| *** JasonCL has quit IRC | 00:01 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: WIP: Add IPv6 support for statsd https://review.openstack.org/556327 | 00:11 |
|---|---|---|
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: autohold: prevent using ',' in change argument https://review.openstack.org/556329 | 00:21 |
| *** odyssey4me has quit IRC | 00:35 | |
| *** odyssey4me has joined #zuul | 00:35 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: WIP: Add IPv6 support for statsd https://review.openstack.org/556327 | 00:37 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Bump statsd to 3.x and uncap https://review.openstack.org/556331 | 00:37 |
| *** JasonCL has joined #zuul | 01:01 | |
| *** JasonCL has quit IRC | 01:05 | |
| *** JasonCL has joined #zuul | 01:13 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: launcher: handle ZK session loss during handler poll https://review.openstack.org/556335 | 01:14 |
| tristanC | Shrews: I think the handler poll is missing a SessionExpired exception too ^ | 01:15 |
| *** JasonCL has quit IRC | 01:21 | |
| *** JasonCL has joined #zuul | 01:21 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 01:24 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 01:37 |
| *** JasonCL has quit IRC | 01:51 | |
| *** JasonCL has joined #zuul | 01:51 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Update to statsd 3.0 https://review.openstack.org/229232 | 01:55 |
| *** JasonCL has quit IRC | 02:23 | |
| *** JasonCL has joined #zuul | 02:25 | |
| *** JasonCL has quit IRC | 02:28 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 02:35 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Update to statsd 3.0 https://review.openstack.org/229232 | 02:35 |
| *** kmalloc has quit IRC | 02:46 | |
| openstackgerrit | Bernd Bausch proposed openstack-infra/zuul master: Add clarity to *Zuul from Scratch* instructions https://review.openstack.org/556344 | 04:01 |
| openstackgerrit | Merged openstack-infra/nodepool master: Add support for STATSD_IPV6 https://review.openstack.org/556209 | 04:09 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 04:19 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Update to statsd 3.0 https://review.openstack.org/229232 | 04:19 |
| *** swest has joined #zuul | 05:06 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 06:08 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Update to statsd 3.0 https://review.openstack.org/229232 | 06:08 |
| *** kmalloc has joined #zuul | 06:17 | |
| *** flepied_ has quit IRC | 06:24 | |
| *** flepied_ has joined #zuul | 06:26 | |
| *** toabctl has joined #zuul | 06:32 | |
| *** hashar has joined #zuul | 07:15 | |
| *** electrofelix has joined #zuul | 07:31 | |
| *** adam_g has quit IRC | 07:39 | |
| *** flepied_ has quit IRC | 07:45 | |
| *** tosky has joined #zuul | 07:46 | |
| *** flepied_ has joined #zuul | 08:12 | |
| *** bhavik1 has joined #zuul | 08:25 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul web: add admin endpoint, enqueue & autohold commands https://review.openstack.org/539004 | 08:26 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul autohold: allow operator to specify nodes TTL https://review.openstack.org/543403 | 08:32 |
| *** flepied__ has joined #zuul | 08:33 | |
| *** flepied_ has quit IRC | 08:36 | |
| LinuxJedi | hey zuulites? zuuls? What is a collective noun for zuul devs? Anyway... I have 1 week to build a demo of Zuul v3. I can probably use a pre-existing GitHub project I developed a long time ago as an example. Any tips? (I'm guessing someone else has done a live demo of a simple (not OpenStack scale) setup before?) | 08:43 |
| *** bhavik1 has quit IRC | 09:07 | |
| *** sshnaidm|off has quit IRC | 09:15 | |
| *** adam_g has joined #zuul | 09:32 | |
| openstackgerrit | Alexander Chadin proposed openstack-infra/zuul-jobs master: Fix live migration of instances https://review.openstack.org/556464 | 09:43 |
| *** flepied_ has joined #zuul | 10:06 | |
| *** flepied__ has quit IRC | 10:08 | |
| *** sshnaidm has joined #zuul | 10:34 | |
| tobiash | LinuxJedi: you could start with https://docs.openstack.org/infra/zuul/admin/zuul-from-scratch.html | 11:26 |
| LinuxJedi | 👍 thanks tobiash | 11:27 |
| tobiash | LinuxJedi: if there are points that are unclear, don't hesitate to ask or submit changes to improve that guide | 11:28 |
| *** dkranz has joined #zuul | 12:00 | |
| *** JasonCL has joined #zuul | 12:11 | |
| *** snapiri has quit IRC | 12:16 | |
| Shrews | LinuxJedi: tip #1- for the nodepool side, if you don't have a cloud to use, set up the static driver and just plug in some pre-existing VM ips to use as your nodes. | 12:20 |
| Shrews | i'm messing around with that myself, atm | 12:20 |
| LinuxJedi | Shrews: cool, thanks. I was considering just using some VirtualBox VMs for now, I can VPN to my house to give the demo | 12:27 |
| Shrews | LinuxJedi: that's exactly what i'm doing | 12:28 |
| LinuxJedi | hmm... I have 1TB of VMs running on this machine, I need to clear a few out first :) | 12:29 |
| Shrews | LinuxJedi: i spent too much time trying to get devstack to give me a usable cloud. i lost sooo much hair doing this | 12:29 |
| LinuxJedi | I'd rather keep my hair for now. I don't want to have to wear the wig you gave me again ;) | 12:30 |
| Shrews | ROFL! | 12:30 |
| Shrews | LinuxJedi: but you looked so good in it! | 12:30 |
| Shrews | like a 80's rocker | 12:30 |
| LinuxJedi | na, I don't have the face paint for that, although my son does. | 12:31 |
| *** rlandy has joined #zuul | 12:38 | |
| *** odyssey4me has quit IRC | 12:47 | |
| *** odyssey4me has joined #zuul | 12:47 | |
| *** chrnils has joined #zuul | 12:48 | |
| LinuxJedi | Shrews: this may not be as easy as I thought. Fedora installer in VirtuaBox did this: https://twitter.com/LinuxJedi/status/978252126538387456 :D | 12:50 |
| Shrews | LinuxJedi: that's... spectacular | 12:54 |
| LinuxJedi | I suspect VirtualBox doesn't like Nvidia's driver (I don't blame it). Appears to only happen when I switch away from the window. | 12:54 |
| *** elyezer has joined #zuul | 12:55 | |
| *** hashar has quit IRC | 13:03 | |
| *** tosky__ has joined #zuul | 13:29 | |
| LinuxJedi | tobiash: stupid question for that guide, with fedora 27 server "bindep" doesn't end up in the path. Easy to workaround to use ~/.local/bin/bindep instead. But is there supposed to be a step to add that to path? | 13:31 |
| *** tosky has quit IRC | 13:31 | |
| *** electrofelix has quit IRC | 13:31 | |
| tobiash | LinuxJedi: does logout/login put it into the path? | 13:32 |
| LinuxJedi | no | 13:33 |
| tobiash | ok, so it looks like this step is missing ;) | 13:33 |
| LinuxJedi | cool, glad it is that and I wasn't missing something :) | 13:34 |
| *** hashar has joined #zuul | 13:38 | |
| *** elyezer has quit IRC | 13:50 | |
| *** elyezer has joined #zuul | 13:51 | |
| dmsimard | tobiash: btw I discussed the new module blacklisting feature with #ansible-devel that's highlighted in 2.5 ( https://www.ansible.com/blog/ansible-2.5-traveling-space-and-time ) | 14:22 |
| dmsimard | Some links: | 14:23 |
| dmsimard | - https://docs.ansible.com/ansible/devel/reference_appendices/config.html#plugin-filters-cfg | 14:23 |
| dmsimard | - https://github.com/ansible/ansible/blob/340a7be7c382d441475b64adbfd65e605786cd4f/examples/plugin_filters.yml | 14:23 |
| dmsimard | - https://docs.ansible.com/ansible/devel/user_guide/plugin_filtering_config.html | 14:23 |
| pabelanger | wouldn't mind a review on an older patchset: https://review.openstack.org/248975/ needed so we can uncap statsd and support ipv6! | 14:24 |
| dmsimard | It sounds like we could use this for Zuul or -- if we can't, maybe this would provide a proper place for us to hook our modifications in | 14:24 |
| clarkb | dmsimard: I dont think we want to blacklist modules instead prevent some of them from affecting localhost in specific ways | 14:26 |
| corvus | pabelanger: lgtm | 14:26 |
| dmsimard | clarkb: what I mean is that they stood up a structure to be able to greenlight the execution of a module or not and that we might be able to hook ourselves at this location to whitelist/blacklist the execution of things according to Zuul's business logic | 14:27 |
| dmsimard | or if it's not possible in the current state, upstream some improvements to be able to do that | 14:27 |
| corvus | pabelanger: i'd like to get a report of that working correctly before we release 3.0 (either by updating/restarting openstack-infra's scheduler, or someone else using graphite reporting success) | 14:27 |
| dmsimard | clarkb: mostly something that'd be a better solution than our current ones which are fairly rudimentary | 14:28 |
| pabelanger | corvus: sure, wfm | 14:29 |
| pabelanger | corvus: we do likely need to discuss: https://review.openstack.org/556327/ for ipv6 statsd in zuul, I'm really not sure how we'd like to do that | 14:30 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: WIP: Add IPv6 support for statsd https://review.openstack.org/556327 | 14:31 |
| clarkb | dmsimard: my reading of that is that you hvae to entirely blacklist the module in config rather than allowing selective execution. I think that means that zuul itself would have to read the ansible playbooks and determine if any of the tasks should be blacklisted before running anible playbook? | 14:38 |
| dmsimard | clarkb: right, but we're like 70% of the way there if the only need we need to add in is to improve on the selectiveness | 14:39 |
| dmsimard | clarkb: like instead of 100% blacklist, only blacklist on specific hosts, for example | 14:40 |
| clarkb | dmsimard: it is more subtle that that in our case but that may be something we can work to improve there. For example with the copy module its ok to copy from safe path to safe path on localhost | 14:40 |
| dmsimard | ++ | 14:40 |
| Shrews | can the <gerrit source> used in setting up a trigger (mentioned here https://docs.openstack.org/infra/zuul/admin/drivers/gerrit.html#trigger-configuration) be anything? Or must that reference something? | 14:43 |
| Shrews | clarkb: corvus: ^^ ? | 14:43 |
| clarkb | Shrews: it has to match your configured source in the ini file iirc | 14:43 |
| clarkb | Shrews: that way you can specify if this is source specific or list all of them to do it for all the gerrits (in a multi gerrit setup) | 14:44 |
| Shrews | clarkb: oh, so [connection gerrit] can be [connection anything]... gotcha | 14:44 |
| clarkb | Shrews: yup | 14:45 |
| Shrews | so long as driver=gerrit. that wasn't immediately clear | 14:45 |
| openstackgerrit | Merged openstack-infra/zuul master: Handle a float in assertReportedStat https://review.openstack.org/248975 | 14:47 |
| openstackgerrit | Merged openstack-infra/zuul master: Update to statsd 3.0 https://review.openstack.org/229232 | 14:47 |
| clarkb | Shrews: might be worth renaming [connection gerrit] to mygerrit and using mygerrit instead in docs | 14:48 |
| clarkb | to make that distinction a little more explicit | 14:48 |
| Shrews | clarkb: yeah. there's more than one place where that would help. it's on my list | 14:49 |
| corvus | ++ | 14:49 |
| *** eventingmonkey has quit IRC | 15:01 | |
| *** leifmadsen has quit IRC | 15:02 | |
| *** eventingmonkey has joined #zuul | 15:03 | |
| *** leifmadsen has joined #zuul | 15:04 | |
| *** dmsimard is now known as dmsimard|cave | 15:05 | |
| *** tosky__ is now known as tosky | 15:07 | |
| *** kmalloc has quit IRC | 15:16 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul master: Fix gerrit connection references https://review.openstack.org/556555 | 15:16 |
| Shrews | clarkb: ^^ | 15:17 |
| *** JasonCL has quit IRC | 15:18 | |
| LinuxJedi | ok, so... I'm trying to get the Zuul web interface built using "npm run build:docker" as mentioned here: https://docs.openstack.org/infra/zuul/developer/javascript.html and I get "Error: EACCES: permission denied, scandir '/usr/src/app'". I know nothing about Node so any ideas what I'm doing wrong? | 15:27 |
| corvus | LinuxJedi: i haven't seen that -- i used the non-docker method starting at https://docs.openstack.org/infra/zuul/developer/javascript.html#yarn-dependency-management | 15:33 |
| corvus | mordred: ^ do you recognize the docker error? | 15:33 |
| LinuxJedi | I may fall back to that, I just wanted to do try lazy thing :) | 15:34 |
| clarkb | if I see modred today I'll ask him to take a look | 15:35 |
| corvus | i'm trying the docker run now to see if i can reproduce | 15:36 |
| LinuxJedi | thanks. No rush. I appear to have gotten Zuul to talk to GitHub so that is a good start I guess :) | 15:36 |
| LinuxJedi | if it helps the host is Fedora 27 | 15:36 |
| clarkb | he has a session first thing this morning but its in the dev summit area of ons which I guess required rsvp'ing which I did not do | 15:37 |
| corvus | okay, i just ran the docker command and it ran without error | 15:38 |
| corvus | here's the output: http://paste.openstack.org/show/714238/ | 15:39 |
| *** JasonCL has joined #zuul | 15:42 | |
| LinuxJedi | I suspect something funky with Fedora's version of Docker | 15:45 |
| *** JasonCL has quit IRC | 15:45 | |
| corvus | i'm on ubuntu xenial 16.04 | 15:45 |
| corvus | i might have a fedora node i can try on | 15:45 |
| LinuxJedi | I have a meeting in a minute and I still need to figure out a sane config for the tree I'm playing with so absolutely no rush on this :) | 15:48 |
| corvus | LinuxJedi: i take it you did: "sudo yum install npm docker" and "sudo npm run build:docker" ? | 15:52 |
| corvus | LinuxJedi: i just did that, and got the error you did: http://paste.openstack.org/show/714250/ | 15:53 |
| Shrews | corvus: what does line 105 in http://paste.openstack.org/show/714249/ mean? | 15:53 |
| LinuxJedi | corvus: correct | 15:55 |
| corvus | Shrews: it's harmless -- when zuul adds a change for testing and its dependency to an independent pipeline, the change for testing is considered a "live" change, and its dependency is considered "non-live". when the "live" change is removed, that leaves only the non-live change in the pipeline. a "non-live" change with no items behind is no longer useful, so it is removed. | 15:56 |
| *** pabelanger has quit IRC | 15:56 | |
| corvus | Shrews: basically, zuul always has a reason for removing each individual change from a pipeline, and the debug logs record that. | 15:56 |
| corvus | Shrews: so this is just the normal procedure for unwinding a dependency chain in an independent pipeline. | 15:57 |
| *** fbo_ has quit IRC | 15:57 | |
| *** weshay has quit IRC | 15:57 | |
| Shrews | corvus: oh. was hoping maybe that was the reason why my zuul didn't seem to be running my job | 15:58 |
| Shrews | guess i should look elsewhere | 15:58 |
| *** mhu has quit IRC | 15:58 | |
| corvus | Shrews: oh, that's probably line 95. | 15:58 |
| Shrews | corvus: oh, hahah! i defined the job, but didn't put it in any pipeline. doh! | 15:59 |
| Shrews | my fingers got in a hurry | 15:59 |
| *** pabelanger has joined #zuul | 16:00 | |
| corvus | LinuxJedi: can i tell you how excited i am that *docker behaves differently on different platforms*? | 16:00 |
| *** mhu has joined #zuul | 16:02 | |
| *** fbo_ has joined #zuul | 16:02 | |
| *** weshay has joined #zuul | 16:02 | |
| *** JasonCL has joined #zuul | 16:04 | |
| *** JasonCL has quit IRC | 16:05 | |
| Shrews | oh, the nodepool static driver doesn't make a node READY again after use? That surprises me | 16:08 |
| corvus | LinuxJedi, mordred: i suspect that selinux permissions are active inside the container. i doubt that alpine linux image has the tools to change directory permissions... | 16:08 |
| *** JasonCL has joined #zuul | 16:10 | |
| Shrews | oh, they just don't show up as READY. *that* surprises me | 16:11 |
| corvus | Shrews: what do you mean? | 16:12 |
| *** JasonCL has quit IRC | 16:13 | |
| Shrews | corvus: 'nodepool list' does not show me static nodes | 16:13 |
| *** JasonCL has joined #zuul | 16:13 | |
| Shrews | corvus: though i can use them | 16:13 |
| *** elyezer has quit IRC | 16:14 | |
| *** JasonCL has quit IRC | 16:14 | |
| *** sshnaidm has quit IRC | 16:14 | |
| corvus | Shrews: yeah. there was a suggestion that the implementation of the static driver should create node entries for all of the nodes. that is not the implementation that landed. instead, the inventory is only in-memory until it's added to a request. i agree that the other way would be more intuitive. | 16:15 |
| *** JasonCL has joined #zuul | 16:15 | |
| Shrews | corvus: in other news, i just had my own zuul successfully run a job! though it did not report it correctly for some reason | 16:16 |
| corvus | Shrews: see comments starting on PS9 of https://review.openstack.org/468624 | 16:16 |
| corvus | Shrews: \o/ | 16:16 |
| corvus | Shrews: those debug messages we were looking at earlier should say why | 16:16 |
| Shrews | corvus: hrm, doesn't appear so. i think it has something to do with my gerrit only recognizing +2 and i have it configured for zuul to vote +1 | 16:18 |
| Shrews | though i did get: Build succeeded. | 16:19 |
| Shrews | job1 finger://wilddog/869882a1ff6f4f8da1a40875823bb755 : SUCCESS in 4s | 16:19 |
| corvus | Shrews: you mean the vote didn't show up right? | 16:19 |
| *** tosky has quit IRC | 16:19 | |
| Shrews | corvus: correct, but that comment did | 16:19 |
| *** JasonCL has quit IRC | 16:20 | |
| corvus | Shrews: maybe there's a problem with the reporter configuration, so zuul isn't sending the right arguments to vote. there should be a debug log entry for the gerrit command it used, you might check that looks correct. | 16:20 |
| Shrews | corvus: doesn't appear so. maybe i need to enable debugging for the driver? http://paste.openstack.org/show/714286/ | 16:22 |
| corvus | Shrews: ah, yes probably so. | 16:22 |
| corvus | Shrews: though {'Verified': 1} helps | 16:23 |
| corvus | LinuxJedi: i think i found the problem; patch incoming | 16:23 |
| *** elyezer has joined #zuul | 16:27 | |
| *** JasonCL has joined #zuul | 16:27 | |
| Shrews | corvus: nothing more useful that i can see: http://paste.openstack.org/show/714295/ | 16:27 |
| Shrews | though, continue helping LinuxJedi. i'll look at gerrit configuration | 16:28 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add selinux flag to docker command https://review.openstack.org/556582 | 16:28 |
| corvus | LinuxJedi: ^ there you go! we were missing a single "z". which is basically always the case with selinux. :) | 16:28 |
| *** JasonCL has quit IRC | 16:29 | |
| *** tosky has joined #zuul | 16:29 | |
| *** JasonCL has joined #zuul | 16:29 | |
| corvus | Shrews: does the zuul account have permissions in gerrit to leave verified votes? | 16:29 |
| Shrews | corvus: it's the adming account, so i'm assuming so | 16:29 |
| Shrews | admin | 16:30 |
| Shrews | gerrit makes it difficult to add users w/o SSO | 16:30 |
| Shrews | s/difficult/impossible/ | 16:30 |
| corvus | Shrews: there's the 'become any user' auth system, that may help. | 16:31 |
| clarkb | you can create system users wirh ssh api | 16:31 |
| corvus | Shrews: but the admin user actually *doesn't* have all perms by default | 16:31 |
| Shrews | corvus: but that came with a big warning "DO NOT USE!!!!11!!" | 16:31 |
| corvus | Shrews: yeah, clarkb's suggestion about creating a system user is probably the way to go | 16:31 |
| corvus | Shrews: here's an example: https://docs.openstack.org/infra/system-config/gerrit.html#users | 16:32 |
| Shrews | oh neat. thx | 16:32 |
| *** JasonCL has quit IRC | 16:33 | |
| *** JasonCL has joined #zuul | 16:36 | |
| *** JasonCL has quit IRC | 16:38 | |
| LinuxJedi | corvus: I figured it would be SELinux, it is always SELinux! :D | 16:43 |
| LinuxJedi | many thanks! | 16:43 |
| corvus | LinuxJedi: thanks for finding bugs :) | 16:47 |
| LinuxJedi | thanks for fixing them so quick! | 16:47 |
| *** sshnaidm has joined #zuul | 16:48 | |
| LinuxJedi | I also had some Fedora firewalld issue with port 9000. So I think it is just that Fedora just wants to be locked down and left alone in the corner :) | 16:49 |
| corvus | oh interesting, i don't think i had a firewall on my image | 16:49 |
| LinuxJedi | although I won't bash it too much as it is my primary OS | 16:49 |
| clarkb | LinuxJedi: is that the cloud image? | 16:49 |
| clarkb | I want to say the cloud image us morw forgiving | 16:49 |
| LinuxJedi | no, I just build a VirtualBox VM from the server ISO | 16:49 |
| corvus | ya i used cloud image | 16:49 |
| LinuxJedi | well this happened whilst installing, so yea. Maybe cloud image next time: https://twitter.com/LinuxJedi/status/978252126538387456 :D | 16:50 |
| *** kmalloc has joined #zuul | 16:52 | |
| LinuxJedi | corvus: one other tiny thing, doc/source/developer/javascript.rst should be a lowercase PWD I think | 16:54 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix gerrit connection references https://review.openstack.org/556555 | 16:58 |
| LinuxJedi | ok, next up, GitHub integration hates me: http://paste.openstack.org/show/714392/ | 17:15 |
| *** flepied_ has quit IRC | 17:18 | |
| *** JasonCL has joined #zuul | 17:29 | |
| Shrews | LinuxJedi: it's not just you. gh hates everyone | 17:31 |
| Shrews | LinuxJedi: i'm not much help w/ that, but did you check https://docs.openstack.org/infra/zuul/admin/drivers/github.html#configure-github ? | 17:32 |
| LinuxJedi | its ok, I fixed it. Turns out I couldn't turn on all the required permissions until after I had linked the app to the project | 17:32 |
| LinuxJedi | final problem I have is "Pipelines may not be defined in untrusted repos" but that is on me to fix as well, after food :) | 17:34 |
| *** JasonCL has quit IRC | 17:38 | |
| *** JasonCL has joined #zuul | 17:38 | |
| *** chrnils has quit IRC | 17:44 | |
| *** nhicher_ has quit IRC | 17:50 | |
| *** nhicher has joined #zuul | 17:50 | |
| *** JasonCL has quit IRC | 17:54 | |
| Shrews | clarkb: corvus: oh, i think i need to enable the Verified label in my gerrit. i think that's why i'm not seeing the votes | 18:08 |
| Shrews | i wonder if i can just use 'Code-Review' in place of 'Verified' in the pipeline config | 18:10 |
| clarkb | Shrews: you should be able to, zuul will just be voting like any other code reviewer in that case | 18:12 |
| Shrews | clarkb: i think that's fine. trying to do this for a 'gerrit quick start' guide that's just covering the basics. more advanced configuration can be covered separately | 18:13 |
| clarkb | ++ | 18:13 |
| clarkb | and by default I think it may just be +/-1 too, no +/-2? | 18:14 |
| Shrews | w00t! Code-Review +1 Zuul | 18:16 |
| Shrews | so many pieces... falling... all together. how neat | 18:16 |
| *** flepied_ has joined #zuul | 18:17 | |
| Shrews | ok, i think i have enough to go on for a gerrit-based quick start that ignores zuul-web and logging things | 18:17 |
| Shrews | now to try and make my chicken scratch notes cohesive and understandable | 18:18 |
| *** JasonCL has joined #zuul | 18:25 | |
| *** JasonCL has quit IRC | 18:26 | |
| *** JasonCL has joined #zuul | 18:26 | |
| tobiash | Yay, the license field problem in github3.py is fixed: https://github.com/sigmavirus24/github3.py/pull/804 | 18:29 |
| tobiash | Hopefully I can test this this week with zuul | 18:29 |
| *** JasonCL has quit IRC | 18:31 | |
| pabelanger | corvus: do you have any suggestions on https://review.openstack.org/556327 I know you are likely busy | 18:37 |
| *** snapiri has joined #zuul | 18:58 | |
| *** JasonCL has joined #zuul | 19:03 | |
| *** JasonCL has quit IRC | 19:04 | |
| *** JasonCL has joined #zuul | 19:05 | |
| *** JasonCL has quit IRC | 19:09 | |
| LinuxJedi | Shrews: next problem you might be able to help with. I followed https://docs.openstack.org/infra/zuul/admin/zuul-from-scratch.html and got it to trigger on a commit. The web log was then complaining that it couldn't connect to a gearman server. So I installed gearmand and got that running. Now the web service appear to be hung. What did I break? :) | 19:13 |
| *** snapiri has quit IRC | 19:15 | |
| Shrews | LinuxJedi: umm, not sure | 19:15 |
| Shrews | LinuxJedi: anything in the logs? | 19:16 |
| LinuxJedi | I'm beginning | 19:16 |
| clarkb | LinuxJedi: Shrews zuul forks its own geard by default | 19:16 |
| LinuxJedi | I'm beginning to thing there are some missing steps in that tutorial :) | 19:16 |
| clarkb | possible the two gearmans are fighting and sad? | 19:16 |
| LinuxJedi | ah, that may be why it broke things. Let me undo what I did and then see why geard didn't fork | 19:17 |
| rcarrillocruz | folks, i think we had at some point the playbooks run in a job as collected data ? e.g. http://logs.openstack.org/93/554993/1/gate/puppet-openstack-lint/d0d753b/zuul-info/ | 19:17 |
| rcarrillocruz | is that gone away | 19:17 |
| rcarrillocruz | we never had it | 19:17 |
| clarkb | LinuxJedi: the process won't be called geard though it will be a zuul-scheduler child process iirc | 19:17 |
| rcarrillocruz | or i'm not looking at right place | 19:17 |
| LinuxJedi | ah! That explains it, there was a missing step around /usr/lib/tmpfiles.d which means zuul-scheduler doesn't survive a reboot | 19:18 |
| clarkb | rcarrillocruz: we had that for zuulv2.5, but I don't think zuulv3 kept it since you are explicitly using playbooks and can find them in the source | 19:18 |
| rcarrillocruz | LinuxJedi: are you the same LinuxJedi from years ago that was in the infra team ? | 19:18 |
| LinuxJedi | rcarrillocruz: yes | 19:18 |
| clarkb | rcarrillocruz: with 2.5 we had them to aid in debugging when zuul did something unexpected | 19:18 |
| rcarrillocruz | clarkb: ok, thx, i thought i wwas going senile | 19:18 |
| clarkb | rcarrillocruz: also ara will include that information as well if you have ara | 19:18 |
| clarkb | rcarrillocruz: so I guess we do include it but only on failures | 19:19 |
| rcarrillocruz | LinuxJedi: nice! awesome to see you around | 19:19 |
| clarkb | (currently for reasons of not enough inodes) | 19:19 |
| LinuxJedi | rcarrillocruz: I am now a Lead Software Engineer for MariaDB and I've pretty much forgotten anything useful for this task :) | 19:19 |
| LinuxJedi | rcarrillocruz: you too :) | 19:19 |
| rcarrillocruz | clarkb: AHA, just in failures | 19:19 |
| rcarrillocruz | sweet, thx sir | 19:19 |
| *** JasonCL has joined #zuul | 19:20 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool master: Add host-key-checking option to opentack providers https://review.openstack.org/556624 | 19:20 |
| clarkb | rcarrillocruz: the ara wsgi middleware is aimed at allowing us to provide ara reports for all jobs again as only the sqlite file is required for that | 19:20 |
| rcarrillocruz | LinuxJedi: does it mean are you going to implement Zuul for MariaDB ? | 19:20 |
| LinuxJedi | rcarrillocruz: I'm demoing it to Monty and the QA people next week. With any luck we will | 19:20 |
| rcarrillocruz | geez, that's fantastic | 19:21 |
| rcarrillocruz | love seeing zuul getting into more places | 19:21 |
| rcarrillocruz | (i'm myself pushing for putting it on ansible, along with other folks) | 19:21 |
| pabelanger | clarkb: corvus: mrhillsman: rcarrillocruz: ^what we talked about last week to toggle ssh-keyscan. Open to suggestions on naming the setting, host-key-checking lines up with ansible | 19:21 |
| rcarrillocruz | pabelanger: yah, host-key-checking would be good | 19:21 |
| rcarrillocruz | ++ from me | 19:22 |
| rcarrillocruz | also, ssh -o is StrictHostKeyChecking , so... | 19:24 |
| *** JasonCL has quit IRC | 19:24 | |
| pabelanger | yah, that is on zuul side. So we'd need to either allow that to be toggled or operator would write pre playbook on executor to maybe ssh-keyscan the node first | 19:25 |
| pabelanger | haven't tested that yet | 19:25 |
| *** JasonCL has joined #zuul | 19:28 | |
| *** JasonCL has quit IRC | 19:29 | |
| *** JasonCL has joined #zuul | 19:30 | |
| *** JasonCL has quit IRC | 19:34 | |
| LinuxJedi | woot! https://github.com/libattachsql/libattachsql/pull/202 thanks everyone who helped so far. I'll spend tomorrow adding real tests to this thing :) | 19:35 |
| *** JasonCL has joined #zuul | 19:36 | |
| *** harlowja has joined #zuul | 19:37 | |
| *** JasonCL has quit IRC | 19:43 | |
| *** JasonCL has joined #zuul | 19:48 | |
| *** JasonCL has quit IRC | 19:50 | |
| tobiash | pabelanger: how would nodepool determine how long to wait until the node is fully booted? | 19:54 |
| tobiash | I guess the idea is to co-locate an executor to the private network. So what about co-locating also a nodepool launcher to it? | 19:56 |
| *** EmilienM has quit IRC | 19:57 | |
| pabelanger | tobiash: it wouldn't wait, once openstack API said the node was online, then it would pass it back to zuul to use. Zuul would then have to validate networking is setup if needed | 19:57 |
| *** EmilienM has joined #zuul | 19:57 | |
| pabelanger | tobiash: in this use case, we don't want to co-location nodepool-launcher with zuul-executor. We could, but if reason is to only ssh-keyscan, I think the toggle option is better | 19:58 |
| tobiash | pabelanger: ok, then I think there should be a note in the docs about this change in behavior in the docs such this needs to be handled by the base job in this case | 20:01 |
| pabelanger | tobiash: yah, we can add a NOTE about if you are expecting ssh host checking, then keys would need to be setup | 20:02 |
| pabelanger | tobiash: in the case of rcarrillocruz right now, there is no zuul for his network testing so that would need be added into his tooling | 20:02 |
| tobiash | pabelanger: ah oky I'd expect a note about key checking but more important about the fact that nodepool could hand out not fully booted and accessible nodes | 20:05 |
| pabelanger | tobiash: right, that would be the case, it would be up to zuul or something else to validate the networking is working (eg: ssh-keyscan) and if not, decline the node and get a new one | 20:06 |
| *** rlandy is now known as rlandy|brb | 20:09 | |
| *** JasonCL has joined #zuul | 20:13 | |
| *** JasonCL has quit IRC | 20:14 | |
| *** JasonCL has joined #zuul | 20:19 | |
| *** JasonCL has quit IRC | 20:27 | |
| *** JasonCL has joined #zuul | 20:28 | |
| *** acozine1 has joined #zuul | 20:33 | |
| *** JasonCL has quit IRC | 20:40 | |
| *** JasonCL has joined #zuul | 20:41 | |
| *** rlandy|brb is now known as rlandy | 20:41 | |
| *** JasonCL has quit IRC | 20:45 | |
| *** JasonCL has joined #zuul | 20:47 | |
| *** dkranz has quit IRC | 20:47 | |
| corvus | LinuxJedi: thanks! i think mrhillsman had an etherpad collecting issues with the zuul-from-scratch doc -- if you have a sec to note the issues you ran into, that'd be great | 20:52 |
| * corvus digs up link | 20:52 | |
| LinuxJedi | sure thing. Will sort it my AM | 20:52 |
| corvus | LinuxJedi: https://etherpad.openstack.org/p/zuul-quickstart-and-zfs(zuul-from-scratch) | 20:52 |
| corvus | Shrews: ^ | 20:52 |
| Shrews | yeah, i saw that a while back | 20:54 |
| *** Wei_Liu has quit IRC | 20:57 | |
| *** hashar has quit IRC | 21:19 | |
| openstackgerrit | Matthew Thode proposed openstack-infra/nodepool master: add gentoo systemd image for testing https://review.openstack.org/556661 | 21:58 |
| corvus | it's zuul meeting time in #openstack-meeting-alt | 22:00 |
| openstackgerrit | Matthew Thode proposed openstack-infra/nodepool master: add gentoo systemd image for testing https://review.openstack.org/556661 | 22:01 |
| openstackgerrit | Matthew Thode proposed openstack-infra/nodepool master: add gentoo systemd image for testing https://review.openstack.org/556661 | 22:05 |
| openstackgerrit | Merged openstack-infra/zuul master: Add selinux flag to docker command https://review.openstack.org/556582 | 22:07 |
| openstackgerrit | Matthew Thode proposed openstack-infra/nodepool master: add gentoo systemd image for testing https://review.openstack.org/556661 | 22:20 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool master: Add host-key-checking option to openstack providers https://review.openstack.org/556624 | 22:29 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Revert "Add support for STATSD_IPV6" https://review.openstack.org/556672 | 22:31 |
| corvus | pabelanger, ianw, jhesketh: ^ | 22:32 |
| mnaser | i'm blanking out | 22:36 |
| mnaser | what do we call the strategy of merging that zuul uses | 22:36 |
| mnaser | the one where we test things all together then pop them out if one fails and restart the stack | 22:36 |
| mnaser | p..something | 22:36 |
| Shrews | speculative? | 22:36 |
| mnaser | oh yes | 22:36 |
| *** pabelanger has quit IRC | 23:01 | |
| *** pabelanger has joined #zuul | 23:01 | |
| *** tosky has quit IRC | 23:08 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool master: DNM: testing gentoo nodepool-dsvm https://review.openstack.org/556679 | 23:12 |
| openstackgerrit | Merged openstack-infra/nodepool master: Revert "Add support for STATSD_IPV6" https://review.openstack.org/556672 | 23:44 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!