| *** tosky has quit IRC | 00:01 | |
| openstackgerrit | Ian Wienand proposed zuul/nodepool master: Require dib 3.5.0 https://review.opendev.org/c/zuul/nodepool/+/767438 | 00:24 |
|---|---|---|
| *** rfolco has joined #zuul | 00:38 | |
| *** wuchunyang has joined #zuul | 00:55 | |
| *** wuchunyang has quit IRC | 01:15 | |
| *** rfolco has quit IRC | 01:20 | |
| *** hamalq_ has quit IRC | 01:24 | |
| *** evrardjp has quit IRC | 01:24 | |
| *** Pilou has quit IRC | 01:24 | |
| *** mnasiadka has quit IRC | 01:24 | |
| *** parallax has quit IRC | 01:24 | |
| *** Open10K8S has quit IRC | 01:24 | |
| *** jbryce has quit IRC | 01:24 | |
| *** leoluk has quit IRC | 01:24 | |
| *** wuchunyang has joined #zuul | 01:28 | |
| *** hamalq_ has joined #zuul | 01:30 | |
| *** evrardjp has joined #zuul | 01:30 | |
| *** Pilou has joined #zuul | 01:30 | |
| *** mnasiadka has joined #zuul | 01:30 | |
| *** parallax has joined #zuul | 01:30 | |
| *** Open10K8S has joined #zuul | 01:30 | |
| *** jbryce has joined #zuul | 01:30 | |
| *** leoluk has joined #zuul | 01:30 | |
| *** wuchunyang has quit IRC | 01:54 | |
| *** sanjayu_ has quit IRC | 02:02 | |
| *** sanjayu_ has joined #zuul | 02:03 | |
| *** hamalq_ has quit IRC | 03:00 | |
| *** bhavikdbavishi has joined #zuul | 03:08 | |
| *** rlandy|bbl is now known as rlandy | 03:59 | |
| *** bhavikdbavishi has quit IRC | 04:02 | |
| *** wuchunyang has joined #zuul | 04:03 | |
| *** mgoddard has quit IRC | 04:03 | |
| *** wuchunyang has quit IRC | 04:08 | |
| *** rlandy has quit IRC | 04:10 | |
| *** bhavikdbavishi has joined #zuul | 04:17 | |
| openstackgerrit | Merged zuul/nodepool master: Require dib 3.5.0 https://review.opendev.org/c/zuul/nodepool/+/767438 | 04:21 |
| *** wuchunyang has joined #zuul | 04:47 | |
| *** wuchunyang has quit IRC | 04:52 | |
| *** mgoddard has joined #zuul | 05:05 | |
| *** bhavikdbavishi has quit IRC | 05:25 | |
| ianw | restarted builders with ^ to hopefully get good centos-8 builds out | 05:29 |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #zuul | 05:33 | |
| *** mgoddard has quit IRC | 06:00 | |
| *** bhavikdbavishi has joined #zuul | 06:04 | |
| *** vishalmanchanda has joined #zuul | 06:17 | |
| *** reiterative has quit IRC | 06:22 | |
| *** reiterative has joined #zuul | 06:23 | |
| *** bhavikdbavishi has quit IRC | 06:45 | |
| *** mgoddard has joined #zuul | 07:04 | |
| *** bhavikdbavishi has joined #zuul | 07:24 | |
| *** mgoddard has quit IRC | 08:01 | |
| *** jcapitao has joined #zuul | 08:03 | |
| *** jpena|off is now known as jpena | 08:07 | |
| *** hashar has joined #zuul | 08:10 | |
| *** bhavikdbavishi1 has joined #zuul | 08:28 | |
| *** bhavikdbavishi has quit IRC | 08:29 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 08:29 | |
| openstackgerrit | Daniel Pawlik proposed zuul/zuul-operator master: Replace Fedora 31 with Fedora 32 in zuul-operator-functional https://review.opendev.org/c/zuul/zuul-operator/+/767473 | 08:37 |
| *** mgoddard has joined #zuul | 09:02 | |
| *** tosky has joined #zuul | 09:12 | |
| *** bhavikdbavishi has quit IRC | 09:22 | |
| *** bhavikdbavishi has joined #zuul | 09:43 | |
| lyr | Hi. My nodepool-launcher is looping build / destroying instances in Openstack since it's unable to connect. How can I reproduce the SSH it's attempting to troubleshoot that ? | 09:43 |
| *** bhavikdbavishi1 has joined #zuul | 09:46 | |
| *** bhavikdbavishi has quit IRC | 09:47 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 09:47 | |
| *** nils has joined #zuul | 09:51 | |
| *** mgoddard has quit IRC | 10:01 | |
| *** mgoddard has joined #zuul | 10:05 | |
| lyr | I tried ssh -o "StrictHostKeyChecking=no" -i /var/lib/nodepool/.ssh/id_rsa zuul-worker@172.16.42.17, seems to be the wrong key | 10:20 |
| tobiash | lyr: nodepool doesn't try to login but just gathers the hostkeys | 10:20 |
| lyr | ok, but I should be able to SSH in nonetheless ? | 10:22 |
| avass | lyr: yeah if authorized_keys are set up correctly | 10:40 |
| avass | but all you need to do to emulator nodepool-launcher you just need to run ssh-keyscan | 10:41 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: Add more installation instructions https://review.opendev.org/c/zuul/zuul-client/+/765999 | 10:42 |
| *** tosky has quit IRC | 10:47 | |
| *** tosky has joined #zuul | 10:47 | |
| *** rpittau|afk is now known as rpittau | 10:59 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: encrypt: add --public-key argument https://review.opendev.org/c/zuul/zuul-client/+/765313 | 11:19 |
| *** hashar is now known as hasharLunch | 11:36 | |
| *** jcapitao is now known as jcapitao_lunch | 11:47 | |
| *** rfolco has joined #zuul | 12:16 | |
| *** wuchunyang has joined #zuul | 12:31 | |
| *** jpena is now known as jpena|lunch | 12:36 | |
| *** wuchunyang has quit IRC | 12:36 | |
| *** piotrowskim has joined #zuul | 12:38 | |
| *** rlandy has joined #zuul | 13:00 | |
| avass | Is it possible to interpolate the nodepool config from environment like zuul.conf does? I believe I've seen it but can't find it in the docs now | 13:04 |
| avass | oh or maybe secrets could just be part of the environment duh | 13:08 |
| tobiash | avass: afaik it's possible (unless the changes did merge) | 13:09 |
| avass | tobiash: do you mean not possible? | 13:10 |
| tobiash | it should be possible, but I don't find the according changes right now | 13:10 |
| tobiash | https://review.opendev.org/c/zuul/zuul/+/579091 | 13:11 |
| tobiash | hrm, abandoned | 13:11 |
| avass | that's zuul/zuul though | 13:12 |
| avass | but it should work by adding secrets to the pods environment right? unless nodepool removes those for any reason | 13:12 |
| *** jcapitao_lunch is now known as jcapitao | 13:12 | |
| tobiash | here it is: https://review.opendev.org/c/zuul/nodepool/+/719599 | 13:13 |
| tobiash | and zuul should have the same mechanism, just didn't find the change | 13:14 |
| tobiash | and guess what, that change misses a doc change :/ | 13:14 |
| avass | ah that explains it | 13:15 |
| tobiash | see https://review.opendev.org/c/zuul/nodepool/+/719599/5/nodepool/tests/fixtures/config_validate/good.yaml for the syntax | 13:15 |
| avass | looks like the same syntax as configparser | 13:15 |
| avass | oh but without the 's' at the end | 13:15 |
| tobiash | keep in mind that the vars must be prefixed with NODEPOOL_ | 13:16 |
| avass | alright | 13:16 |
| tobiash | zuul supports the same: https://opendev.org/zuul/zuul/src/branch/master/zuul/cmd/__init__.py#L135 | 13:18 |
| avass | yep we're already using that | 13:18 |
| tobiash | zuul also has it documented, so just the nodepool docs are missing | 13:19 |
| *** jpena|lunch is now known as jpena | 13:31 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/nodepool master: config: support for interpollation with a closing `)s` and add doc https://review.opendev.org/c/zuul/nodepool/+/767556 | 13:34 |
| *** bhavikdbavishi has quit IRC | 13:36 | |
| *** hasharLunch is now known as hashar | 13:47 | |
| *** zenkuro has joined #zuul | 14:23 | |
| *** bhavikdbavishi has joined #zuul | 14:25 | |
| *** bhavikdbavishi1 has joined #zuul | 14:27 | |
| *** bhavikdbavishi has quit IRC | 14:29 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 14:29 | |
| *** SpamapS has joined #zuul | 15:09 | |
| *** SpamapS has quit IRC | 15:10 | |
| *** SpamapS has joined #zuul | 15:12 | |
| *** hashar has quit IRC | 15:30 | |
| *** zenkuro has quit IRC | 15:31 | |
| *** zenkuro has joined #zuul | 15:31 | |
| *** zenkuro has quit IRC | 15:36 | |
| *** bhavikdbavishi has quit IRC | 15:36 | |
| *** zenkuro has joined #zuul | 15:36 | |
| *** bhavikdbavishi has joined #zuul | 15:38 | |
| *** bhavikdbavishi has quit IRC | 15:50 | |
| *** bhavikdbavishi has joined #zuul | 15:53 | |
| *** wuchunyang has joined #zuul | 16:32 | |
| *** wuchunyang has quit IRC | 16:37 | |
| *** hashar has joined #zuul | 16:38 | |
| pabelanger | fungi: mordred: do you mind looking at https://review.opendev.org/c/zuul/zuul-jobs/+/767361 that will get us working with tox_siblings | 16:39 |
| corvus | pabelanger: can you catch me up on how using tox facilitates siblings with non-python projects? | 16:39 |
| pabelanger | corvus: sure, the basic idea is, we have container build tooling in zuul. That project isn't python, but we use tox to manage the python dependencies needed to builder the containerfile: https://github.com/ansible/network-ee | 16:42 |
| mordred | corvus: it's a non-python project that has python depends - so the tox env installs all the python depends into a virtualenv - but the repo in question doesn't itself have python code to install into the virtualenv | 16:42 |
| * mordred shuts up and lets pabelanger drive :) | 16:42 | |
| pabelanger | to avoid the need of installing a setup.cfg file, for the project name, I've updated the tox role to support passing it in via zuul.project.short_name | 16:42 |
| corvus | oh interesting | 16:43 |
| pabelanger | at the moment, we don't have a good story for zuul job, that needs cross-project python dependencies, outside of using tox | 16:43 |
| corvus | pabelanger: ansible-builder is the thing you want to have CRD for? | 16:47 |
| corvus | pabelanger: tox installs https://github.com/ansible/network-ee/blob/main/requirements.txt as specified in https://github.com/ansible/network-ee/blob/main/tox.ini and that's what's used to build the container image? | 16:48 |
| pabelanger | corvus: yes, ansible-builder right now | 16:49 |
| corvus | pabelanger: there might be others in the future? like some dependency that ansible-builder has? | 16:50 |
| pabelanger | yes! that is possible | 16:50 |
| pabelanger | but right now, we don't have that dependency | 16:50 |
| corvus | pabelanger: but to be clear, it's not the stuff in the container right? it's not https://github.com/ansible/network-ee/blob/main/requirements_combined.txt | 16:50 |
| pabelanger | corvus: right, that is what ansible-builder generates | 16:50 |
| pabelanger | and it uses | 16:50 |
| corvus | pabelanger: 767361 makes sense to me, and sounds reasonable since ultimately ansible-buildir is a python program. an alternative might be to make an ansible-builder role that either installs ansible-builder from pypi or from the zuul src dir if it's present. i don't see a reason to do that if you're happy with tox though. | 16:52 |
| corvus | pabelanger: thanks for the catch up :) | 16:52 |
| *** bhavikdbavishi has quit IRC | 16:52 | |
| pabelanger | yes, I think a role is fair too | 16:53 |
| corvus | pabelanger: out of curiosity is quay.io/ansible/python-builder a copy/fork of opendev's? | 16:54 |
| pabelanger | I thought of hacking on some sort of generic virtualenv zuul job, that used ZUUL_SIBLINGS logic to prepare a virtualenv for the user | 16:54 |
| *** ikhan has quit IRC | 16:54 | |
| *** jcapitao has quit IRC | 16:54 | |
| pabelanger | corvus: it is, yes. except centos/ubi based | 16:54 |
| corvus | pabelanger: neat :) | 16:55 |
| pabelanger | yah, people in ansible are really excited by it. It helps solves the distro package / pip package issue nicely | 16:55 |
| pabelanger | using bindep.txt / requirements / upper-constraints | 16:55 |
| *** ikhan has joined #zuul | 16:58 | |
| corvus | mhu: 765999 approved but see comment from tobiash (i think re-ordering those is a good idea) | 17:04 |
| *** piotrowskim has quit IRC | 17:06 | |
| mhu | corvus, thanks, looking | 17:07 |
| *** jpena is now known as jpena|off | 17:08 | |
| mordred | pabelanger, corvus: I don't have any bandwith so have a low opinion threshold - but is there any value in considering an opendev/python-base:ubi variant so there doens't have to be a fork? I'm not usually a fan of multi-distro container images for things like zuul - but the argument could certainly be made that it's a valid use case for a base image | 17:10 |
| corvus | mordred: probably worth moving this to #opendev; but if the tripleo/rdo folks might be interested in using it then sounds like it could be a cool collaboration with ansible | 17:13 |
| mordred | corvus: yah - I was going to ask there- but it didn't seem pabelanger was in there | 17:13 |
| corvus | i dunno if there's an advantage to using it for zuul? | 17:13 |
| pabelanger | let me join #opendev | 17:15 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: Reorganize the installation documentation, specify versions https://review.opendev.org/c/zuul/zuul-client/+/767592 | 17:18 |
| mhu | tobiash, corvus ^ | 17:18 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Zuul Cache role with s3 implementation. https://review.opendev.org/c/zuul/zuul-jobs/+/764808 | 17:19 |
| corvus | mhu: i left a -1 on 765313 (cc tobiash) | 17:19 |
| openstackgerrit | Tristan Cacqueray proposed zuul/nodepool master: config: support for interpollation with a closing `)s` and add doc https://review.opendev.org/c/zuul/nodepool/+/767556 | 17:21 |
| openstackgerrit | Merged zuul/zuul-client master: Add more installation instructions https://review.opendev.org/c/zuul/zuul-client/+/765999 | 17:25 |
| openstackgerrit | Merged zuul/zuul-client master: API, CLI: Allow use of tenant-scoped API URLs https://review.opendev.org/c/zuul/zuul-client/+/765203 | 17:28 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: encrypt: add --public-key argument https://review.opendev.org/c/zuul/zuul-client/+/765313 | 17:37 |
| *** zenkuro has quit IRC | 17:42 | |
| *** zenkuro has joined #zuul | 17:42 | |
| *** hamalq has joined #zuul | 17:50 | |
| *** hamalq_ has joined #zuul | 17:52 | |
| *** hamalq has quit IRC | 17:56 | |
| *** vishalmanchanda has quit IRC | 18:06 | |
| *** rpittau is now known as rpittau|afk | 18:20 | |
| *** hashar is now known as hasharDinner | 19:03 | |
| *** nils has quit IRC | 19:09 | |
| *** AJaeger has joined #zuul | 19:18 | |
| *** AJaeger has quit IRC | 19:56 | |
| *** hasharDinner is now known as hashar | 20:11 | |
| *** rfolco has quit IRC | 21:09 | |
| *** hashar has quit IRC | 21:14 | |
| *** josefwells has joined #zuul | 21:18 | |
| josefwells | Hello zuulers. I'm back again this week.. Trying now to just use the examples tutorial straight up with gerrit and the whole bit | 21:20 |
| josefwells | my web interface doesn't seem to be coming up: | 21:20 |
| josefwells | web_1 | 2020-12-17 21:19:07,289 WARNING zuul.zk.ZooKeeper: Retrying zookeeper connection | 21:20 |
| fungi | josefwells: https://zuul-ci.org/docs/zuul/discussion/components.html if you're firewalling between the server where zuul-web is running and the zookeeper daemon, make sure the right ports are open | 21:21 |
| josefwells | I don't see anything crazy in the zookeeper logs, it seems to ahve started | 21:21 |
| josefwells | this is all using the docker-compose | 21:22 |
| fungi | ahh | 21:22 |
| fungi | so they're running on the same host | 21:22 |
| josefwells | yes | 21:22 |
| fungi | just in different containers | 21:22 |
| josefwells | right | 21:22 |
| fungi | josefwells: and you're deploying tip of master, or the most recent versioned release? | 21:23 |
| josefwells | yes, as of a couple days ago | 21:23 |
| fungi | current master (soon to be 4.0.0) requires tls for zookeeper communication, so you might double-check the configuration against this: https://zuul-ci.org/docs/zuul/howtos/zookeeper.html#zk-encrypted-connections | 21:25 |
| josefwells | ok, I'll pull and check the config as well | 21:25 |
| fungi | basically these are the two things i can think of which would produce that error. well or the zookeeper daemon not running at all | 21:28 |
| fungi | or other zk connection details being mismatched for some reason | 21:28 |
| mordred | fungi: yah - although I'd still expect the docker-compose to work since we run it in the gate | 21:31 |
| fungi | right | 21:31 |
| josefwells | I can dump out my zk logs if that would help | 21:32 |
| fungi | point being we run all of that on every commit, so something's different | 21:32 |
| josefwells | gerrit came up, so I didn't notice till I got to the point where I was supposed to see stuff on web | 21:32 |
| josefwells | not sure how your check ensures that everything is really up | 21:33 |
| fungi | could the "Retrying zookeeper connection" warning be a red herring and the zuul-web container merely races the zk container and so keeps retrying after start until it can connect? | 21:34 |
| josefwells | Last week I was trying this with a modified setup that used our enterprise github, but I ended up with trouble getting my static external nodes to have working ssh | 21:34 |
| josefwells | so I figured I would go back to basics | 21:34 |
| josefwells | notably last week, I just left off the zoo.cfg and all the cert stuff, because I didn't have it last time I messed with this | 21:34 |
| josefwells | and I didn't hit these problems with zookeeper | 21:35 |
| fungi | are you sure zuul-web didn't connect to zk? does it continually log that it's retrying to connect? | 21:37 |
| fungi | or was it just the once or a few times? | 21:37 |
| josefwells | fungi: if you mean a transient thing, no, once the network comes up (wait-to-start-gearman finishes) and I get endless zookeeper connection retries | 21:38 |
| fungi | got it | 21:38 |
| fungi | because yeah these different daemons are hopefully robust against asynchronous starting (we consider it a bug if they're not), but if it never establishes a connection then that suggests connectivity issues or some sort of configuration mismatch (or zookeeper isn't starting at all?) | 21:39 |
| josefwells | Yes, I saw the removal of ansible from some containers and some better checking for databases to be not just up, but functional, etc | 21:40 |
| josefwells | I had a patch when this first was added to help with athat | 21:40 |
| *** zenkuro has quit IRC | 21:51 | |
| josefwells | sigh.. deleting all containers/volumes/networks and bringing it all back up seems to work | 22:10 |
| josefwells | I love computers | 22:10 |
| corvus | josefwells: you turned it off and back on again! :) | 22:11 |
| fungi | nondeterminism as a service | 22:12 |
| fungi | heisenberg uncertainty computing | 22:12 |
| corvus | josefwells: but seriously, maybe there was some local state from before a zk-related change that wasn't accounted for | 22:12 |
| corvus | josefwells: for example, i think the quickstart now runs zk with tls, but didn't before; we have release notes for version upgrades like that, but for a small test installation running latest master it may just be easier to rebuild it | 22:14 |
| josefwells | corvus: for sure. I'm still no docker expert, so it was only way in the back of my mind that I thought the volumes or other bits might be interfering here | 22:16 |
| josefwells | This whole process is "take it apart, put it back together", I've really built a career on those simple steps. | 22:17 |
| josefwells | sometimes there are extra screws, etc.. clearly overengineered to begin with | 22:17 |
| * corvus hands josefwells a box of random screws | 22:18 | |
| * josefwells dumps the box in my other box | 22:18 | |
| mordred | the best part about boxes of screws is that no matter how many screws are in the box, it never contains the screw I need | 22:28 |
| fungi | i've tapped some die blocks with various diameters and thread pitches so i can sort mine | 22:30 |
| fungi | except making the machine screw checker was the fun part, not actually sorting the screws. so instead i still have a giant pile of assorted machine screws and end up just ordering new packs of what i need | 22:31 |
| fungi | though i do at least have an indexed set of baggies with reference screws i can use to work out what i need to order when i have something missing some screws | 22:32 |
| *** wuchunyang has joined #zuul | 22:35 | |
| *** wuchunyang has quit IRC | 22:40 | |
| *** rlandy has quit IRC | 23:23 | |
| *** Tahvok has quit IRC | 23:24 | |
| *** Tahvok_ has joined #zuul | 23:24 | |
| *** Tahvok_ is now known as Tahvok | 23:24 | |
| *** tosky has quit IRC | 23:43 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!