Yes. That is what I am attempting to do. | 01:29 |
| jlvillal | naohirot, Does it make sense what I am saying? | 01:30 |
| naohirot | jlvillal: I believe that even if FQND changed to example.com, output will not change, I think | 01:31 |
| jlvillal | naohirot, I disagree. Let me find link to code | 01:31 |
| naohirot | jlvillal: https://review.openstack.org/#/c/253498/5/etc/ironic/ironic.conf.sample L.123 | 01:31 |
| jlvillal | naohirot, https://github.com/openstack/ironic/blob/master/ironic/common/config_generator/generator.py#L254 | 01:32 |
| naohirot | jlvillal: 'value' is localhost | 01:32 |
| jlvillal | So the default value passed in is 'localhost' but the FQDN is also 'localhost'. | 01:32 |
| naohirot | jlvillal: if hostname is ubuntu, and fqdn is example.com | 01:32 |
| naohirot | jlvillal: then the logic is not applied | 01:33 |
| jlvillal | naohirot, Yes, so it did change. Correct? | 01:33 |
| jlvillal | naohirot, Because the FQDN changed from 'localhost' | 01:33 |
| naohirot | jlvillal: that means that value is returned as is. which is 'localhost'. | 01:33 |
| jlvillal | Yes | 01:34 |
| jlvillal | https://github.com/openstack/ironic/blob/master/ironic/common/config_generator/generator.py#L254-L256 | 01:34 |
| jlvillal | Is what happens when hostname or FQDN = 'localhost' and the value = 'localhost' and 'host' is part of the name of the value ( like rabbit_host ) | 01:34 |
| jlvillal | naohirot, Does it make sense? Do you agree with me or disagree? | 01:35 |
| * naohirot thinking | 01:35 | |
| lintan | hi naohirot, jlvilla, I read both of your patch, you both change to a same behavior but with different implementation detail... | 01:37 |
| naohirot | jlvillal: Yes I agree. it generate #host=ironic | 01:37 |
| *** david-lyle has joined #openstack-ironic | 01:37 | |
| jlvillal | naohirot, And #rabbit_host ???? | 01:37 |
| naohirot | jlvillal: #rabbit_host becames ironit too, not locahost. | 01:38 |
| naohirot | lintan: Yes, both changes are basically same. | 01:39 |
| jlvillal | naohirot, I'm confused. In which case? With FQDN = 'localhost' or FQDN != 'localhost'? | 01:39 |
| jlvillal | naohirot, I'm unsure if you agree with me or not :) | 01:40 |
| naohirot | jlvillal: let's discuss concretely, my case is localhost = ubuntu, fqdn = localhost | 01:42 |
| jlvillal | lintan, Yes similar work on the two patches. | 01:42 |
| jlvillal | naohirot, Okay. So then it becomes rabbit_host=ironic | 01:42 |
| naohirot | jlvillal: my case is host=ubuntu, fqdn=localhost, sorry | 01:43 |
| jlvillal | naohirot, I understood :) | 01:43 |
| lintan | so it looks like jlvilla will keep all 'localhost' as 'localhost', and naohirot will only keep 'localhost' for key/value pair which has 'host' as part of key name, | 01:43 |
| naohirot | jlvillal: even if I update my dev env, host=ubuntu, fqdn=example.com | 01:44 |
| jlvillal | naohirot, Now if change FQDN to something else. Maybe 'ubuntu' | 01:44 |
| jlvillal | Okay that also works | 01:44 |
| jlvillal | naohirot, And when you update. What does rabbit_host become? | 01:45 |
| naohirot | jlvillal: #rabbit_host becames localhost, but #host also becomes localhost. | 01:45 |
| jlvillal | naohirot, So do you see how changing the FQDN or hostname causes the config to change? | 01:46 |
| naohirot | jlvillal: I couldn't figure out the network settings which generate #rabbit_host=localhost, #host=ironic. | 01:46 |
| jlvillal | naohirot, Is your testing using any patches? | 01:46 |
| naohirot | jlvillal: yes, I did test using your code too. | 01:47 |
| *** spandhe has quit IRC | 01:47 | |
| jlvillal | naohirot, But the test just now? With my patch? | 01:48 |
| naohirot | jlvillal: In your env, can generator.py generate #rabbit_host=localhost, #host=ironic? what are your host and fqdn? | 01:48 |
| jlvillal | naohirot, I have NO patches installed | 01:48 |
| openstackgerrit | vinay kumar muddu proposed openstack/proliantutils: Use existing Ilo object for each Ilo operation https://review.openstack.org/256172 | 01:49 |
| openstackgerrit | vinay kumar muddu proposed openstack/proliantutils: Added test case and logs https://review.openstack.org/256173 | 01:49 |
| naohirot | jlvillal: you mean you didn't test? | 01:49 |
| jlvillal | Though I did add a 'return' on line 274 | 01:49 |
| jlvillal | naohirot, I was testing to see how changing the hostname caused the values to change | 01:50 |
| *** [1]cdearborn has joined #openstack-ironic | 01:50 | |
| naohirot | jlvillal: is your patch same code you made comment in my patch, right? | 01:52 |
| jlvillal | naohirot, I did test my patch before. But I was talking right now | 01:52 |
| jlvillal | naohirot, Yes | 01:52 |
| naohirot | jlvillal: I certainly tested your code, and confirmed that #host became localhost too in my dev env. | 01:53 |
| jlvillal | naohirot, So without any patches. When I change from 'ubuntu' hostname to 'localhost' hostname. | 01:53 |
| jlvillal | Then #rabbit_host=ironic happens | 01:54 |
| jlvillal | Before it was #rabbit_host=localhost with hostname/FQDN = 'ubuntu' | 01:54 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add grenade plugin https://review.openstack.org/256174 | 01:55 |
| jroll | woo | 01:55 |
| naohirot | jlvillal: then what was #host=? | 01:55 |
| jlvillal | naohirot, #host=ironic on my system | 01:56 |
| naohirot | jlvillal: Okay, then what is the result of applying the patch? | 01:57 |
| jlvillal | naohirot, Same as what it is currently in the git repo | 01:57 |
| *** rajathagasthya has joined #openstack-ironic | 01:57 | |
| jlvillal | naohirot, A few moments | 01:58 |
| *** Marga_ has joined #openstack-ironic | 02:00 | |
| *** jamielennox is now known as jamielennox|away | 02:01 | |
| *** rajathagasthya has quit IRC | 02:01 | |
| jlvillal | naohirot, I found one issue. Changes rpc_zmq_host=localhost | 02:02 |
| jlvillal | naohirot, If my system is called 'localhost' | 02:02 |
| jlvillal | naohirot, I changed my patch to -1 and will investigate. | 02:02 |
| jlvillal | naohirot, I want output to be the same | 02:02 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove override-defaults https://review.openstack.org/255914 | 02:02 |
| naohirot | jlvillal: okay, but hostname=localhost is not typical. | 02:02 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Clean up variables in plugin.sh https://review.openstack.org/255936 | 02:03 |
| jroll | I think that's everything. | 02:03 |
| jlvillal | naohirot, It is somewhat typical. It is a decent default. | 02:03 |
| jroll | it's a horrible default | 02:03 |
| jroll | but everything else is a more horrible default | 02:03 |
| jlvillal | naohirot, localhost will exist. Where 'ironic' probably doesn't | 02:03 |
| jlvillal | :) | 02:03 |
| naohirot | jlvillal: because when we install ubuntu KVM or other vmguest image, hostname is ubuntu. | 02:04 |
| jlvillal | naohirot, But 'localhost' works also. | 02:04 |
| jroll | localhost will almost always be resolvable. anything else probably will not. | 02:04 |
| jlvillal | naohirot, I'll dig into this a little more and maybe upload a follow up patch tonight. Or tomorrow. | 02:05 |
| jlvillal | naohirot, Thanks for the discussion. | 02:05 |
| *** cdearborn has quit IRC | 02:05 | |
| * jlvillal goes off to find where rpc_zmp_host value is found... | 02:05 | |
| naohirot | jlvillal: hostname should be output of 'hostname' command | 02:05 |
| jlvillal | naohirot, I don't understand. | 02:06 |
| jroll | jlvillal: that's also oslo.messaging | 02:06 |
| jlvillal | jroll, Thanks | 02:06 |
| naohirot | jlvillal: I mean that hostname = socket.gethostname() is same as 'hostname' unix command. | 02:07 |
| naohirot | jlvillal: if 'hostname' unix command returned 'localhost', it means that the host dosn't have any hostname at all. | 02:09 |
| *** smoriya_ has joined #openstack-ironic | 02:09 | |
| jlvillal | naohirot, Okay. But I'm not understanding what trying to tell me. | 02:09 |
| jlvillal | naohirot, Could be because I am also trying to troubleshoot at same time... | 02:09 |
| jlvillal | :) | 02:09 |
| naohirot | jlvillal: so it is not typical case, host ususually has one name beside 'localhost'. | 02:09 |
| jlvillal | naohirot, Yes it is not typical but I think it can happen with people developing using VMs | 02:10 |
| jlvillal | naohirot, Or the FQDN = 'localhost' | 02:10 |
| naohirot | jlvillal: I'm trying tell you when you test, use hostname=ubuntu, and fqdn=localhost, but not hostname=localhost and fqdn=example.com. | 02:11 |
| jlvillal | naohirot, Okay. Thanks! :) | 02:12 |
| naohirot | jlvillal: wc, so let's continue to discuss on gerrit :-) | 02:12 |
| jlvillal | naohirot, Okay. I probably will stop working soon. | 02:13 |
| *** Marga_ has quit IRC | 02:15 | |
| naohirot | jlvillal: one thing I need to tell you that rpc_zmq_host= is no longer used in oslo.messaging 3.1.0. it disappeared. | 02:15 |
| jlvillal | naohirot, Oh. So maybe I shouldn't worry :) | 02:16 |
| *** Marga_ has joined #openstack-ironic | 02:16 | |
| *** [1]cdearborn has quit IRC | 02:16 | |
| naohirot | jlvillal: Yep, it' minor thing :) | 02:16 |
| jlvillal | naohirot, I'll do some investigating first. Thank you though. | 02:17 |
| naohirot | jlvillal: Okay, I'm looking forward to seeing the result :) | 02:18 |
| jlvillal | Thanks | 02:18 |
| *** r-daneel has quit IRC | 02:22 | |
| *** cemason has joined #openstack-ironic | 02:29 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove override-defaults https://review.openstack.org/255914 | 02:29 |
| *** jamielennox|away is now known as jamielennox | 02:29 | |
| *** baoli has quit IRC | 02:30 | |
| *** baoli has joined #openstack-ironic | 02:32 | |
| *** cemason has quit IRC | 02:34 | |
| *** baoli has quit IRC | 02:35 | |
| *** sinval has left #openstack-ironic | 02:35 | |
| *** alexpilo_ has joined #openstack-ironic | 02:56 | |
| *** alexpilotti has quit IRC | 02:56 | |
| *** hoangcx has quit IRC | 02:59 | |
| openstackgerrit | Tan Lin proposed openstack/ironic: Add a config option [upgrade_levels]/conductor https://review.openstack.org/253355 | 03:03 |
| *** jaybeale has joined #openstack-ironic | 03:03 | |
| *** Fdaisuke has quit IRC | 03:10 | |
| *** jaybeale has quit IRC | 03:13 | |
| *** Fdaisuke has joined #openstack-ironic | 03:13 | |
| *** aginwala has quit IRC | 03:19 | |
| *** boris-42_ has quit IRC | 03:23 | |
| *** Nisha has joined #openstack-ironic | 03:24 | |
| *** alexpilo_ has quit IRC | 03:32 | |
| openstackgerrit | Merged openstack/ironic: Add oslo_config.Opt support in Ironic config generator https://review.openstack.org/253498 | 03:35 |
| *** cemason has joined #openstack-ironic | 03:37 | |
| *** links has joined #openstack-ironic | 03:42 | |
| *** Marga_ has quit IRC | 03:51 | |
| *** cemason has quit IRC | 03:57 | |
| *** Marga_ has joined #openstack-ironic | 03:57 | |
| *** cemason has joined #openstack-ironic | 03:57 | |
| *** cemason has quit IRC | 03:57 | |
| *** dims has quit IRC | 04:12 | |
| openstackgerrit | Merged openstack/ironic: Copy devstack code to ironic tree https://review.openstack.org/255879 | 04:15 |
| *** dims has joined #openstack-ironic | 04:19 | |
| *** hoangcx has joined #openstack-ironic | 04:24 | |
| *** spandhe has joined #openstack-ironic | 04:26 | |
| *** smoriya has quit IRC | 04:30 | |
| *** dims has quit IRC | 04:30 | |
| *** dims has joined #openstack-ironic | 04:31 | |
| *** alexpilotti has joined #openstack-ironic | 04:33 | |
| *** garthb has quit IRC | 04:34 | |
| *** spandhe_ has joined #openstack-ironic | 04:35 | |
| *** spandhe has quit IRC | 04:36 | |
| *** spandhe_ is now known as spandhe | 04:36 | |
| *** smoriya_afk has joined #openstack-ironic | 04:36 | |
| *** smoriya_afk is now known as smoriya | 04:36 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: Fix issue where system hostname can impact genconfig https://review.openstack.org/255834 | 04:37 |
| *** alexpilotti has quit IRC | 04:37 | |
| *** baoli has joined #openstack-ironic | 04:46 | |
| *** Marga_ has quit IRC | 04:50 | |
| *** kkoski has quit IRC | 04:50 | |
| *** baoli has quit IRC | 04:51 | |
| *** Marga_ has joined #openstack-ironic | 04:51 | |
| *** Marga_ has quit IRC | 04:56 | |
| *** hoangcx has quit IRC | 04:58 | |
| *** Nisha has quit IRC | 05:01 | |
| *** HimanshuGarg has quit IRC | 05:07 | |
| *** dims has quit IRC | 05:08 | |
| *** HimanshuGarg has joined #openstack-ironic | 05:09 | |
| *** ig0r__ has joined #openstack-ironic | 05:15 | |
| *** ig0r_ has quit IRC | 05:19 | |
| *** ig0r__ has quit IRC | 05:21 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Update ironic.conf.sample by applying the bug fix #1522841 https://review.openstack.org/255738 | 05:22 |
| *** Marga_ has joined #openstack-ironic | 05:23 | |
| *** Marga_ has quit IRC | 05:23 | |
| *** Marga_ has joined #openstack-ironic | 05:23 | |
| *** Marga_ has quit IRC | 05:24 | |
| *** Marga_ has joined #openstack-ironic | 05:25 | |
| *** smoriya has quit IRC | 05:30 | |
| *** amotoki has joined #openstack-ironic | 05:31 | |
| *** alexpilotti has joined #openstack-ironic | 05:34 | |
| *** alexpilotti has quit IRC | 05:38 | |
| *** alexpilotti has joined #openstack-ironic | 05:52 | |
| devananda | TheJulia: bifrost doesn't seem to be setting up rootwrap config? | 05:53 |
| *** alexpilotti has quit IRC | 05:57 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Update ironic.conf.sample by applying the bug fix #1522841 https://review.openstack.org/255738 | 05:59 |
| *** hoangcx has joined #openstack-ironic | 06:04 | |
| devananda | TheJulia: testing pxe_amt out on a 5th gen nuc, and noticed bifrost isn't setting up rootwrap. don't know if this matters to anyone, but the fix is easy. details here: https://bugs.launchpad.net/bifrost/+bug/1525082 | 06:10 |
| openstack | Launchpad bug 1525082 in bifrost ""pxe" drivers not supported" [Undecided,New] | 06:10 |
| *** chlong has quit IRC | 06:15 | |
| *** Marga_ has quit IRC | 06:20 | |
| *** ChubYann has quit IRC | 06:21 | |
| *** yonglihe has quit IRC | 06:21 | |
| *** zhenguo has joined #openstack-ironic | 06:24 | |
| *** smoriya_afk has joined #openstack-ironic | 06:28 | |
| *** smoriya_afk is now known as smoriya | 06:28 | |
| *** harshs has joined #openstack-ironic | 06:29 | |
| *** jrist has quit IRC | 06:33 | |
| *** jamielennox is now known as jamielennox|away | 06:33 | |
| *** jrist has joined #openstack-ironic | 06:35 | |
| *** jrist has quit IRC | 06:35 | |
| *** jrist has joined #openstack-ironic | 06:35 | |
| *** harshs has quit IRC | 06:40 | |
| *** harshs has joined #openstack-ironic | 06:42 | |
| *** sandhya has joined #openstack-ironic | 06:45 | |
| sandhya | Ramakrishnan, I have questions regarding the UEFI iSCSI boot for Proliant hardware patch https://review.openstack.org/#/c/207337/7. | 06:47 |
| *** dims has joined #openstack-ironic | 06:52 | |
| *** alexpilotti has joined #openstack-ironic | 06:53 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic: Use Tempest plugin interface https://review.openstack.org/246161 | 06:53 |
| *** dims has quit IRC | 06:56 | |
| *** alexpilotti has quit IRC | 06:58 | |
| *** zer0c00l has quit IRC | 06:59 | |
| *** dims has joined #openstack-ironic | 07:10 | |
| *** Marga_ has joined #openstack-ironic | 07:12 | |
| *** zer0c00l has joined #openstack-ironic | 07:16 | |
| *** pas-ha has joined #openstack-ironic | 07:16 | |
| *** spandhe_ has joined #openstack-ironic | 07:31 | |
| *** ig0r_ has joined #openstack-ironic | 07:32 | |
| *** spandhe has quit IRC | 07:32 | |
| *** spandhe_ is now known as spandhe | 07:32 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic-specs: Add allowed transitions to API https://review.openstack.org/224022 | 07:43 |
| *** spandhe has quit IRC | 07:54 | |
| *** ionutbalutoiu has joined #openstack-ironic | 07:55 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add console feature to ssh driver https://review.openstack.org/241453 | 07:56 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Enable ironic node pty console by default https://review.openstack.org/256272 | 07:56 |
| *** sandhya has quit IRC | 07:58 | |
| *** romcheg1 has quit IRC | 08:09 | |
| *** Nisha has joined #openstack-ironic | 08:10 | |
| *** ig0r_ has quit IRC | 08:11 | |
| *** GheRivero has left #openstack-ironic | 08:16 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add hardware inspection module for iRMC driver https://review.openstack.org/196480 | 08:17 |
| *** ig0r_ has joined #openstack-ironic | 08:19 | |
| *** harshs has quit IRC | 08:25 | |
| *** MattMan has quit IRC | 08:32 | |
| *** MattMan has joined #openstack-ironic | 08:32 | |
| *** jcoufal has joined #openstack-ironic | 08:38 | |
| *** rcernin has joined #openstack-ironic | 08:44 | |
| *** ig0r_ has quit IRC | 08:53 | |
| *** ifarkas has joined #openstack-ironic | 08:54 | |
| *** mbound_ has joined #openstack-ironic | 09:04 | |
| *** mbound_ has quit IRC | 09:12 | |
| *** mbound_ has joined #openstack-ironic | 09:12 | |
| *** athomas has joined #openstack-ironic | 09:17 | |
| *** ig0r_ has joined #openstack-ironic | 09:20 | |
| *** mbound_ has quit IRC | 09:28 | |
| *** kalpase has joined #openstack-ironic | 09:28 | |
| *** mbound_ has joined #openstack-ironic | 09:29 | |
| *** dims is now known as dimsum__ | 09:29 | |
| *** kozhukalov_ has joined #openstack-ironic | 09:37 | |
| *** _degorenko|afk is now known as degorenko | 09:39 | |
| *** praneshp has quit IRC | 09:40 | |
| *** chenke_ has quit IRC | 09:44 | |
| *** ig0r_ has quit IRC | 09:47 | |
| *** amotoki has quit IRC | 09:50 | |
| *** alexpilotti has joined #openstack-ironic | 09:52 | |
| *** jistr has joined #openstack-ironic | 09:55 | |
| *** dtantsur|afk is now known as dtantsur | 09:56 | |
| dtantsur | Morning Ironic, TGIF finally | 09:56 |
| *** alexpilotti has quit IRC | 09:57 | |
| *** naohirot has quit IRC | 09:57 | |
| *** zhenguo has quit IRC | 10:01 | |
| *** ndipanov has joined #openstack-ironic | 10:05 | |
| *** ionutbalutoiu has quit IRC | 10:05 | |
| *** ionutbalutoiu has joined #openstack-ironic | 10:06 | |
| *** mkovacik has joined #openstack-ironic | 10:11 | |
| *** Nisha has quit IRC | 10:12 | |
| *** hoangcx has left #openstack-ironic | 10:25 | |
| *** alexpilotti has joined #openstack-ironic | 10:34 | |
| *** e0ne has joined #openstack-ironic | 10:36 | |
| *** alexpilotti has quit IRC | 10:38 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/255552 | 10:46 |
| *** alexpilotti has joined #openstack-ironic | 10:51 | |
| *** alexpilotti has quit IRC | 10:56 | |
| *** ndipanov has quit IRC | 10:58 | |
| sambetts | o/ dtantsur, and everyone ese TGIF | 10:59 |
| dtantsur | o/ | 11:00 |
| *** ig0r_ has joined #openstack-ironic | 11:09 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-specs: Implement libvirt-based driver for virtual HW https://review.openstack.org/254421 | 11:10 |
| aarefiev | morning! | 11:10 |
| sambetts | o/ aarefiev | 11:11 |
| aarefiev | sambetts, dtantsur: morning | 11:11 |
| *** admin0 has joined #openstack-ironic | 11:11 | |
| admin0 | is it possible to use ironic without ipmi but just pxe ? | 11:12 |
| *** mgould has joined #openstack-ironic | 11:12 | |
| sambetts | admin0: do you want manual power control? or some other power control method instead of ipmi? | 11:15 |
| admin0 | i am trying to setup a PoC at home and the hardware i have just supports pxe :) | 11:16 |
| sambetts | Does it support wake on lan? | 11:17 |
| sambetts | we have a driver for that | 11:17 |
| admin0 | 4 of them do . wake on lan whch auto activates the PXE | 11:17 |
| admin0 | 3 of them do have wake-on-lan, but i have to select for pxe after that | 11:17 |
| lucasagomes | morning all | 11:18 |
| openstackgerrit | Cameron.C proposed openstack/ironic: Fix not check duplicated node when enrolling node https://review.openstack.org/256356 | 11:18 |
| sambetts | can't you set PXE to always be the first boot device on those machines? | 11:19 |
| *** alexpilotti has joined #openstack-ironic | 11:19 | |
| admin0 | i can | 11:19 |
| admin0 | so wtih wake on lan and first boot to pxe on those 3, i should be abel to use ironic ? | 11:19 |
| sambetts | yeah that should work, because ironic will even use pxe to boot the user image after deploy | 11:20 |
| admin0 | ok .. i will give it a try today ( using rdo-manager ) | 11:20 |
| aarefiev | sambetts: have a min to talk about discovery ? | 11:22 |
| aarefiev | sambetts | 11:23 |
| sambetts | o/ | 11:23 |
| aarefiev | oops second time was accidentally | 11:23 |
| aarefiev | sambetts: I'm thinking about not create node with fake driver in ironic, but just create node_info and after all rules applies create node with all info | 11:24 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Allow to build user image with DIB https://review.openstack.org/256363 | 11:24 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: refactor ironic enroll-node code https://review.openstack.org/256364 | 11:24 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add Link-Local-Connection info to ironic port https://review.openstack.org/256365 | 11:24 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update Ironic VM network connection https://review.openstack.org/256366 | 11:24 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add configure_provision_network function https://review.openstack.org/256367 | 11:24 |
| sambetts | aarefiev: but what if there aren't any rules to set the driver type, I still want that node discovered and added to my ironic db so that it can be identified if I decided to reinspect it later | 11:26 |
| aarefiev | sambetts: we can have default rule for this, also it's possible to merge actions(rules?) to avoid multiple calls to ironic | 11:27 |
| aarefiev | It's about optimization | 11:28 |
| aarefiev | or not rule just create fake node in this case | 11:28 |
| sambetts | aarefiev: I would keep the node create, and the node update separate, leave the node not found hook to do the addition, but perhaps we can optimise the node update and make it a bulk operation, although if I remember correctly we used to do that and got rid of it | 11:31 |
| sambetts | dtantsur: ^ | 11:31 |
| *** chlong has joined #openstack-ironic | 11:33 | |
| aarefiev | sambetts: I just don't see any benefit of this, why not doing all this stuff in inspector and than one update ironic state? | 11:34 |
| aarefiev | I mean only about one node | 11:35 |
| dtantsur | sorry, I'm not quite following. could someone give me a tl;dr? | 11:36 |
| aarefiev | dtantsur: I'm proposing to create or update node in one request to ironic, after applying all rules | 11:38 |
| aarefiev | we can merge all rules on import time | 11:39 |
| dtantsur | aarefiev, that's kind of a big rewrite. how will hooks work after that? | 11:39 |
| dtantsur | how do we figure out which changes to apply actually? | 11:39 |
| sambetts | my opinion is that we can implement this without drastic changes to the way that ironic inspector operates today, doing it the way I see would only require the addition of the node not found hook because it uses the rules just like it does for any other node, doing what your suggesting isn't so much implementing auto discovery but optimising rules and processing hooks, the way I see it is that w | 11:39 |
| sambetts | e're only adding 1 more call to ironic | 11:39 |
| aarefiev | also few call to update in case a few actions I guess | 11:40 |
| aarefiev | for one node | 11:40 |
| sambetts | no more calls than if we had actions for non-autodiscovered node | 11:41 |
| dtantsur | I still don't fully understand the context, but let's not optimize calls to ironic at the cost of internal complexity | 11:42 |
| sambetts | ++ | 11:42 |
| dtantsur | we used to do that some time ago, and it didn't work really well | 11:42 |
| openstackgerrit | Dmitry Tantsur proposed openstack/python-ironicclient: Scale back on how many warnings we issue https://review.openstack.org/256375 | 11:45 |
| *** ig0r_ has quit IRC | 11:46 | |
| aarefiev | dtantsur:AFAIU inspector make new calls on each action, what the problem with merging them, current approach isn't scalable | 11:47 |
| dtantsur | how do you merge them? | 11:48 |
| dtantsur | also "isn't scalable" needs to be proved. I'm not convinced that issuing 4-5 HTTP requests per node once in a while doesn't scale | 11:49 |
| aarefiev | dtantsur: the same operation I mean set-attribute, why not set all attributes in one request | 11:50 |
| dtantsur | re merge, simple exercise: hook/rule 1: "set /properties/xxx to 1", hook/rule 2: "if /properties/yyy == 3 then /properties/xxx += 1" | 11:51 |
| dtantsur | please merge them | 11:51 |
| dtantsur | if it's not fun enough, add something modifying /properties/yyy | 11:51 |
| dtantsur | our biggest problem was /properties/capabilities which can't be updated atomically | 11:52 |
| *** lobaille has joined #openstack-ironic | 11:53 | |
| *** sinval-afk is now known as sinval | 11:54 | |
| sinval | morning everyone | 11:54 |
| sambetts | o/ sinval | 11:54 |
| dtantsur | morning sinval | 11:55 |
| aarefiev | sinval: morning | 11:55 |
| sinval | o/ | 11:57 |
| mgould | morning Ironic | 11:58 |
| sinval | mgould, good morning | 11:58 |
| *** alexpilotti has quit IRC | 11:59 | |
| *** jdk has quit IRC | 12:00 | |
| aarefiev | dtantsur: in your example it's operator 's error or not error, don't see any diff in current approach and my, this problem present's in both cases; ok never mind if folks disagree on this | 12:00 |
| aarefiev | mgould: morning | 12:00 |
| dtantsur | sorry, where do you seen an error? I'm just saying that it's pretty hard to merge updates to capabilities, for example | 12:01 |
| *** alexpilotti has joined #openstack-ironic | 12:03 | |
| dtantsur | again, maybe I don't know the full context, I didn't check the spec yet | 12:03 |
| dtantsur | morning mgould | 12:03 |
| openstackgerrit | Sinval Vieira Mendes Neto proposed openstack/ironic: Improving capabilities_to_dict and adding helpful tests https://review.openstack.org/254972 | 12:03 |
| *** smoriya has quit IRC | 12:05 | |
| aarefiev | dtantsur: ok, maybe I understand it wrong, I mean make this updates on inspector node_info and then update ironic | 12:07 |
| dtantsur | this would require us to duplicate the whole JSON-patch parsing logic in inspector | 12:07 |
| *** raildo-afk is now known as raildo | 12:08 | |
| aarefiev | dtantsur: ok, nevermind, it's not a problem now | 12:10 |
| *** kalpase has quit IRC | 12:10 | |
| sambetts | aarefiev, dtantsur: We can always add a spec for optimising rules and processing hooks | 12:11 |
| dtantsur | yes, if someone feels like | 12:11 |
| aarefiev | sambetts: sure | 12:11 |
| *** ig0r_ has joined #openstack-ironic | 12:15 | |
| sambetts | tbh scaling inspector isn't something we've really talked about yet, but for bulk inspection of many nodes in paralel we probably want to be able to horizontally scale the inspector service | 12:15 |
| sambetts | I'd expect it should work under a loadbalancer now except for the semaphores | 12:16 |
| sambetts | which when we start using toos is solved anyway | 12:16 |
| *** ionutbalutoiu has quit IRC | 12:17 | |
| *** ionutbalutoiu has joined #openstack-ironic | 12:18 | |
| *** rcernin has quit IRC | 12:18 | |
| *** links has quit IRC | 12:19 | |
| dtantsur | sambetts, yeah, mkovacik is working on an HA spec, have you seen it? | 12:19 |
| *** degorenko is now known as _degorenko|afk | 12:24 | |
| *** thrash|g0ne is now known as thrash | 12:24 | |
| *** ig0r_ has quit IRC | 12:24 | |
| *** admin0 has quit IRC | 12:28 | |
| sambetts | dtantsur: I think I remember hearing about it, we don't have a BP for it though | 12:34 |
| dtantsur | sambetts, we have a spec up already | 12:34 |
| dtantsur | sambetts, https://review.openstack.org/253675 | 12:34 |
| sambetts | dtantsur: awesome :), just had a quick scan, it looks good so far, we need a BP or RFE and need to update the inspector inbox with specs like the ironic one has | 12:38 |
| *** dprince has joined #openstack-ironic | 12:39 | |
| *** mgoddard has joined #openstack-ironic | 12:44 | |
| mkovacik | dtantsur, sambetts, yeah, the spec is ready for review. I'm just about to file a [RFE] bug to track the implementation progress | 12:54 |
| *** _degorenko|afk is now known as degorenko | 12:58 | |
| jlvillal | Good morning Ironic | 12:59 |
| mgould | jlvillal, morning! | 13:00 |
| jlvillal | :) | 13:00 |
| sambetts | dtantsur: Are we going port all our existing BPs to RFEs? | 13:00 |
| dtantsur | sambetts, yes! | 13:01 |
| dtantsur | morning jlvillal | 13:01 |
| dtantsur | sambetts, I just didn't have any time to do it | 13:01 |
| mgould | sambetts, that's kinda relevant to my current python-tripleoclient patch | 13:01 |
| dtantsur | we should also update you spec template to mention bug links | 13:01 |
| sambetts | ++ | 13:01 |
| mgould | bulk introspection currently inspects all nodes in series, including timeouts | 13:01 |
| mgould | I'm wondering if it's in-scope to make it happen in parallel | 13:02 |
| jlvillal | dtantsur, lucasagomes, sambetts or anyone else. https://review.openstack.org/#/c/255885/1/ironic/dhcp/neutron.py Any thoughts if that should be 'network' instead of 'neutron' | 13:02 |
| dtantsur | mgould, not sure if question belongs here, bulk introspection is part of tripleo | 13:02 |
| lucasagomes | jlvillal, morning! | 13:03 |
| * lucasagomes looks | 13:03 | |
| jlvillal | lucasagomes, morning :) | 13:03 |
| dtantsur | jlvillal, depending on what get_service_url accepts: service type or name | 13:03 |
| jlvillal | dtantsur, The function claims it is service_type, I believe | 13:03 |
| dtantsur | jlvillal, then it should be network | 13:04 |
| *** mgoddard_ has joined #openstack-ironic | 13:04 | |
| lucasagomes | jlvillal, hmmm needs some investigation, but I think it's network | 13:04 |
| jlvillal | lucasagomes, dtantsur: Thanks. I thought so, but wanted a 2nd opinion :) | 13:04 |
| jlvillal | I assume most times the CONF value is getting used. | 13:04 |
| lucasagomes | jlvillal, https://github.com/openstack/ironic/blob/master/ironic/common/keystone.py#L110 | 13:05 |
| mgould | dtantsur, thanks - I'll ask over in #tripleo :-) | 13:05 |
| lucasagomes | jlvillal, yeah, that's why we don't actually hit that yet | 13:05 |
| mkovacik | dtantsur, sambetts FYI: https://bugs.launchpad.net/ironic-inspector/+bug/1525218 | 13:05 |
| openstack | Launchpad bug 1525218 in Ironic Inspector "[RFE] Introduce High Availability to Ironic Inspector Service" [Undecided,New] | 13:05 |
| sambetts | good catch jlvillal, I agree I think it should be network, I've just grepped over nova and horizon and neither of them use get_service_url | 13:05 |
| *** mgoddard has quit IRC | 13:06 | |
| sambetts | mkovacik: Thanks! | 13:06 |
| mkovacik | np | 13:06 |
| mkovacik | ;) | 13:06 |
| jlvillal | lucasagomes, https://github.com/openstack/ironic/blob/master/ironic/common/keystone.py#L110-L136 | 13:06 |
| sambetts | dtantsur: Do we delete the BPs once we've made them into RFEs? | 13:06 |
| jlvillal | lucasagomes, never mind! you did that :) | 13:06 |
| lucasagomes | :-) | 13:06 |
| jlvillal | sambetts, That is an Ironic custom function: https://github.com/openstack/ironic/blob/master/ironic/common/keystone.py#L110-L136 | 13:07 |
| sambetts | ah | 13:07 |
| dtantsur | sambetts, mmmm... good question. that would break some links, but I don't know how to make blueprints read-only | 13:07 |
| * dtantsur looks | 13:07 | |
| dtantsur | sambetts, jroll, what about "needs-spec" tag for RFE's which need spec? | 13:09 |
| sambetts | I think neutron uses one of the statuses to indicate it | 13:10 |
| sambetts | as well as comment I guess | 13:10 |
| *** jistr is now known as jistr|biab | 13:10 | |
| *** kozhukalov_ has quit IRC | 13:13 | |
| dtantsur | sambetts, btw when creating an RFE, please set it to either confirmed/wishlist or triaged/wishlist depending on whether we accepted it | 13:13 |
| sambetts | dtantsur: which is which? | 13:14 |
| sambetts | confirmed == accepted? | 13:15 |
| dtantsur | sambetts, I guess confirmed means we merely accept it as a valid request, triaged means we accepted it for inclusion ("approved" status in blueprints) | 13:15 |
| sambetts | ok :) btw I don't have the power to set the importance | 13:16 |
| sambetts | :( | 13:16 |
| *** ig0r_ has joined #openstack-ironic | 13:16 | |
| sambetts | dtantsur: can you add me to the inspector bug team? | 13:17 |
| *** athomas has quit IRC | 13:18 | |
| dtantsur | sambetts, just join https://launchpad.net/~ironic-bugs | 13:19 |
| sambetts | in other projects I've worked on the bug team is invite only | 13:19 |
| sambetts | so I assumed the same | 13:19 |
| dtantsur | :) | 13:21 |
| *** athomas has joined #openstack-ironic | 13:24 | |
| *** ig0r_ has quit IRC | 13:25 | |
| jlvillal | dtantsur, Do you know who shuquan huang is? I noticed they grabbed the bug I added. | 13:29 |
| TheJulia | devananda: I guess nothing has really required it thus far... I'll take a look at it, thank you | 13:29 |
| jlvillal | dtantsur, But then I noticed they have 38 bugs assigned to them. 33 of which are open.... | 13:29 |
| jlvillal | https://launchpad.net/~shuquan/+assignedbugs?search=Search | 13:30 |
| dtantsur | jlvillal, no, no idea, but I also noticed that | 13:30 |
| jlvillal | dtantsur, Okay. I'm just worried they grab bugs but then don't have the time to work on them... | 13:30 |
| dtantsur | I was going to start pinging and unassigning people who are assigned for >2 weeks without a patch posted | 13:30 |
| dtantsur | jlvillal, I'd really appreciate help in the endeavor :) | 13:31 |
| *** Nisha has joined #openstack-ironic | 13:31 | |
| jlvillal | :) Yeah, I'm not sure what to do. Ping them in the bug and ask them if they really have the time to work on it? Or send email.... | 13:31 |
| *** dims has joined #openstack-ironic | 13:32 | |
| *** mgoddard_ has quit IRC | 13:33 | |
| *** mgould has quit IRC | 13:33 | |
| *** dimsum__ has quit IRC | 13:36 | |
| *** alexpilotti has quit IRC | 13:36 | |
| *** alexpilotti has joined #openstack-ironic | 13:39 | |
| dtantsur | jlvillal, ping in the bug, but make sure they're subscribed to them | 13:41 |
| jlvillal | dtantsur, Okay | 13:41 |
| dtantsur | jlvillal, just being an assignee does not mean subscribing (which is weird, but still) | 13:41 |
| sambetts | dtantsur: I'm pretty sure we've got an RFE for every BP now | 13:41 |
| dtantsur | awesome thanks! | 13:41 |
| jlvillal | dtantsur, That is wierd :) | 13:42 |
| jlvillal | dtantsur, done | 13:43 |
| dtantsur | sambetts, there's no option to make blueprints read-only, not sure what to do | 13:44 |
| *** baoli has joined #openstack-ironic | 13:45 | |
| *** mjturek1 has joined #openstack-ironic | 13:45 | |
| sambetts | dtantsur: I think we can delete the ones that have no spec right? because there isn't really anything linked back to them? | 13:45 |
| sambetts | dtantsur: if we turn off bp tracking in LP does it just delete everything? | 13:45 |
| *** derekh has joined #openstack-ironic | 13:45 | |
| dtantsur | sambetts, I don't know, I can only try :) | 13:45 |
| sambetts | I'd hope it'll throw up a confirmation box if it does | 13:46 |
| sambetts | :-P | 13:46 |
| *** dprince has quit IRC | 13:46 | |
| *** mgould has joined #openstack-ironic | 13:46 | |
| *** baoli_ has joined #openstack-ironic | 13:46 | |
| dtantsur | well, lemme try it.. broken links is probably better than people submitting blueprints without feedback.. | 13:47 |
| dtantsur | sambetts, awesome, old links still work | 13:47 |
| sambetts | awesome :D | 13:47 |
| sambetts | you just can't see them from blueprints.launchpad.net/ironic-inspector | 13:48 |
| dtantsur | yes | 13:49 |
| *** baoli has quit IRC | 13:50 | |
| *** dprince has joined #openstack-ironic | 13:50 | |
| dtantsur | sambetts, also dropping unreleased 2.4.0 milestone, right? | 13:50 |
| sambetts | I'm not sure, we can still use it for targeting things right? | 13:51 |
| sambetts | or are we just going use the prioritys to target what we should be doing next? | 13:51 |
| *** lucasagomes is now known as lucas-afk | 13:51 | |
| sambetts | have everything priorised and then cut a release when say we've completed all the highs an above? | 13:52 |
| dtantsur | sambetts, yeah. we also make releases on more or less timely fashion | 13:53 |
| sambetts | cool, then yeah we can get rid of the milestones too then | 13:54 |
| jlvillal | dtantsur, They grabbed both the bugs I opened... So I added comment to both bugs. | 13:55 |
| dtantsur | good | 13:55 |
| *** kozhukalov_ has joined #openstack-ironic | 13:55 | |
| *** jrist has quit IRC | 13:56 | |
| sambetts | we had someone like that at one point when I was working on horizon, they ended up with ~40+ bugs assigned to them we very little activity, so we just had to take action | 13:56 |
| *** jrist has joined #openstack-ironic | 13:58 | |
| *** jrist has quit IRC | 13:58 | |
| *** jrist has joined #openstack-ironic | 13:58 | |
| *** dprince_ has joined #openstack-ironic | 14:02 | |
| dtantsur | I like how our bugs list looks now | 14:02 |
| sambetts | its pretty cool :D | 14:02 |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: Use 'service_type' of 'network'. Not 'neutron' https://review.openstack.org/256420 | 14:03 |
| sambetts | we can now priorised the RFEs against each other too :D | 14:04 |
| *** admin0 has joined #openstack-ironic | 14:04 | |
| TheJulia | mordred: bad news for that rev, posted comments and a backtrace in the review | 14:05 |
| *** dtantsur is now known as dtantsur|brb | 14:08 | |
| *** jistr|biab is now known as jistr | 14:09 | |
| *** mbound_ has quit IRC | 14:10 | |
| *** derekh has quit IRC | 14:11 | |
| *** mbound_ has joined #openstack-ironic | 14:12 | |
| *** mbound_ has quit IRC | 14:12 | |
| *** mbound_ has joined #openstack-ironic | 14:12 | |
| *** Nisha has quit IRC | 14:13 | |
| *** jrist has quit IRC | 14:14 | |
| *** jrist has joined #openstack-ironic | 14:14 | |
| jlvillal | sambetts, I hope that doesn't happen for us :) | 14:16 |
| jlvillal | sambetts, 40+ bugs is a lot to sit on... | 14:16 |
| * jlvillal wonders how many he is sitting on.... | 14:17 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Remove override-defaults https://review.openstack.org/255914 | 14:17 |
| jroll | mornin yall | 14:17 |
| jlvillal | Good morning jroll sambetts TheJulia and everyone else. I already said hello to lucas-afk and dtantsur|brb :) | 14:18 |
| *** mgoddard has joined #openstack-ironic | 14:18 | |
| TheJulia | Good morning | 14:19 |
| sinval | morning jlvillal TheJulia | 14:19 |
| jlvillal | sinval, :) | 14:20 |
| *** dprince_ has quit IRC | 14:21 | |
| *** dprince_ has joined #openstack-ironic | 14:21 | |
| *** [1]cdearborn has joined #openstack-ironic | 14:23 | |
| *** mgoddard has quit IRC | 14:25 | |
| *** dprince_ has quit IRC | 14:26 | |
| *** smoriya_ has quit IRC | 14:26 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector-specs: Ironic node auto-discovery https://review.openstack.org/252946 | 14:27 |
| *** robcresswell has joined #openstack-ironic | 14:28 | |
| *** deray has joined #openstack-ironic | 14:31 | |
| *** mgoddard has joined #openstack-ironic | 14:32 | |
| deray | hello all | 14:32 |
| deray | I am facing an issue with our CI builds.. | 14:33 |
| deray | failure details are as follows: | 14:33 |
| deray | File "tempest/scenario/manager.py", line 1252, in _node_state_timeout | 14:33 |
| deray | raise exceptions.TimeoutException(msg) | 14:33 |
| deray | tempest.exceptions.TimeoutException: Request timed out | 14:33 |
| deray | Details: Timed out waiting for node b6b3ffdb-d22e-429e-8543-cfb2e30d9e16 to reach power_state state(s) ['power on'] | 14:33 |
| deray | has anybody got any idea about how to resolve this? | 14:34 |
| jroll | deray: that sounds like slow or unreachable or broken ilo | 14:36 |
| jroll | I believe what that means is "ironic turned the power on, and the oob controller is not reporting power on after x seconds" | 14:36 |
| jroll | I would manually play with the ilo and see what happens | 14:36 |
| deray | jroll, I am able to power on and off the ilo from the oob tool. "manually play with the ilo and see what happens" - can you elaborate? | 14:37 |
| jroll | deray: that's what I meant, see if the ilo works manually :) | 14:38 |
| jroll | does the ilo report power on after you turn it on? | 14:38 |
| deray | jroll, yes | 14:39 |
| jroll | are you doing that from where the conductor is running? | 14:39 |
| jroll | conductor/devstack/whatever | 14:39 |
| *** Marga_ has quit IRC | 14:39 | |
| *** Marga_ has joined #openstack-ironic | 14:40 | |
| jroll | also, how long is the time from power on to the ilo reporting that power is on? | 14:40 |
| deray | jroll, I use this to power on and power off. https://github.com/openstack/proliantutils#usage | 14:40 |
| jroll | deray: right, are you doing that from where the conductor is running? I want to rule out a network problem | 14:41 |
| *** ig0r_ has joined #openstack-ironic | 14:41 | |
| deray | instantly if I do the power on () and chk () | 14:41 |
| deray | yes .. | 14:41 |
| jroll | hmmm | 14:42 |
| deray | let me try it out one more time | 14:42 |
| *** mgoddard has quit IRC | 14:42 | |
| mgould | deray, is that related to http://status.openstack.org/elastic-recheck/#1408067 ? | 14:42 |
| jroll | that seems different | 14:43 |
| jroll | deray: do you have the full traceback btw? | 14:43 |
| *** ig0r_ has quit IRC | 14:43 | |
| jroll | (and can you put it on paste.openstack.org) | 14:43 |
| deray | ys .. will do | 14:43 |
| jroll | thanks | 14:43 |
| NobodyCam | Good morning Ironicers... and.... | 14:44 |
| NobodyCam | TGIF! | 14:44 |
| jroll | :D hey NobodyCam | 14:44 |
| jroll | mgould: oh, maybe it is similar, during provisioning at least | 14:45 |
| jroll | I'm guessing this is where it comes from https://github.com/openstack/tempest/blob/master/tempest/scenario/manager.py#L1316 | 14:45 |
| NobodyCam | morning dtantsur|brb lucas-afk jroll jlvillal JayF rloo sinval sambetts vdrok and all others not directly listed here | 14:45 |
| sinval | hey NobodyCam | 14:45 |
| *** mkovacik has quit IRC | 14:45 | |
| NobodyCam | :) | 14:46 |
| jroll | NobodyCam: I love that you say good morning to people that aren't even connected ;D (rloo) | 14:46 |
| *** ndipanov has joined #openstack-ironic | 14:47 | |
| sambetts | o/ NobodyCam jlvillal jroll | 14:47 |
| deray | jroll, http://paste.openstack.org/show/481659/ | 14:48 |
| jroll | deray: so looking at the code more, it looks like your server isn't powering on to do the deploy | 14:48 |
| NobodyCam | hehehe | 14:48 |
| jroll | deray: full logs would be good too, if you have a link :) | 14:49 |
| deray | mgould, it seems similar | 14:49 |
| jroll | deray: or conductor logs, at least | 14:49 |
| NobodyCam | jroll: you've 255880 rechecking? | 14:50 |
| NobodyCam | you've got * | 14:50 |
| *** dansmith is now known as superdan | 14:53 | |
| mgould | NobodyCam, morning :-) | 14:53 |
| NobodyCam | morning mgould :) | 14:53 |
| deray | jroll, conductor log: http://paste.openstack.org/show/481660/ | 14:53 |
| deray | I dont have a public link to share with.. may be this one (this is of some subsequent build). https://15.184.134.36/20/256420/1/check/gate-tempest-ironic-ilo-driver-iscsi_ilo/5983b39/ | 14:55 |
| *** kozhukalov_ has quit IRC | 14:56 | |
| deray | jroll, mgould, sorry I woudl have to log off now. Can you drop me a mail: debayan.ray@gmail.com if you find anythign interesting? :) | 14:57 |
| *** kozhukalov_ has joined #openstack-ironic | 15:01 | |
| *** mgoddard has joined #openstack-ironic | 15:01 | |
| *** ig0r_ has joined #openstack-ironic | 15:02 | |
| *** deray has quit IRC | 15:02 | |
| *** mtanino has joined #openstack-ironic | 15:03 | |
| *** Marga_ has quit IRC | 15:03 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: [PROOF OF CONCEPT] Ansible deploy driver https://review.openstack.org/238183 | 15:04 |
| *** ig0r_ has quit IRC | 15:06 | |
| *** yonglihe has joined #openstack-ironic | 15:10 | |
| *** links has joined #openstack-ironic | 15:11 | |
| jroll | NobodyCam: yep, it's wrapping up now | 15:11 |
| devananda | morning, all | 15:12 |
| *** mbound_ has quit IRC | 15:12 | |
| jroll | ohai devananda | 15:13 |
| *** admin0 has quit IRC | 15:13 | |
| NobodyCam | good morning devananda :) | 15:13 |
| jroll | the job did pass \o/ | 15:13 |
| jroll | BAM https://review.openstack.org/#/c/255880/ | 15:14 |
| * jroll gets excited | 15:14 | |
| jroll | who wants to land that? | 15:14 |
| NobodyCam | jroll: BAM! | 15:22 |
| *** kkoski has joined #openstack-ironic | 15:23 | |
| jroll | NobodyCam: awesome, ty sir! | 15:23 |
| NobodyCam | :) now windows is forcing me to reboot.. so brb! | 15:23 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/256518 | 15:25 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/256519 | 15:25 |
| *** lucas-afk is now known as lucasagomes | 15:26 | |
| *** vishwanathj has quit IRC | 15:27 | |
| *** vishwanathj has joined #openstack-ironic | 15:27 | |
| lucasagomes | NobodyCam, jroll sinval deray morning | 15:31 |
| lucasagomes | jroll, a hurdle... it's easy to translate the properties to the get_node_by* methods | 15:31 |
| lucasagomes | but things like get_node_list that sucks... | 15:31 |
| NobodyCam | morning lucasagomes :) | 15:32 |
| *** dims has quit IRC | 15:32 | |
| jroll | lucasagomes: it sholdn't be too bad, but idk how you're doing it | 15:32 |
| jroll | if you're joining in the db layer and then assembling that in python | 15:32 |
| lucasagomes | jroll, no that's fine... but we are keeping the old column right? | 15:33 |
| lucasagomes | so I have properties which is the old column | 15:33 |
| lucasagomes | and properties_ which is the join | 15:33 |
| jroll | sure... | 15:37 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Correct inspection doc sudo reference https://review.openstack.org/256546 | 15:39 |
| *** admin0 has joined #openstack-ironic | 15:42 | |
| *** dprince has quit IRC | 15:43 | |
| *** dprince has joined #openstack-ironic | 15:43 | |
| *** david-ly_ has joined #openstack-ironic | 15:50 | |
| *** david-lyle has quit IRC | 15:53 | |
| *** david-ly_ is now known as david-lyle | 15:53 | |
| *** jaybeale has joined #openstack-ironic | 15:54 | |
| *** smoriya has joined #openstack-ironic | 15:56 | |
| *** jaybeale has quit IRC | 15:56 | |
| *** dprince has quit IRC | 16:05 | |
| JayF | would love a core IPA review on https://review.openstack.org/#/c/254896/ | 16:07 |
| *** dprince has joined #openstack-ironic | 16:08 | |
| *** links has quit IRC | 16:09 | |
| *** mgould has quit IRC | 16:10 | |
| *** kozhukalov_ has quit IRC | 16:15 | |
| *** ig0r_ has joined #openstack-ironic | 16:21 | |
| *** romcheg1 has joined #openstack-ironic | 16:21 | |
| *** jlvillal has quit IRC | 16:22 | |
| *** jlvillal has joined #openstack-ironic | 16:22 | |
| TheJulia | /win 7 | 16:22 |
| TheJulia | doh | 16:22 |
| *** jlvillal has quit IRC | 16:24 | |
| *** jlvillal has joined #openstack-ironic | 16:28 | |
| *** kozhukalov_ has joined #openstack-ironic | 16:28 | |
| *** admin0 has quit IRC | 16:29 | |
| *** jlvillal has quit IRC | 16:30 | |
| *** jlvillal has joined #openstack-ironic | 16:31 | |
| lucasagomes | JayF, lgtm, but I didn't test it yet | 16:42 |
| lucasagomes | JayF, you are runnning it already ? | 16:43 |
| JayF | lucasagomes: well, our downstream ramdisk is a little different than upstream, but we've been running in a chroot for a while | 16:43 |
| lucasagomes | ack | 16:43 |
| lucasagomes | JayF, boom +a | 16:43 |
| *** dtantsur|brb is now known as dtantsur | 16:44 | |
| JayF | https://review.openstack.org/#/c/189969/ could use an ipa core review too /me 1 for 1 this morning trying for two | 16:44 |
| JayF | lol | 16:44 |
| *** devananda has quit IRC | 16:45 | |
| dtantsur | morning NobodyCam, dev<TAB> | 16:45 |
| *** devananda has joined #openstack-ironic | 16:47 | |
| *** kozhukalov_ has quit IRC | 16:47 | |
| *** ig0r_ has quit IRC | 16:47 | |
| *** alex_xu has quit IRC | 16:53 | |
| dtantsur | jroll, I'd say pyudev upstream is not too friendly (again): https://bugzilla.redhat.com/show_bug.cgi?id=1288581 | 16:54 |
| openstack | bugzilla.redhat.com bug 1288581 in python-pyudev "Do NOT upgrade to pyudev 0.18" [Unspecified,New] - Assigned to dshea | 16:54 |
| dtantsur | maybe we should not rely on pyudev in long term | 16:54 |
| *** alex_xu has joined #openstack-ironic | 16:54 | |
| JayF | dtantsur: did you see my note yesterday? | 16:55 |
| jroll | dtantsur: what do we rely on it for? | 16:55 |
| JayF | dtantsur: pyudev is leaking in unit tests for IPA as well | 16:55 |
| JayF | dtantsur: test_get_os_install_device() is actually running the pyudev stuff locally, even though it's passing | 16:56 |
| dtantsur | JayF, yeah, that's another problem. But if it didn't leak, we would not notice the potential breakage ;) | 16:56 |
| dtantsur | jroll, it's used in IPA hardware manager to fetch additional data used e.g. for root device hints | 16:56 |
| JayF | I mean, sure, whatever, but I get really nervous when a rest API to erase a machine has leaky tests | 16:56 |
| JayF | lol | 16:56 |
| *** garthb has joined #openstack-ironic | 16:57 | |
| sambetts | haha | 16:57 |
| jroll | dtantsur: ah, I see | 16:57 |
| dtantsur | we don't have this code covered in gate FWIW, which is a separate problem | 16:57 |
| dtantsur | also the maintainer explicitly said that future breakages are possible and even expected, until they reach 1.0 | 16:58 |
| dtantsur | so I dunno... | 16:58 |
| *** smoriya is now known as smoriya_brb | 17:00 | |
| JayF | dtantsur: wow, that guy kinda made an ass out of himeslf in a public bug :/ | 17:00 |
| *** kozhukalov_ has joined #openstack-ironic | 17:00 | |
| openstackgerrit | Sam Betts proposed openstack/python-ironicclient: Updates supporting ironic-neutron integration https://review.openstack.org/206144 | 17:00 |
| JayF | dtantsur: like hard to use software maintained by someone who dgaf about breaking upstream users who are trying to help fix it :( | 17:00 |
| dtantsur | JayF, yeah, that was my impression as well | 17:01 |
| dtantsur | especially since fix would be extremely trivial and not too hacky: inherit your DeviceNotFound from EnvironmentError | 17:01 |
| jroll | I am so sick of this timeout bug in the gate | 17:02 |
| jroll | I'm not doing anything else until that's fixed | 17:02 |
| *** Marga_ has joined #openstack-ironic | 17:05 | |
| *** thrash is now known as thrash|f00dz | 17:05 | |
| dtantsur | jroll, yeah, it becomes worse with time | 17:05 |
| jroll | yep | 17:05 |
| jroll | and nobody wants to work on it :D | 17:06 |
| jroll | once we get devstack code in our tree it will be easier to work on | 17:06 |
| dtantsur | ++ for devstack code here. then you can experiment more more with various changes | 17:07 |
| *** jcoufal has quit IRC | 17:08 | |
| dtantsur | also, the ilo 3rd party CI has not worked for a while.. | 17:08 |
| dtantsur | oh, I see, it was just discussed today | 17:09 |
| *** sinval is now known as sinval-afk | 17:10 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Split the node's propeties column into its own table. https://review.openstack.org/252531 | 17:11 |
| lucasagomes | jroll, ^ if you have a time, it already migrates stuff | 17:12 |
| lucasagomes | on the fly, when properties is updated | 17:12 |
| lucasagomes | or for new nodes that is created it will use the new tables already | 17:12 |
| jroll | well, I guess it's not a timeout problem, this hasn't significantly changed the failure rate https://github.com/openstack-infra/devstack-gate/commit/0541b3e16feeaebeb6d3bb2c77457b4aeee9a643 | 17:12 |
| lucasagomes | doing more tests, but so far so good | 17:12 |
| jroll | lucasagomes: nice, I'll try to get a look | 17:12 |
| lucasagomes | thanks | 17:12 |
| dtantsur | jroll, sounds bad... so it might be a real failure? | 17:13 |
| jroll | dtantsur: well, the other theory was pxe failure, so.... probably | 17:14 |
| dtantsur | we need to fix out baremetal console logging, I guess | 17:14 |
| jroll | yeah | 17:14 |
| dtantsur | aka https://bugs.launchpad.net/devstack/+bug/1506829 | 17:14 |
| openstack | Launchpad bug 1506829 in devstack "Ironic bare metal console logs only show the last boot" [Undecided,New] - Assigned to Caio Oliveira (caiobo) | 17:14 |
| jroll | yep | 17:14 |
| *** ifarkas has quit IRC | 17:14 | |
| jroll | assigned for a month | 17:14 |
| * jroll removes | 17:14 | |
| jroll | or not, LP hates me | 17:15 |
| lucasagomes | ++ for fixing that | 17:16 |
| lucasagomes | I think caiobo was looking into it | 17:16 |
| dtantsur | ironic-lib patch https://review.openstack.org/184443 looks good to me, I suggest someone to have the final +2/+A :) | 17:18 |
| sambetts | jroll: inspector is now completely RFEs now btw :D | 17:18 |
| * dtantsur mentions it on the whiteboard | 17:18 | |
| sambetts | :) | 17:19 |
| *** dims has joined #openstack-ironic | 17:19 | |
| dtantsur | sambetts, oh, we have to fix a spec template and -1 all the specs :D wanna do that or should I? | 17:19 |
| jroll | sambetts: nice work | 17:19 |
| sambetts | dtantsur: Just about to do the patch for the spec template :-P | 17:20 |
| dtantsur | thanks a lot, you saved me a lot of time this Friday :) | 17:22 |
| sambetts | :) glad to be of service haha | 17:22 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector-specs: Update template to use RFE instead of BP https://review.openstack.org/256596 | 17:24 |
| dtantsur | jroll, so what do you think about specs that reference python-ironicclient or ironic-python-agent bugs? I have hanging -1 on my specs change due to that | 17:24 |
| jroll | dtantsur: make them add ironic to the bug and link like that | 17:25 |
| dtantsur | pas-ha, ^^ | 17:25 |
| jroll | lp redirects anyway | 17:25 |
| jroll | dtantsur: like this redirects to devstack bug https://bugs.launchpad.net/ironic/+bug/1506829 | 17:25 |
| openstack | Launchpad bug 1506829 in devstack "Ironic bare metal console logs only show the last boot" [Undecided,New] - Assigned to Caio Oliveira (caiobo) | 17:25 |
| jroll | if you click it just works | 17:25 |
| jroll | so I don't think it's a huge deal | 17:25 |
| dtantsur | wow, launchpad magic! | 17:26 |
| jroll | :) | 17:26 |
| * jroll grabs breakfast | 17:26 | |
| *** Nisha has joined #openstack-ironic | 17:26 | |
| *** praneshp has joined #openstack-ironic | 17:26 | |
| *** smoriya_brb is now known as smoriya | 17:27 | |
| *** penick has joined #openstack-ironic | 17:32 | |
| *** mgould has joined #openstack-ironic | 17:33 | |
| * sambetts is calling it an evening for today | 17:37 | |
| sambetts | night everyone, have a good weekend | 17:37 |
| openstackgerrit | Merged openstack/ironic-python-agent: Run IPA in chroot instead of container in CoreOS https://review.openstack.org/254896 | 17:37 |
| *** sambetts is now known as sambetts-afk | 17:38 | |
| dtantsur | g'night sambetts-afk | 17:38 |
| *** kozhukalov_ has quit IRC | 17:40 | |
| dtantsur | JayF, chroot change has landed \o/ | 17:40 |
| JayF | \o/ | 17:40 |
| lucasagomes | dtantsur, night! | 17:41 |
| dtantsur | lucasagomes, I didn't say I'm leaving, but that's a great idea :D | 17:41 |
| * dtantsur has some wine and cheese for tonight | 17:41 | |
| lucasagomes | oh sorry misread | 17:41 |
| lucasagomes | sambetts-afk, g'night! | 17:41 |
| * lucasagomes is concentrated on the db stuff, testing it | 17:41 | |
| dtantsur | jroll, when you're back please vote on https://review.openstack.org/#/c/255807/ | 17:48 |
| jroll | dtantsur: done | 17:51 |
| dtantsur | thnx | 17:52 |
| jroll | this is fun, we return the traceback in the api for 409? http://logs.openstack.org/93/255793/3/gate/gate-tempest-dsvm-ironic-agent_ssh/25175ed/logs/screen-n-cpu.txt.gz?level=TRACE#_2015-12-10_16_11_48_799 | 17:52 |
| *** kozhukalov_ has joined #openstack-ironic | 17:52 | |
| * jroll waiting for devstack to verify | 17:53 | |
| dtantsur | jroll, yes | 17:53 |
| lucasagomes | jroll, if debug is enabled I think we return the traceback yes | 17:53 |
| jroll | o.o | 17:53 |
| jroll | that seems wrong :/ | 17:54 |
| dtantsur | jroll, lucasagomes, which is something Nova hates use for :D | 17:54 |
| *** ndipanov has quit IRC | 17:54 | |
| jroll | right | 17:54 |
| dtantsur | did you guys see the bug from today? | 17:54 |
| jroll | mriedem brought this up to me | 17:54 |
| jroll | no? | 17:54 |
| lucasagomes | I feel like it's a feature | 17:54 |
| dtantsur | jroll, well, that's from mriedem, so you now know | 17:54 |
| jroll | I feel like it's a borderline vulnerability | 17:54 |
| jroll | heh | 17:54 |
| jroll | got a link to the bug? | 17:54 |
| lucasagomes | https://github.com/openstack/ironic/blob/master/ironic/api/hooks.py#L141-L144 | 17:54 |
| dtantsur | looking for it | 17:54 |
| openstackgerrit | Merged openstack/ironic-lib: Updated from global requirements https://review.openstack.org/256484 | 17:54 |
| jroll | lucasagomes: mehhhhh | 17:55 |
| lucasagomes | jroll, it was done by design | 17:55 |
| jroll | yeah, I see that | 17:55 |
| dtantsur | jroll, lucasagomes, https://bugs.launchpad.net/ironic/+bug/1525002 | 17:55 |
| openstack | Launchpad bug 1525002 in python-ironicclient "The ironic driver needs to scale back how many errors it traces out" [Low,In progress] - Assigned to Dmitry Tantsur (divius) | 17:55 |
| jroll | it's spammy as hell when we retry 409s though | 17:55 |
| dtantsur | lucasagomes, maybe, but it sucks. or we have to change ironicclient to strip it a way, which is also an option | 17:55 |
| lucasagomes | yeah | 17:56 |
| lucasagomes | that's true | 17:56 |
| jroll | I'd rather just not do it :| | 17:56 |
| jroll | or have a separate config | 17:56 |
| lucasagomes | couple of problems tangled there tho | 17:56 |
| dtantsur | yeah, I have a patch for ironicclient side of things | 17:56 |
| lucasagomes | our 409 is one of them, not everything that is 409 is retriable | 17:56 |
| dtantsur | but dunno what to do about traceback itself | 17:56 |
| jroll | dtantsur: I feel like tracebacks shouldn't be returned in production | 17:57 |
| jroll | and well, since everyone uses debug in prod | 17:57 |
| jroll | it's a problem | 17:57 |
| dtantsur | ++ | 17:57 |
| dtantsur | I'll fix it on Monday, if nobody takes it earlier | 17:57 |
| jroll | I'm doing it now | 17:57 |
| dtantsur | (please do, cause I'm pretty busy) | 17:57 |
| dtantsur | oh good :) | 17:57 |
| mgould | it sucks that people are using debug in prod, that's not what it should be for IMO | 17:58 |
| jroll | give us more time to bikeshed about the config name | 17:58 |
| *** ionutbalutoiu has quit IRC | 17:58 | |
| jroll | mgould: it does. but errors aren't debuggable without debug level in any openstack project | 17:58 |
| mgould | :-( | 17:58 |
| dtantsur | jroll, thanks :) also please take a look at related ironicclient patch | 17:58 |
| lucasagomes | jroll, we need a way to return tracebacks imo | 17:58 |
| *** spandhe has joined #openstack-ironic | 17:58 | |
| jroll | lucasagomes: I'm making a config for it. | 17:59 |
| dtantsur | https://review.openstack.org/256375 | 17:59 |
| lucasagomes | jroll, ack | 17:59 |
| dtantsur | lucasagomes, logs? ;) | 17:59 |
| lucasagomes | dtantsur, jroll [api]enable_serverside_traceback ? | 17:59 |
| lucasagomes | jroll, dtantsur actually btw, can we have a "troubleshoot" config option section | 17:59 |
| *** cdearborn has joined #openstack-ironic | 17:59 | |
| *** praneshp_ has joined #openstack-ironic | 17:59 | |
| dtantsur | my wild guess is that people who can't look at logs do not need traceback | 18:00 |
| jroll | dtantsur: ++ | 18:00 |
| lucasagomes | I feel we should have ways to run Ironic in troubleshoot mode | 18:00 |
| dtantsur | yeah, e.g. your change to not power the node off after failure.. I would LOVE it to land | 18:00 |
| *** penick has quit IRC | 18:00 | |
| lucasagomes | dtantsur, idk, I like to be able to deal with it interactively | 18:01 |
| lucasagomes | I don't see what we gain by not allowing it | 18:02 |
| lucasagomes | it's troubleshoot we have to make it easy | 18:02 |
| lucasagomes | openstack is already too complicated | 18:02 |
| dtantsur | ++ | 18:02 |
| jroll | I mean, we shouldn't need special configs to troubleshoot issues | 18:02 |
| * lucasagomes votes for config option | 18:02 | |
| jroll | issues are going to happen | 18:02 |
| dtantsur | inspector does not power off the node on error, which saved me a lot of time | 18:02 |
| jroll | and production apps need to be ready for it | 18:02 |
| lucasagomes | jroll, sometimes we do IMO | 18:02 |
| jroll | I don't want to restart my services to debug something | 18:02 |
| lucasagomes | there's one patch that me and dtantsur always talk about | 18:02 |
| lucasagomes | the one for not powering off the node in case of deploy failure | 18:03 |
| mgould | jroll, "sucks to be you, you should have had troubleshoot mode enabled" | 18:03 |
| jroll | mgould: ++ | 18:03 |
| lucasagomes | that's extremely useful for troubleshooting | 18:03 |
| dtantsur | jroll, that's fair, but we're talking about debugging ramdisks and situation when you need a machine with failed deployment to stay intact until we log in and take a look | 18:03 |
| jroll | lucasagomes: yeah, I think that should just always happen | 18:03 |
| *** praneshp has quit IRC | 18:03 | |
| *** praneshp_ is now known as praneshp | 18:03 | |
| lucasagomes | jroll, and for developing stuff too | 18:03 |
| lucasagomes | when you need to modify a ramdisk and it fails and power off, it's a PITA | 18:03 |
| jroll | lucasagomes: I agree it shouldn't power off on failure | 18:04 |
| lucasagomes | if I had one section for troubleshoot that are def not meant for production (and we should make it very clear) | 18:04 |
| * mgould points anyone who hasn't read it at http://www.drmaciver.com/2013/03/a-rewritten-manifesto-for-error-reporting/ | 18:04 | |
| lucasagomes | I think it makes total sense | 18:04 |
| jroll | lucasagomes: I just don't think it should be a config, I think that should always be true | 18:04 |
| jroll | lucasagomes: so then what do you do when a deploy fails in prod? | 18:04 |
| lucasagomes | jroll, right, that can be that way too | 18:04 |
| lucasagomes | but it's harder with intergrated systems | 18:04 |
| jroll | not sure what you mean | 18:05 |
| lucasagomes | e.g nova calls destroy() for a failed instance | 18:05 |
| lucasagomes | so ironic will power it off | 18:05 |
| jroll | oh, right | 18:05 |
| lucasagomes | because that's the same as one issuing an "deleted" in the api | 18:05 |
| lucasagomes | we don't have means to identify it | 18:05 |
| lucasagomes | so the config option, at least for some stuff makes sense | 18:05 |
| dtantsur | mgould, "You should not mask lower level errors" that's what we do a lot in ironic :) | 18:05 |
| *** penick has joined #openstack-ironic | 18:05 | |
| lucasagomes | means to idenfity whether it came from nova or not * | 18:06 |
| lucasagomes | mgould, thanks, will read that | 18:06 |
| *** harlowja has quit IRC | 18:06 | |
| lucasagomes | jroll, anyway, food for thought | 18:06 |
| mgould | dtantsur, yeah, it's not 100% applicable to every project, but I think there's a lot of sensible stuff in there | 18:06 |
| *** harlowja has joined #openstack-ironic | 18:07 | |
| mgould | the original rantier version is also a fun read :-) | 18:08 |
| lucasagomes | jrist, ironic_dev_mode=True | 18:08 |
| lucasagomes | jk hah | 18:08 |
| *** sinval-afk is now known as sinval | 18:08 | |
| * dtantsur renames jrist for his own sake | 18:09 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Don't return tracebacks in API response in debug mode https://review.openstack.org/256615 | 18:09 |
| jroll | ^ that should do it | 18:09 |
| *** thrash|f00dz is now known as thrash | 18:10 | |
| *** kozhukalov_ has quit IRC | 18:11 | |
| devananda | MattMan: is there anything which prevents IPA from deploying a Solaris golden image? | 18:13 |
| lucasagomes | jroll, thanks, it actually do not return for server errors | 18:13 |
| *** [1]cdearborn has quit IRC | 18:13 | |
| lucasagomes | so it's not any error in the api (unelss you remove that conditional) | 18:13 |
| *** aginwala has joined #openstack-ironic | 18:14 | |
| jroll | lucasagomes: oh, oops | 18:14 |
| dtantsur | mmm, I think we should not return traceback for HTTP 500 too | 18:15 |
| jroll | yeah I agree | 18:15 |
| jroll | blah | 18:15 |
| *** degorenko is now known as _degorenko|afk | 18:15 | |
| devananda | jroll: we should return tracebacks if debug is on | 18:15 |
| devananda | why would you propose to change that? | 18:15 |
| jroll | devananda: because people run prod in debug, and you shouldn't do that in prod | 18:15 |
| jroll | also it's spammy as heck | 18:15 |
| jroll | because 409 retry bs | 18:16 |
| devananda | ah | 18:16 |
| devananda | you're adding a new option to enable that, ok | 18:16 |
| *** lobaille has quit IRC | 18:16 | |
| jroll | yeah | 18:16 |
| jroll | https://bugs.launchpad.net/ironic/+bug/1525002 fwiw | 18:16 |
| openstack | Launchpad bug 1525002 in python-ironicclient "The ironic driver needs to scale back how many errors it traces out" [Low,In progress] - Assigned to Dmitry Tantsur (divius) | 18:16 |
| dtantsur | devananda, yeah, nova folks were not pleased with seeing up to 60 warnings with tracebacks every time we retry.. | 18:16 |
| devananda | yea... | 18:17 |
| lucasagomes | devananda, yeah new opt is fine | 18:17 |
| devananda | our gate logs are pretty ugly | 18:17 |
| lucasagomes | devananda, do you think we should have a "troubleshoot" config group? | 18:17 |
| devananda | jroll: now that i've read the patch, i'm fine with it :) | 18:17 |
| jroll | dtantsur: so for 500s, those are actually in a different key in the body, I think | 18:17 |
| * jroll tests | 18:18 | |
| devananda | lucasagomes: no, but perhaps a doc page on it | 18:18 |
| dtantsur | we have "troubleshooting" docs page now | 18:19 |
| dtantsur | we can add it there | 18:19 |
| lucasagomes | right, yeah cause I was wondering if having the group would be easier to identify options that should not be enabled in production | 18:19 |
| dtantsur | btw jroll, can your patch also update this documentation? | 18:19 |
| *** aginwala has quit IRC | 18:19 | |
| * dtantsur does not insist, but would be good | 18:19 | |
| jroll | dtantsur: actually, there's no traceback in 500s right now, I think | 18:19 |
| jroll | RESP BODY: {"error_message": "{\"debuginfo\": null, \"faultcode\": \"Server\", \"faultstring\": \"ohai\"}"} | 18:20 |
| jroll | not seeing this in 409 either, wth | 18:21 |
| dtantsur | jroll, how do you reproduce? this traceback is only added by oslo.messaging | 18:21 |
| dtantsur | jroll, so e.g. adding raises in API level is not the best reproducer | 18:21 |
| * dtantsur was confused like hell today | 18:21 | |
| lucasagomes | I will update the spec, should I open an RFE already? | 18:22 |
| * lucasagomes will do | 18:22 | |
| jroll | dtantsur: oh! | 18:22 |
| * jroll tries | 18:22 | |
| dtantsur | lucasagomes, RFE sounds good | 18:23 |
| dtantsur | now that we have a lightweight process for tracking features :) | 18:23 |
| lucasagomes | yeah | 18:23 |
| dtantsur | now it's a good idea to wish you a nice weekend | 18:24 |
| jroll | dtantsur: yeah, no traceback on 500 | 18:24 |
| dtantsur | see you all on Monday | 18:24 |
| dtantsur | mmm, I see | 18:24 |
| jroll | RESP BODY: {"error_message": "{\"debuginfo\":null,\"faultcode\":\"Server\",\"faultstring\":\"ohai\"}"} | 18:25 |
| jroll | have a good weekend \o | 18:25 |
| *** dtantsur is now known as dtantsur|afk | 18:25 | |
| lucasagomes | dtantsur|afk, enjoy the weekend | 18:26 |
| lucasagomes | jroll, that's because that hook is removing it? | 18:26 |
| lucasagomes | or did you remove that conditional | 18:26 |
| jroll | lucasagomes: yeah, I meant current behavior | 18:27 |
| jroll | I guess I should remove that, I was confused | 18:27 |
| *** e0ne has quit IRC | 18:27 | |
| *** ijw has joined #openstack-ironic | 18:28 | |
| * jroll fixes and also generates sample config and moar docs | 18:29 | |
| *** ijw has quit IRC | 18:29 | |
| *** ijw has joined #openstack-ironic | 18:29 | |
| maratoid | Hello! I'm getting an image writing problem with ironic: http://paste.openstack.org/show/481696/ | 18:31 |
| maratoid | This is with an image i built manually : DIB_RELEASE=trusty disk-image-create -o /httpboot/deployment_image.qcow2 ubuntu simple-init | 18:31 |
| * TheJulia looks | 18:32 | |
| *** alexpilotti has quit IRC | 18:33 | |
| TheJulia | maratoid: try adding vm to the element list | 18:33 |
| *** alexpilotti has joined #openstack-ironic | 18:33 | |
| TheJulia | There should likely be a thing in IPA to toss an error of "no partition table found", since I bet that is what is happening | 18:34 |
| lucasagomes | TheJulia, ++ I think IPA doesn't work with partition images yet | 18:34 |
| lucasagomes | s/I think// | 18:35 |
| TheJulia | it does not | 18:35 |
| TheJulia | and nothing along the way truly validates the contents of the image as far as I remember | 18:36 |
| * maratoid googles for 'adding cm to element list' furiously | 18:36 | |
| maratoid | vm. dangit. vm. | 18:36 |
| TheJulia | maratoid: one moment | 18:36 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Don't return tracebacks in API response in debug mode https://review.openstack.org/256615 | 18:36 |
| TheJulia | https://github.com/openstack/diskimage-builder/tree/master/elements/vm <-- worst named disk image builder element ever | 18:37 |
| jroll | lucasagomes: ^ tested, doc'd, fixed the 5xx thing | 18:37 |
| lucasagomes | jroll, ta much | 18:37 |
| * lucasagomes looks | 18:37 | |
| jroll | np, ty! | 18:37 |
| *** alexpilotti has quit IRC | 18:37 | |
| maratoid | so 'DIB_RELEASE=trusty disk-image-create -o /httpboot/deployment_image.qcow2 ubuntu simple-init vm' ? | 18:39 |
| TheJulia | maratoid: yes | 18:39 |
| maratoid | WIll do. Thanks! | 18:39 |
| TheJulia | maratoid: no problem | 18:40 |
| TheJulia | maratoid: let us know if it works... or not :) | 18:40 |
| maratoid | sure thing. | 18:41 |
| TheJulia | maratoid: by the way, w/r/t serial:1 in the playbooks and waiting for node deployment, the code for the intermediate shade library merged today, but hast not been released yet. Once that releases, I'll allow the update to bifrost that can leverage that functionality to land. If you want to manually patch stuff. I can give you the revision numbers :) | 18:44 |
| *** shakamunyi has quit IRC | 18:44 | |
| *** Marga_ has quit IRC | 18:44 | |
| *** jistr has quit IRC | 18:45 | |
| maratoid | Thanks, but I think I'd rather get to some form of things deploying semi-reliably first. Don't wanna add any more opportunities for myself to mess up quite yet :) | 18:47 |
| *** toabctl has quit IRC | 18:47 | |
| TheJulia | maratoid: Okay, well, when the time comes, feel free to ping me and I'll help you out with those | 18:47 |
| maratoid | thanks! | 18:47 |
| maratoid | looks like adding the vm element worked. not seeing any more of those errors. | 18:48 |
| *** alexpilotti has joined #openstack-ironic | 18:49 | |
| TheJulia | Awesome | 18:50 |
| maratoid | Sigh. All nodes are now "active" but i still can't ssh to them. | 18:50 |
| *** Sukhdev has joined #openstack-ironic | 18:51 | |
| *** toabctl has joined #openstack-ironic | 18:51 | |
| TheJulia | maratoid: so give them a few minutes | 18:51 |
| TheJulia | maratoid: do you have a display attached to one? | 18:51 |
| maratoid | yep | 18:51 |
| TheJulia | Is the machine at a login prompt? | 18:51 |
| *** athomas has quit IRC | 18:53 | |
| maratoid | interesting | 18:54 |
| maratoid | seems like they booted into netwrok boot mode, which fialed. | 18:54 |
| TheJulia | devananda: did you run into something like ^^^? I remember somebody mentioning such... :( | 18:55 |
| *** dims has quit IRC | 18:56 | |
| devananda | maratoid: try running these commands against the node you're watching | 18:58 |
| devananda | ironic node-set-boot-device <NAME> disk | 18:58 |
| devananda | ironic node-set-power-state <NAME> reboot | 18:58 |
| maratoid | ok | 18:58 |
| devananda | I ran into an issue testing the pxe_amt driver where it wrote the MBR, cleaned up the tftp config, rebooted the node -- but didn't change the boot device | 18:58 |
| *** mgoddard_ has joined #openstack-ironic | 18:59 | |
| *** mgoddard has quit IRC | 18:59 | |
| *** jdk has joined #openstack-ironic | 19:02 | |
| *** athomas has joined #openstack-ironic | 19:03 | |
| *** yuanying has quit IRC | 19:03 | |
| openstackgerrit | Satoru Moriya proposed openstack/ironic-specs: Volume connection information for Ironic nodes https://review.openstack.org/200496 | 19:04 |
| jdk | hi, I am using bifrost/ironic to install ubuntu 15.10, I built a custom disk image with disk image builder. not sure if I got all the right elements. what is the preferred way to set the hostname? | 19:05 |
| mgould | goodnight, everyone, and have a good weekend! | 19:06 |
| *** mgould has quit IRC | 19:06 | |
| *** smoriya has quit IRC | 19:09 | |
| NobodyCam | night mgould | 19:10 |
| *** e0ne has joined #openstack-ironic | 19:11 | |
| *** Marga_ has joined #openstack-ironic | 19:14 | |
| jroll | lucasagomes: hm, on the topic of the timeout bug in the gate | 19:15 |
| lucasagomes | jroll, shoot | 19:15 |
| jroll | what if we made everything ipxe, and had one non-voting pxe job | 19:15 |
| jroll | ipxe seems much more stable I think | 19:15 |
| openstackgerrit | Merged openstack/ironic: Add devstack plugin https://review.openstack.org/255880 | 19:16 |
| jroll | yesssss | 19:16 |
| NobodyCam | :) | 19:16 |
| lucasagomes | jroll, I took a quick look in the logs on tuesday after we talked about it | 19:16 |
| lucasagomes | I believe that there's a problem with using TFTP | 19:17 |
| lucasagomes | the node took more than 6 minutes to boot | 19:17 |
| lucasagomes | so I think iPXE is the way to go, let's make it use iPXE and check if it will solves the timeout problem | 19:17 |
| jroll | lucasagomes: yeah, your comment is what I was thinking of | 19:17 |
| lucasagomes | or at least minimize it | 19:17 |
| jroll | we'll still want to make sure we test pxe | 19:18 |
| jroll | but we can leave it nv | 19:18 |
| lucasagomes | jroll, yeah, that's fine | 19:18 |
| jroll | so it doesn't kill the gate | 19:18 |
| lucasagomes | ++ true | 19:18 |
| jroll | alright, I'll think about it | 19:18 |
| lucasagomes | jroll, we can compare pxe_ipa with pxe_ipa-ipxe jobs | 19:18 |
| jroll | going to check out early today | 19:18 |
| lucasagomes | about stability | 19:18 |
| jroll | right | 19:18 |
| lucasagomes | cause the are same, only diff is one is ipxe the other is not | 19:18 |
| jroll | definitely going to do that | 19:18 |
| NobodyCam | holiday shopping jroll ??? | 19:19 |
| lucasagomes | jroll, thanks! | 19:19 |
| jroll | NobodyCam: nah, it's friday! | 19:19 |
| jroll | (plus birthday weekend so why not) | 19:19 |
| NobodyCam | lol +++++ | 19:19 |
| *** krtaylor has joined #openstack-ironic | 19:19 | |
| jroll | thank you lucasagomes | 19:19 |
| NobodyCam | oh wow... | 19:19 |
| NobodyCam | Happy BirthDay man | 19:19 |
| jroll | thanks :) | 19:19 |
| lucasagomes | jroll, oh it's ur bd? Happy birthday! | 19:20 |
| jroll | thanks! lol | 19:20 |
| jroll | should probably wait for monday to turn on the devstack plugin for all jobs | 19:21 |
| jroll | agree/disagree? ^ | 19:21 |
| jroll | I don't want to risk breaking the gate on a friday | 19:21 |
| NobodyCam | agree here | 19:21 |
| NobodyCam | ++++ | 19:22 |
| * jroll wips | 19:22 | |
| *** sinh has quit IRC | 19:22 | |
| * NobodyCam has had to many broken gate fridays | 19:22 | |
| NobodyCam | lol | 19:22 |
| lucasagomes | jroll, yeah leave it for monday indeed | 19:23 |
| lucasagomes | it's almost end of the day for me | 19:23 |
| lucasagomes | and you should enjoy ur bd weekend | 19:23 |
| jroll | mhm | 19:24 |
| *** dprince has quit IRC | 19:24 | |
| *** e0ne_ has joined #openstack-ironic | 19:26 | |
| jdk | SpamapS: are you there? | 19:26 |
| jdk | or TheJulia? | 19:26 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Add indexable JSON fields spec https://review.openstack.org/253605 | 19:28 |
| lucasagomes | jroll, devananda ^ if you guys have some time | 19:28 |
| *** e0ne has quit IRC | 19:28 | |
| lucasagomes | there's a patch up implementing it for propperties already, needs unittests but I've tested it locally here and seems good: https://review.openstack.org/#/c/252531/ | 19:28 |
| *** garthb_ has joined #openstack-ironic | 19:28 | |
| jroll | lucasagomes: I'll come back to it, gonna get out of here | 19:29 |
| lucasagomes | no worries | 19:29 |
| jroll | thanks for doing that, I'll be sure to review next week | 19:29 |
| NobodyCam | have a great weekend jroll | 19:29 |
| jroll | have a good weekend! | 19:29 |
| NobodyCam | lucasagomes: if I click on that will it say add mongodb backend? | 19:29 |
| NobodyCam | lol | 19:30 |
| * NobodyCam *ducks* | 19:30 | |
| lucasagomes | jroll, np! Feel free to test that patch too, apply it, run "ironic-dbsync --config-file=/etc/ironic/ironic.conf upgrade" and it all should work | 19:30 |
| lucasagomes | you can see the migration working when updating the properties | 19:30 |
| lucasagomes | and new nodes will use the new tables already | 19:30 |
| lucasagomes | NobodyCam, hah no | 19:30 |
| *** liliars-afk has quit IRC | 19:30 | |
| *** sinval has quit IRC | 19:30 | |
| lucasagomes | NobodyCam, couchdb | 19:30 |
| lucasagomes | jk | 19:30 |
| NobodyCam | lol | 19:30 |
| NobodyCam | +++ | 19:30 |
| *** garthb has quit IRC | 19:30 | |
| lucasagomes | hah | 19:30 |
| *** raildo is now known as raildo-afk | 19:30 | |
| *** raildo-afk is now known as raildo | 19:30 | |
| *** liliars-afk has joined #openstack-ironic | 19:31 | |
| *** sinval has joined #openstack-ironic | 19:31 | |
| TheJulia | jdk: I'm here | 19:32 |
| jdk | hi TheJulia, so I got a custom image deployed but the hostname didn't get set. how is that supposed to work? | 19:33 |
| devananda | jroll: o you want to do anything to prevent changes landing in devstack that should be going to ironic now (beyond your -1's) ? | 19:34 |
| TheJulia | jdk: through metadata in the config drive or the metadata service, depending on your configuration and infrastructure | 19:34 |
| lucasagomes | devananda, he sent a message to the ML about it too | 19:37 |
| * lucasagomes needs to move some of my changes to ironic now | 19:37 | |
| jdk | TheJulia: I'd like to use the config drive for now. What element(s) must I include in disk image builder? | 19:37 |
| lucasagomes | devananda, but would be good to get a -2 since the code is now merged in our tree | 19:37 |
| lucasagomes | IMO | 19:37 |
| devananda | lucasagomes: saw that. just thinking of folks who may not see the ML, might upload another change that erases the -1 | 19:37 |
| lucasagomes | devananda, yeah totally, it's a bit weak | 19:37 |
| devananda | exactly. -2's on those should be done now IMO | 19:37 |
| lucasagomes | devananda, if we abandon the change and someone tries to put a new patch set | 19:38 |
| lucasagomes | will it error out? Or un-abandon it? | 19:38 |
| devananda | un-abandon | 19:38 |
| lucasagomes | damn | 19:38 |
| lucasagomes | ok | 19:38 |
| lucasagomes | yeah -2 is it | 19:38 |
| devananda | also, I dont think we can abandon devstack changes | 19:38 |
| lucasagomes | fair enough | 19:38 |
| devananda | but if there's a list of them, we can ask a devstack core to -2 them | 19:39 |
| devananda | *the list | 19:39 |
| TheJulia | jdk: config drives can be consumed by cloud init or simple-init. What is the source that is generating the configuration drives? | 19:39 |
| * devananda gets ready to walk into the office | 19:39 | |
| *** kozhukalov_ has joined #openstack-ironic | 19:39 | |
| *** harlowja has quit IRC | 19:41 | |
| *** harlowja has joined #openstack-ironic | 19:41 | |
| lucasagomes | devananda, http://paste.openstack.org/show/481704/ | 19:42 |
| lucasagomes | those are the ones I was notified o | 19:42 |
| lucasagomes | of* | 19:42 |
| jdk | TheJulia: nodes are standalone right now. I see bifrost creating a config drive. do i use the cloud-init-datasources element with DIB_CLOUD_INIT_DATASOURCES="ConfigDrive" | 19:43 |
| *** pas-ha has quit IRC | 19:46 | |
| devananda | lucasagomes: thanks. I'll check during lunch, if you or jroll don't get to it first | 19:47 |
| lucasagomes | devananda, ack, I probably won't cause I'm almost calling it a day | 19:47 |
| devananda | lucasagomes: also, it's gotta be late for you -- enjoy your friday night :) | 19:47 |
| lucasagomes | yeah | 19:47 |
| devananda | yea, have a good weekend! | 19:48 |
| lucasagomes | devananda, thanks a lot! Enjoy the weekend over there too | 19:48 |
| NobodyCam | yes lucasagomes have a great weekend | 19:48 |
| lucasagomes | NobodyCam, thanks! You too, enjoy it | 19:48 |
| *** shakamunyi has joined #openstack-ironic | 19:49 | |
| *** aginwala has joined #openstack-ironic | 19:50 | |
| TheJulia | jdk: in that case, use simple-init, it will turn off cloud-init since bifrost writes the metadata it needs for network configuration | 19:51 |
| lucasagomes | ok folks, have a good night! | 19:53 |
| *** baoli_ has quit IRC | 19:53 | |
| lucasagomes | enjoy the weekend | 19:53 |
| *** lucasagomes is now known as lucas-beer | 19:53 | |
| *** baoli has joined #openstack-ironic | 19:53 | |
| *** e0ne_ has quit IRC | 19:54 | |
| *** shakamunyi has quit IRC | 19:55 | |
| *** Marga_ has quit IRC | 19:57 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/191505 | 20:02 |
| sinval | have a great weekend lucas-beer | 20:02 |
| *** penick has quit IRC | 20:02 | |
| *** mgoddard_ has quit IRC | 20:04 | |
| JayF | jroll: if you do have an objection to https://review.openstack.org/#/c/191505 I suggest you put it on the patchset sooner rather than later. It will have my +2 as soon as it passes tests. I know you had been concerned about upstream coreos being able to break IPA's gate. | 20:04 |
| *** mgoddard has joined #openstack-ironic | 20:05 | |
| jdk | TheJulia: thanks , I will try that | 20:08 |
| *** romcheg1 has quit IRC | 20:09 | |
| jdk | TheJulia: I should also mention that I had to create a custom element for network configuration because bonding and bridging is involved | 20:09 |
| jdk | TheJulia: I am using dhcp as well | 20:10 |
| *** ionutbalutoiu has joined #openstack-ironic | 20:10 | |
| TheJulia | jdk: Interesting. I'm trying to remember if the network metadata spec addresses that. I think simple-init does not support that completely yet :( | 20:10 |
| *** romcheg1 has joined #openstack-ironic | 20:11 | |
| jdk | TheJulia: can I still use simple-init for the hostname? | 20:12 |
| openstackgerrit | Merged openstack/ironic: Fix iPXE template for whole disk image https://review.openstack.org/255362 | 20:13 |
| TheJulia | jdk: absolutely. If a mac address matches the config info it gets, it will still write a /etc/network/interfaces/interfaces.d or /etc/sysconfig/network-scripts/ifcfg-blah file though, but we do have support for just not passing that through and allowing the node to rely upon dhcp | 20:13 |
| *** aginwala has quit IRC | 20:14 | |
| *** aginwala has joined #openstack-ironic | 20:14 | |
| *** penick has joined #openstack-ironic | 20:17 | |
| openstackgerrit | Ondřej Nový proposed openstack/ironic: Deprecated tox -downloadcache option removed https://review.openstack.org/256686 | 20:17 |
| openstackgerrit | Ondřej Nový proposed openstack/ironic-python-agent: Deprecated tox -downloadcache option removed https://review.openstack.org/256687 | 20:18 |
| TheJulia | jdk: for what its worth, some may be interested with what you've come up with element wise, so if you can, I don't think there would be much objection for it to land if it has a general use case that makes sense to the rest of the world :) | 20:20 |
| *** rajathagasthya has joined #openstack-ironic | 20:22 | |
| jdk | TheJulia: I can probably do that but I will have to separate out some other non-networking stuff that I included in it and make it more general. | 20:23 |
| jdk | TheJulia: i would be interested in how to stop passing through the network config and allow the node to rely on dhcp | 20:24 |
| jdk | TheJulia: do i need to modify the bifrost-configdrives-dynamic role | 20:27 |
| *** vishwanathj has quit IRC | 20:27 | |
| TheJulia | jdk: don't define an ipv4_address I think | 20:28 |
| *** raildo is now known as raildo-afk | 20:28 | |
| *** vishwanathj has joined #openstack-ironic | 20:28 | |
| jdk | TheJulia: ok I will try that, thanks | 20:29 |
| TheJulia | jdk: addressing_mode needs to set to "dhcp" | 20:30 |
| *** ig0r_ has joined #openstack-ironic | 20:37 | |
| openstackgerrit | Merged openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/254495 | 20:39 |
| *** e0ne has joined #openstack-ironic | 20:44 | |
| *** Sukhdev has quit IRC | 20:48 | |
| *** aginwala has quit IRC | 20:48 | |
| *** ChubYann has joined #openstack-ironic | 20:51 | |
| devananda | jroll: where are we w.r.t. the stable/liberty break? I'm looking over the bug and patches and trying to piece it together, but you may have more context | 20:52 |
| JayF | devananda: I believe he's gone for the day already :( | 20:53 |
| *** Marga_ has joined #openstack-ironic | 20:55 | |
| *** penick has quit IRC | 20:55 | |
| *** rcernin has joined #openstack-ironic | 20:56 | |
| *** ionutbalutoiu has quit IRC | 20:58 | |
| *** ionutbalutoiu has joined #openstack-ironic | 20:59 | |
| *** penick has joined #openstack-ironic | 20:59 | |
| *** baoli has quit IRC | 21:00 | |
| *** Marga_ has quit IRC | 21:02 | |
| jroll | <devananda:#openstack-ironic> jroll: o you want to do anything to prevent changes landing in devstack that should be going to ironic now (beyond your -1's) ? <- is there something I *can* do? | 21:02 |
| jroll | <devananda:#openstack-ironic> jroll: where are we w.r.t. the stable/liberty break? I'm looking over the bug and patches and trying to piece it together, but you may have more context <- the bad o.msg version is blacklisted, but the next release will break again unless this lands: https://review.openstack.org/#/c/254904/ . been trying to catch up with dims and not much luck | 21:03 |
| *** aginwala has joined #openstack-ironic | 21:04 | |
| NobodyCam | jroll: your still here? | 21:04 |
| JayF | a wild jroll appears | 21:04 |
| *** jdk has quit IRC | 21:04 | |
| JayF | I thought you were blowing out candles and eating cake | 21:04 |
| jroll | shhhhh | 21:04 |
| NobodyCam | lol | 21:04 |
| jroll | I sat down to fill my vape and may have looked at irc >.> | 21:04 |
| JayF | this is why I close IRC when not working, lol | 21:05 |
| NobodyCam | lol | 21:05 |
| jroll | well, I did it intentionally, to be fair | 21:05 |
| JayF | NobodyCam: did you and John ever get the stickers I sent over? | 21:05 |
| NobodyCam | have to check mail today | 21:05 |
| NobodyCam | ourmail here can be slow | 21:06 |
| JayF | zer0c00l: jfyi; chroot change merged. I rebased your coreos stable change on top of it, and if it passes tests, I'll +2 it | 21:07 |
| JayF | zer0c00l: \o/ very glad to have more folks working on the ramdisk, and please pping me if you're doing anything else like this now that I have time to help some | 21:07 |
| *** penick has quit IRC | 21:13 | |
| *** kkoski has quit IRC | 21:15 | |
| *** mjturek1 has quit IRC | 21:18 | |
| *** penick has joined #openstack-ironic | 21:19 | |
| *** Marga_ has joined #openstack-ironic | 21:21 | |
| zer0c00l | JayF: :) | 21:22 |
| zer0c00l | JayF: Thanks | 21:22 |
| zer0c00l | JayF: I have plans to work on this feature https://review.openstack.org/#/c/248968/ | 21:23 |
| zer0c00l | The spec needs some review :P | 21:23 |
| JayF | I'll review it right now | 21:25 |
| *** krtaylor has quit IRC | 21:25 | |
| zer0c00l | Appreciate it Thanks | 21:25 |
| JayF | zer0c00l: one thing though | 21:25 |
| JayF | zer0c00l: it depends on a backlog spec that nobody is working on | 21:26 |
| JayF | zer0c00l: so not sure it adds value at all to get that spec merged without the backlog spec for agent partition image support being worked on | 21:26 |
| zer0c00l | JayF: I don't think it depends on the partition image support. May be the wording in the spec makes it look like it depends on partition image support | 21:28 |
| zer0c00l | I don't see why we can't implement tar-payload without having to depend on partition image support | 21:29 |
| JayF | zer0c00l: I do: IPA doesn't know how to do any partitioning | 21:30 |
| JayF | zer0c00l: at least until all the ironic-lib stuff gets landed and that partition image support spec goes in | 21:30 |
| zer0c00l | JayF: I thought the disk_partitioner code could do what we want to | 21:31 |
| zer0c00l | It is part of ironic-lib? | 21:31 |
| JayF | zer0c00l: all images deployed by IPA today are done either as a full disk image or using iscsi deploy method (where conductor handles teh partition/imaging) | 21:31 |
| JayF | ironic-lib hasn't been integrated into IPA ytet | 21:31 |
| JayF | so it's in line, partiiton image support will get done | 21:31 |
| JayF | I just don't think the spec you proposed can be started on until at least the partition image one is possibly done | 21:31 |
| JayF | and like I said, AFAIK nobody is actively working on that. I don't know everything though :P | 21:31 |
| *** mgoddard has quit IRC | 21:32 | |
| zer0c00l | JayF: So if i work to include ironic-lib in IPA first and then start working on tar-payload | 21:32 |
| zer0c00l | Will it work? | 21:32 |
| JayF | zer0c00l: potentially, sure, but in the process you'll be implementing like 95% of stuff needed for parition image support | 21:33 |
| zer0c00l | I agree! | 21:33 |
| JayF | which would make me, at least, want to just implement that spec instead of just the few bits needed for the tar piece | 21:33 |
| zer0c00l | Yes implementing partition image support from there is easy | 21:33 |
| zer0c00l | okay. So you are saying we should get the partition image support in first and then tar-payload? | 21:34 |
| zer0c00l | i emailed the spec owner long back | 21:34 |
| zer0c00l | I haven't recieved any response yet | 21:35 |
| JayF | Generally speaking with things liek that we're a do-ocracy | 21:37 |
| *** baoli has joined #openstack-ironic | 21:37 | |
| JayF | meaning if the person abandoned that for the most part, just pick it up and do it, esepcially if they aren't responding to your messages | 21:37 |
| *** krtaylor has joined #openstack-ironic | 21:37 | |
| zer0c00l | JayF: Okay! | 21:38 |
| zer0c00l | JayF: brb, lunch time here | 21:38 |
| JayF | zer0c00l: given it's late on a friday | 21:38 |
| JayF | zer0c00l: if you can come to ironic meeting on Friday morning (PST, IDK what timezone you're in, but gotta be close if it's lunch for you) | 21:39 |
| JayF | s/Friday/Monday/ | 21:39 |
| zer0c00l | I am on PST | 21:39 |
| JayF | zer0c00l: announce your intentions to pick that up and you'll probably get a lot of \o/ and help in response | 21:39 |
| JayF | zer0c00l: in NorCal by chance? | 21:39 |
| zer0c00l | Yes | 21:39 |
| JayF | if you're in/near SF we could grab lunch sometime | 21:39 |
| JayF | I work from home most days but don't mind trekking into the soma office if I have plans | 21:40 |
| zer0c00l | JayF: Awesome! I work from sunnyvale | 21:40 |
| zer0c00l | You could come and have lunch with us at Yahoo | 21:40 |
| zer0c00l | meet the team! | 21:40 |
| JayF | zer0c00l: I have eaten at yahoo before when the mid-cycle was there. Delicious delicious food in ridiculous quantities. | 21:41 |
| zer0c00l | :D | 21:41 |
| zer0c00l | :) | 21:41 |
| JayF | I'm in SJ a lot b/c I'm a huge sharks fan | 21:41 |
| zer0c00l | hahaha it is sharks territory | 21:41 |
| JayF | maybe one day I can find a game to go to and catch lunch with you, work from coffee shop, then go to the game | 21:42 |
| zer0c00l | sure! | 21:42 |
| JayF | if I fight that traffic to go south, there has to be some kind of other reason to do it :P | 21:42 |
| JayF | lol | 21:42 |
| zer0c00l | just drop me an email | 21:42 |
| zer0c00l | or message | 21:42 |
| JayF | sure; probably after the new year but absolutely | 21:42 |
| zer0c00l | cool! | 21:43 |
| zer0c00l | brb | 21:43 |
| *** aginwala has quit IRC | 21:45 | |
| *** e0ne has quit IRC | 21:47 | |
| *** e0ne has joined #openstack-ironic | 21:49 | |
| *** e0ne has quit IRC | 21:49 | |
| *** Marga_ has quit IRC | 21:52 | |
| *** mkovacik has joined #openstack-ironic | 21:52 | |
| *** aginwala has joined #openstack-ironic | 21:56 | |
| *** Marga_ has joined #openstack-ironic | 21:57 | |
| *** Nisha has quit IRC | 21:58 | |
| *** Marga_ has quit IRC | 22:01 | |
| *** Sukhdev has joined #openstack-ironic | 22:08 | |
| *** Marga_ has joined #openstack-ironic | 22:16 | |
| davidlenwell | poseur | 22:18 |
| openstackgerrit | Ondřej Nový proposed openstack/pyghmi: Deprecated tox -downloadcache option removed https://review.openstack.org/256807 | 22:20 |
| *** aginwala has quit IRC | 22:20 | |
| NobodyCam | just had a poweroutage and our router lost all of its settings... :( | 22:25 |
| *** thrash is now known as thrash|wknd | 22:29 | |
| *** aginwala has joined #openstack-ironic | 22:30 | |
| *** baoli has quit IRC | 22:31 | |
| *** penick has quit IRC | 22:36 | |
| *** openstackgerrit has quit IRC | 22:37 | |
| *** rcernin has quit IRC | 22:37 | |
| *** openstackgerrit has joined #openstack-ironic | 22:37 | |
| *** cdearborn has quit IRC | 22:39 | |
| devananda | NobodyCam: ugh :( | 22:44 |
| NobodyCam | ya | 22:44 |
| NobodyCam | how are you today mr devananda | 22:44 |
| devananda | NobodyCam: it's afternoon coffee time :) | 22:44 |
| * BadCub appreciates afternoon coffee time | 22:45 | |
| NobodyCam | nice ... I have my starbucks in front of me too :) | 22:45 |
| *** shakamunyi has joined #openstack-ironic | 22:51 | |
| *** mkovacik has quit IRC | 22:54 | |
| *** dims has joined #openstack-ironic | 22:55 | |
| *** rajathagasthya has quit IRC | 22:56 | |
| *** shakamunyi has quit IRC | 23:00 | |
| *** dims has quit IRC | 23:00 | |
| *** dims has joined #openstack-ironic | 23:01 | |
| *** dims has quit IRC | 23:04 | |
| *** ijw has quit IRC | 23:06 | |
| *** dims has joined #openstack-ironic | 23:12 | |
| *** baoli has joined #openstack-ironic | 23:22 | |
| *** jcoufal has joined #openstack-ironic | 23:23 | |
| mordred | TheJulia: thanks for the feedback on https://review.openstack.org/#/c/254399/ - I updated it - if you get bored, try again? | 23:26 |
| *** shakamunyi has joined #openstack-ironic | 23:26 | |
| *** baoli has quit IRC | 23:27 | |
| *** kozhukalov_ has quit IRC | 23:27 | |
| NobodyCam | hey hey mordred happy Friday to you! | 23:29 |
| mordred | same to you NobodyCam ! | 23:29 |
| NobodyCam | :) | 23:29 |
| zer0c00l | ping JayF | 23:30 |
| zer0c00l | JayF: You were suggesting that we should be using instance_info to store part_data | 23:31 |
| zer0c00l | Does that mean we add code to nova to recognize instance_info/part_data ? | 23:31 |
| *** mkovacik has joined #openstack-ironic | 23:31 | |
| zer0c00l | or we could reuse nova metadata and pass part_data as one of its keys | 23:32 |
| *** shakamunyi has quit IRC | 23:32 | |
| *** mkovacik has quit IRC | 23:39 | |
| *** baoli has joined #openstack-ironic | 23:44 | |
| JayF | zer0c00l: if you look at how image information is handled today | 23:50 |
| JayF | zer0c00l: nova populates it in instance_info before building the instance | 23:51 |
| JayF | I see a similar pattern working for tar images -- you'd put the "how I want these image pieces arranged" in instance_info for each node on build | 23:51 |
| JayF | or not, it's just an idea :) I really don't like overloading the raid details though | 23:52 |
| *** baoli has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!