| *** yolanda has quit IRC | 00:48 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: bubblewrap: adds --die-with-parent option https://review.openstack.org/473164 | 00:59 |
|---|---|---|
| mordred | tristanC: ^^ is that supposed to be "sleep && disown" rather than "sleep & disown" ? | 02:53 |
| mordred | or is the & correct? | 02:53 |
| mordred | tristanC: ah! the & is correct - I have learned something! | 02:53 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Support booting cloud-images by name or id https://review.openstack.org/472959 | 02:57 |
| tristanC | mordred: heh :-) disown is like nohup without the pts redirection to a file | 03:04 |
| tristanC | well I'm not super happy with that test, but since bwrap double fork it's not trivial to detect process leak | 03:06 |
| *** yolanda has joined #zuul | 03:11 | |
| *** yolanda has quit IRC | 04:35 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: zuul_stream: do not collect delegated task https://review.openstack.org/473219 | 05:53 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: zuul_stream: harden log processing https://review.openstack.org/473222 | 06:01 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement a static driver for Nodepool https://review.openstack.org/468624 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Extend Nodepool configuration syntax to support multiple drivers https://review.openstack.org/468751 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Collect request handling implementation in an OpenStack driver https://review.openstack.org/468750 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool provider management code https://review.openstack.org/468749 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool request handling code https://review.openstack.org/468748 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Add support for custom ssh port https://review.openstack.org/468752 | 06:11 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement an OpenContainer driver https://review.openstack.org/468753 | 06:11 |
| *** jiaohaolin1 has joined #zuul | 06:13 | |
| jiaohaolin1 | hi ,everybody , I set up my CI with zuul 2.5.2 ,it works, but I can not open zuul page (master ip),it jump to https://jenkins and still can't open it | 06:17 |
| jiaohaolin1 | the last version of zuul can show the page, what should I do ? | 06:18 |
| tristanC | jiaohaolin1: did you setup the vhost like http://git.openstack.org/cgit/openstack-infra/puppet-zuul/tree/templates/zuul.vhost.erb ? | 06:19 |
| jiaohaolin1 | I'm not sure , I setup it follow https://docs.openstack.org/infra/openstackci/third_party_ci.html | 06:23 |
| jiaohaolin1 | tristanC: what should I do ? setup the vhost? how to do it ? Thank you very much | 06:25 |
| jiaohaolin1 | tristanC: :-D | 06:26 |
| *** yolanda has joined #zuul | 06:28 | |
| tristanC | the "zuul page" isn't listening by default on port 80, it needs a redirection like the one in zuul.vhost | 06:35 |
| *** isaacb has joined #zuul | 07:13 | |
| *** yolanda has quit IRC | 07:22 | |
| *** yolanda has joined #zuul | 07:43 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: github: handle ping event https://review.openstack.org/473249 | 07:44 |
| *** yolanda has quit IRC | 08:03 | |
| *** bhavik1 has joined #zuul | 08:10 | |
| *** isaacb has quit IRC | 08:12 | |
| *** lennyb has joined #zuul | 08:15 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: github: retry pull_request() https://review.openstack.org/473301 | 08:20 |
| *** isaacb has joined #zuul | 08:21 | |
| *** hashar has joined #zuul | 08:41 | |
| *** smyers has quit IRC | 09:02 | |
| *** smyers has joined #zuul | 09:13 | |
| *** yolanda has joined #zuul | 09:23 | |
| *** bhavik1 has quit IRC | 09:30 | |
| *** jkilpatr has quit IRC | 10:45 | |
| *** _ari_ is now known as _ari_|conf | 10:45 | |
| *** yolanda has quit IRC | 10:48 | |
| *** jkilpatr has joined #zuul | 11:05 | |
| *** hashar has quit IRC | 11:21 | |
| *** hashar has joined #zuul | 12:20 | |
| *** yolanda has joined #zuul | 12:24 | |
| *** yolanda has quit IRC | 14:45 | |
| *** isaacb has quit IRC | 15:02 | |
| *** isaacb has joined #zuul | 15:12 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add extra-test-setup role https://review.openstack.org/472472 | 16:10 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add python unit test jobs https://review.openstack.org/472484 | 16:10 |
| rcarrillocruz | yo yo | 16:15 |
| rcarrillocruz | http://38.145.33.129:9191/reports/ | 16:15 |
| rcarrillocruz | automated integration tests of ansible OVS | 16:16 |
| rcarrillocruz | powered by nodepool v3 cloud-images | 16:16 |
| rcarrillocruz | and ARA | 16:16 |
| rcarrillocruz | where's dmsimard dammit | 16:16 |
| rcarrillocruz | i'll tweet him later | 16:16 |
| mordred | periodic jobs with nodepool as a stop-gap til you can spin up the v3, yeah? | 16:16 |
| *** isaacb has quit IRC | 16:16 | |
| rcarrillocruz | yeah | 16:16 |
| mordred | rcarrillocruz: I bet since you have that, transitioning to periodic jobs in v3 will be a decently easy next step | 16:17 |
| rcarrillocruz | i need to find time to poke at zuul github stuff | 16:17 |
| rcarrillocruz | right | 16:17 |
| rcarrillocruz | the playbook should be just the same | 16:17 |
| mordred | yup | 16:17 |
| rcarrillocruz | ansible all the way down \o/ | 16:17 |
| mordred | \o/ | 16:17 |
| rcarrillocruz | so kudos folks for all the work you're doing | 16:17 |
| mordred | it's almost like this ansible idea is gonna work out :) | 16:17 |
| rcarrillocruz | haha, indeed | 16:17 |
| gundalow | \o/ | 16:42 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add 'description' field to jobs https://review.openstack.org/472483 | 16:43 |
| Shrews | rcarrillocruz: v3 builder _and_ launcher? | 16:43 |
| rcarrillocruz | no builds, just launcher with cloud-images | 16:44 |
| rcarrillocruz | but yeah, v3 branch | 16:44 |
| Shrews | awesome | 16:44 |
| Shrews | pabelanger: you had requested this I think? https://review.openstack.org/470364 Would be nice to get that merged | 16:45 |
| clarkb | rcarrillocruz: you are using the as of yet unreleased feature on the v3 branch to supprot launching off of existing cloud images? | 16:46 |
| rcarrillocruz | yup | 16:47 |
| clarkb | good to hear that is working, maybe you can respond to the nodepool + arm64 thread with how you set that up? | 16:48 |
| clarkb | I think they would be interested | 16:48 |
| pabelanger | Shrews: thanks | 16:48 |
| clarkb | http://lists.openstack.org/pipermail/openstack-infra/2017-June/005397.html that thread | 16:48 |
| jeblair | clarkb: i'm not sure i'd recommend that someone try to use the v3 nodepool with a v2 openstackci setup. are you sure you want to do that? | 16:52 |
| clarkb | jeblair: I think paul already did? he mentioned teh change at least | 16:52 |
| clarkb | it was more of "heres a thing that exists it works for some people and this is how they are using it" | 16:53 |
| clarkb | perhaps that would make them interested in backporting the feature to master? | 16:53 |
| jeblair | clarkb: i'm not aware of anyone using v3 nodepool with v2 zuul. in fact, i think we've even shelved the idea of doing that in infra... | 16:54 |
| rcarrillocruz | although, not sure if i read that correctly, i'm running off feature/zuulv3, i mean , support for that is already in that branch | 16:55 |
| rcarrillocruz | not running a local patch | 16:55 |
| clarkb | rcarrillocruz: yes the feature you are using is only on zuulv3 branch, not on master | 16:55 |
| rcarrillocruz | yep | 16:55 |
| rcarrillocruz | gah | 16:56 |
| rcarrillocruz | wifi died | 16:56 |
| rcarrillocruz | clarkb: was that arm comment to me | 16:56 |
| rcarrillocruz | ? | 16:56 |
| rcarrillocruz | ah yeah | 16:58 |
| rcarrillocruz | gmail refreshed | 16:58 |
| rcarrillocruz | i'll respond with a paste for my nodepool.yaml | 16:58 |
| clarkb | ya was just thinking it might be useful information to them. And maybe we can suggest they backport the feature? | 16:58 |
| clarkb | jeblair: ^ | 16:58 |
| jeblair | i don't think we should backport the feature | 16:58 |
| pabelanger | https://review.openstack.org/#/c/472861/ is an easier review for missing bindep.txt dependencies on openstack-zuul-roles | 16:58 |
| jeblair | it's not straightforward, and it sends some pretty mixed signals about nodepool development | 16:58 |
| jeblair | (also, reviewer time is at a premium right now) | 16:59 |
| rcarrillocruz | has nova servers metadata tags ever been considered and asked? from docs I see meta cna be put on images, but not in a server-by-server basis ? | 17:01 |
| rcarrillocruz | for nodepool i mean | 17:01 |
| clarkb | rcarrillocruz: we use it for leak detection. Nodepool adds metadata that says nodepool launched this node so it can check for them later | 17:02 |
| *** hashar has quit IRC | 17:02 | |
| clarkb | rcarrillocruz: we also use it for key management now | 17:02 |
| rcarrillocruz | hmm, so you can put a meta param on a label? | 17:02 |
| clarkb | rcarrillocruz: I don't know if we expose the ability to add arbitrary meta data if that is what you mean | 17:03 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add documentation to jobs https://review.openstack.org/472485 | 17:03 |
| clarkb | but we do use metadata in nodepool on nova instances | 17:03 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add Sphinx module for Zuul jobs https://review.openstack.org/472743 | 17:03 |
| rcarrillocruz | yeah, that's what i'm after | 17:03 |
| *** morgan is now known as mordgan | 17:03 | |
| pabelanger | I thought we did have a meta field at one point | 17:03 |
| rcarrillocruz | yeah, but for images | 17:03 |
| *** mordgan is now known as morgan | 17:03 | |
| pabelanger | ah, right | 17:04 |
| rcarrillocruz | would people be opposed to it ? I think it can be useful on just nodepool scenarios, thinking of getting automatic group form Ansible openstack.py via metadata | 17:05 |
| rcarrillocruz | like label name for example | 17:05 |
| clarkb | rcarrillocruz: I think it would likely be fine. The biggest thing to watch out for is conflicting with the keys nodepool already needs to use and adding so much metadata that an instance exceeds its metadata limit | 17:09 |
| jeblair | rcarrillocruz: provider name and image name are already automatic group entries | 17:09 |
| jeblair | rcarrillocruz: also, note that zuulv3 (which i know you aren't using) has support for inventory groups | 17:09 |
| rcarrillocruz | yeah saw that, but no label right? | 17:09 |
| jeblair | rcarrillocruz: right, i don't think label is included | 17:09 |
| rcarrillocruz | maybe auto adding label instad of arbitrary meta is a better compromise | 17:10 |
| rcarrillocruz | i'll get a look at the code | 17:10 |
| jeblair | rcarrillocruz: yeah, i suspect if we were to write the automatic group metadata stuff today, we would have included label; i think it may have been a v3 oversight. | 17:11 |
| rcarrillocruz | ++ | 17:11 |
| jlk | o/ | 17:28 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Add log streaming test https://review.openstack.org/471079 | 17:41 |
| Shrews | ^^^ that *might* actually work now | 17:41 |
| Shrews | heavily commented because computers are hard | 17:42 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Enhance sphinx plugin https://review.openstack.org/473544 | 17:42 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Add log streaming test https://review.openstack.org/471079 | 17:46 |
| Shrews | gah. pep8 | 17:46 |
| Shrews | \o/ for green things | 18:00 |
| *** cinerama` is now known as cinerama | 18:24 | |
| *** Thelo_ has joined #zuul | 18:26 | |
| *** Thelo_ has left #zuul | 18:27 | |
| mordred | Shrews: \o/ | 19:00 |
| mordred | Shrews: btw - don't know if you saw in scrollback - but starting https://review.openstack.org/#/c/437764 I did a bit of work on getting finger urls into the status page | 19:01 |
| Shrews | mordred: sweet. will begin looking at those | 19:03 |
| Shrews | mordred: responded to your call for HALP on that one | 20:10 |
| clarkb | jeblair: while I'm poking at log related items I recall that zuul 2.5 emits the job compelte event after all logs are completely copied. Is that accurate? and that behavior will carry through to zuul v3? wondering if I can simplify the logstash workers | 20:49 |
| jeblair | clarkb: v2.5: yes | 20:55 |
| jeblair | clarkb: v3: there are no zmq events in v3. instead, we will write a post playbook to deal with log processing. first, we'll probably just have it either send a compatible zmq event, or maybe submit a background gearman job. later, i think it can perhaps take on some of the actions we actually do in the processors. the timing will depend on which job we attach the post-playbook to. i'm guessing our base job since we want to include console logs. | 20:57 |
| *** harlowja has quit IRC | 21:02 | |
| clarkb | and we can assume that happens after the console iw copied right? | 21:06 |
| clarkb | basically no more need to delay | 21:06 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul feature/zuulv3: Default bubblewrap to work_root https://review.openstack.org/473099 | 21:23 |
| pabelanger | jeblair: left comment on 473099 | 21:25 |
| jeblair | clarkb: i think that's correct in both cases, yes. | 21:26 |
| mordred | jeblair: most of our tests do not define a success_url or a url_pattern, so the final url reported is the status_url - do you have thoughts on what you think we _should_ be doing there? | 21:27 |
| jeblair | mordred: define it if it's important for the test? | 21:29 |
| *** jkilpatr has quit IRC | 21:30 | |
| jeblair | pabelanger: thanks; replied. https://review.openstack.org/473164 is also related. | 21:30 |
| mordred | jeblair: indeed. lemme ask better questions: is url_pattern still a thing? do you think we should prefer a zuul-level url_pattern? or a pipline-level success_url? | 21:30 |
| mordred | jeblair: (like, I'm looking at test_json_status which righ tnow has thesame values in url and report_url and I'd like to improve it a little bit to have different values | 21:31 |
| pabelanger | jeblair: thanks | 21:32 |
| jeblair | mordred: neither zuul-level url_pattern nor pipeline success-url are things anymore; success-url is job-levl only now | 21:32 |
| mordred | jeblair: gotcha. so defining a success-url on the base job in that test would be the thing I shoujld do | 21:32 |
| jeblair | mordred: yep i think so | 21:32 |
| jeblair | mordred: looking at the test | 21:32 |
| jeblair | mordred: yep | 21:33 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make logging helper method in zuul_stream https://review.openstack.org/472963 | 21:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Change log streaming link to finger protocol https://review.openstack.org/437764 | 21:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Move streaming url formatting to model https://review.openstack.org/473090 | 21:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Only prepend hostname on multi-node plays https://review.openstack.org/472964 | 21:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Support finger ports in finger URL https://review.openstack.org/473103 | 21:34 |
| mordred | jeblair: ok. I'll add that to this stack | 21:34 |
| mordred | jeblair: also - paul pointed me at the proposed rfc for finger urls, which is finger://{host}[:{port}]/{resource} rather than finger://{resource}@{host}[:{port] - you good with that? | 21:35 |
| jeblair | mordred: that makes me sad... does anyone implement it? (but i think we should probably follow the rfc or implementations on this, not make up our own) | 21:37 |
| mordred | jeblair: lynx implements support for both forms | 21:38 |
| mordred | jeblair: I do sort of like host/resource if for no other reason that it should make requests of specific files not look as strange | 21:39 |
| mordred | jeblair: although I agree with the sad feeling | 21:39 |
| jeblair | mordred: 'links' seems to send "/W foo" when you finger://foo@host/. and it sends "/W" when you finger://host/foo. | 21:41 |
| jeblair | mordred: i'm not sure that helps us? | 21:41 |
| mordred | hrm | 21:41 |
| jeblair | what does "/W" mean? | 21:42 |
| jeblair | https://tools.ietf.org/html/rfc1288#section-2.5.4 | 21:43 |
| jeblair | aha | 21:43 |
| jeblair | mordred: that makes me think that 'links' supports only foo@host, and sends "/W foo" as the query, meaning additional verbosity. | 21:43 |
| jeblair | fungi: you may want in on this bikeshed | 21:44 |
| fungi | mmm | 21:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add a success-url for status.json test https://review.openstack.org/473604 | 21:52 |
| fungi | so was /resource in a ietf draft then, or is there an accepted numbered rfc? | 21:54 |
| mordred | fungi: just a draft | 21:55 |
| fungi | i've seen so many (sometimes directly conflicting!) drafts about extending other protocols i tend not to give them much weight | 21:55 |
| mordred | https://tools.ietf.org/html/draft-ietf-uri-url-finger-03 | 21:55 |
| mordred | fungi: I believe that's the only thing we've found that defines this at all | 21:55 |
| fungi | keeping in mind that even rfcs are not necessarily "standards" in the strictest sense | 21:55 |
| mordred | fungi, jeblair: also: http://lynx.browser.org/lynx2.8.8/lynx_help/lynx_url_support.html#finger_url | 21:56 |
| fungi | draft-ietf-uri-url-finger-03 expired more than 20 years ago without gaining consensus | 21:57 |
| mordred | so if we emit as resource@host, we'll work with both links and lynx - if we emit as host/resource we'll be consistent with an expired non-adopted rfc | 21:57 |
| fungi | yup. in the ensuing 20 years, consensus emerged on its own in my opinion | 21:58 |
| mordred | fungi: there is an o'reilly HTTP book that lists the host/resource form: | 21:59 |
| mordred | https://books.google.com/books?id=qEoOl9bcV_cC&pg=PT518&lpg=PT518&dq=rfc+finger+url&source=bl&ots=zxuYVpNzbj&sig=S7x_lyrd-Fg1tfwcu_9Sn1vqswE&hl=en&sa=X&ved=0ahUKEwiG4N_JprnUAhXDz4MKHc64DFsQ6AEIUzAI#v=onepage&q=rfc%20finger%20url&f=false | 21:59 |
| fungi | (even though finger != http) | 21:59 |
| mordred | yah | 21:59 |
| * mordred is mostly trolling to see what, if anything, we might reasonably break if we choose one over the other | 22:00 | |
| fungi | looks like rfc 1738 is the reason why they didn't go with user@host | 22:01 |
| fungi | ooh, meeting time | 22:01 |
| jeblair | zuul meeting time in #openstack-meeting-alt | 22:01 |
| mordred | fungi: http://tidbits.com/article/1632 says that Peter's MacTCP based finger client supports finger://resource@ format directly | 22:02 |
| jeblair | jlk: are you around for the zuul meeting? | 22:03 |
| *** harlowja has joined #zuul | 22:06 | |
| *** yolanda has joined #zuul | 22:34 | |
| Shrews | mordred: so, i that check for py35 in that streaming test is unintentionally left in. I meant it for a future test of the websocket stuff. There's actually nothing py3 specific there | 23:02 |
| tristanC | python3 is actually not really an issue on centos... but how about nodepoolv3, can it run py3 too? | 23:02 |
| clarkb | ya use software collections | 23:04 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Add log streaming test https://review.openstack.org/471079 | 23:04 |
| pabelanger | clarkb: from emails I have seen for Red hat, that is not an option. | 23:04 |
| Shrews | mordred: ^^^ just corrected that | 23:04 |
| pabelanger | it is missing a lot of python dependencies | 23:04 |
| clarkb | pabelanger: you mean system packages for python deps? | 23:05 |
| jamielennox | can someone tell me if i really must run zuul-executor as root? i know it drops privs but is the only reason so that it can get the finger port? | 23:07 |
| pabelanger | clarkb: right. If I understand the base python3-devel packages is that but only python packages related to django (I think) have actually been packaged as RPMs. I would expect pip install to work, but obviously red hat like RPMs | 23:07 |
| pabelanger | jamielennox: right, finger port is the only reason IIRC | 23:07 |
| jamielennox | i don't really plan on exposing 79 directly anyway so if that's the only problem i can skip that | 23:08 |
| pabelanger | yes, should be. I haven't tested it yet | 23:08 |
| jlk | jeblair: sorry no my kid had a school event. | 23:11 |
| jlk | Could I get some reviews of https://review.openstack.org/#/c/472468/ ? | 23:20 |
| *** dkranz has quit IRC | 23:30 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!