| *** eernst has quit IRC | 00:00 | |
| mordred | clarkb: is etherpad.o.o runniner older code? | 00:01 |
|---|---|---|
| clarkb | mordred: no, reading the docs this is still supported | 00:01 |
| mordred | *awesome* | 00:01 |
| clarkb | http://etherpad.org/doc/v1.7.0/#index_custom_static_files on current etherpad | 00:01 |
| clarkb | its just you have to create the dir I guess | 00:01 |
| clarkb | I'll update the puppet | 00:01 |
| *** sthussey has quit IRC | 00:02 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: [wip] ubuntu element - ignore missing openstack repos https://review.openstack.org/610430 | 00:02 |
| *** rlandy is now known as rlandy|bbl | 00:03 | |
| ianw | #status log nl04 in emergency while I fiddle with ovh-gra quotas to see what works | 00:05 |
| openstackgerrit | Clark Boylan proposed openstack-infra/puppet-etherpad_lite master: Ensure the custom js dir is created https://review.openstack.org/610755 | 00:05 |
| openstackstatus | ianw: finished logging | 00:05 |
| clarkb | ianw: mordred ^ I am hopeful that is the last fix we need for new etherpad lite | 00:05 |
| clarkb | I actually need to go look at making dinner nowish. But I think you can safely approve that and the server will puppet overnight sometime and I'll get back to it in the morning | 00:06 |
| ianw | clarkb: cool, will have a look | 00:06 |
| clarkb | I wonder how this ends up working with the rspec tests. Maybe we use older version there that does work? | 00:09 |
| clarkb | probably worth looking into | 00:09 |
| dmsimard | clarkb: gerritbot fixed for #ara, thanks ! | 00:12 |
| *** agopi has quit IRC | 00:29 | |
| *** agopi has joined #openstack-infra | 00:30 | |
| *** jamesmcarthur has joined #openstack-infra | 00:39 | |
| *** longkb has joined #openstack-infra | 00:40 | |
| *** smarcet has quit IRC | 00:45 | |
| smcginnis | Fixed for #openstack-cinder too! | 00:46 |
| *** jamesmcarthur has quit IRC | 00:47 | |
| *** ssbarnea|bkp2 has quit IRC | 00:48 | |
| *** ssbarnea has joined #openstack-infra | 00:48 | |
| *** agopi_ has joined #openstack-infra | 00:49 | |
| *** agopi has quit IRC | 00:50 | |
| *** jamesmcarthur has joined #openstack-infra | 00:51 | |
| *** jamesmcarthur has quit IRC | 00:52 | |
| *** jamesmcarthur has joined #openstack-infra | 00:52 | |
| *** smarcet has joined #openstack-infra | 00:53 | |
| *** smarcet has quit IRC | 00:56 | |
| openstackgerrit | Merged openstack-infra/gerritbot master: Don't autojoin channels https://review.openstack.org/610745 | 01:05 |
| *** rkukura has joined #openstack-infra | 01:06 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/zuul-website master: Add three more case studies https://review.openstack.org/610757 | 01:06 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/zuul-website master: Add three more case studies https://review.openstack.org/610757 | 01:10 |
| *** slaweq has joined #openstack-infra | 01:11 | |
| *** slaweq has quit IRC | 01:16 | |
| *** mrsoul has joined #openstack-infra | 01:26 | |
| tonyb | Where in system-config? can I see the size (RAM/Disk) of the VMs used for jobs? | 01:28 |
| *** rlandy|bbl is now known as rlandy | 01:28 | |
| clarkb | tonyb https://docs.openstack.org/infra/manual/testing.html are the docs on test envs | 01:30 |
| tonyb | clarkb: Thanks | 01:31 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul-jobs master: Revert "Add ensure-python role" https://review.openstack.org/610758 | 01:32 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul-jobs master: Revert "Add ensure-python role" https://review.openstack.org/610758 | 01:33 |
| gmann | clarkb: AJaeger yes, this will cover the both scenario. 1. testing openstack with lib master version by lib-forward-testing and lib-forward-testing-python3 2. testing openstack with lib pypi version in integrated tempest-full job | 01:35 |
| pabelanger | AJaeger: frickler: coreycb (I think), I'd like to consider a revert of ^. It seems to duplicate functionality we already support with bindep | 01:35 |
| *** openstackgerrit has quit IRC | 01:35 | |
| pabelanger | clarkb: also, sine you on original review | 01:36 |
| ianw | clarkb: are we tracking the "openstack.cloud.exc.OpenStackCloudCreateException: Error creating server: 083df9db-4db9-447f-a9df-73bf26a0dc9d" in ovh? | 01:37 |
| ianw | bhs1 is hitting this all the time | 01:37 |
| *** yamamoto has quit IRC | 01:41 | |
| *** yamamoto has joined #openstack-infra | 01:41 | |
| clarkb | ianw not that I know of. | 01:41 |
| fungi | pabelanger: how is it a duplicate? | 01:42 |
| fungi | are you suggesting that every project which wants to run python 3.7 tests with a tox job needs an entry in their bindep.txt so that a job which says it's going to run tests under python 3.7 will actually have python 3.7 available? | 01:43 |
| pabelanger | fungi: it provides the same functionality that bindep could, and our tox jobs parent to unittests, which include bindep by default | 01:44 |
| pabelanger | fungi: yes | 01:44 |
| pabelanger | I think that is a sane approach | 01:44 |
| fungi | i think it's not sane at all | 01:47 |
| pabelanger | why? we do it today for python3-devel | 01:47 |
| pabelanger | for example: https://git.zuul-ci.org/cgit/zuul/tree/bindep.txt#n32 | 01:48 |
| pabelanger | i don't think it is a good idea to have 2 places to install python from | 01:48 |
| fungi | which is why i had encouraged coreycb to abandon mass patching to add python3.7 entries to the bindep.txt file of every single python-based project we host | 01:48 |
| fungi | bindep.txt is not a tool to control our ci system | 01:48 |
| fungi | it's instructions to a developer to let them know what unusual packages they should install in their development environment to be able to test changes they're writing | 01:49 |
| pabelanger | that is now how I understood it, otherwise, we can just have a global bindep_fallback.txt again for tox specific things over pre ansible role package installs | 01:49 |
| fungi | if they choose to run `tox -e py37` i think 1. they ought to know they need a python 3.7 interpreter installed, and 2. the error message tox provides if they don't is quite clear | 01:49 |
| fungi | this isn't a global bindep fallback | 01:50 |
| fungi | it's like deciding that our pypi upload jobs should install twine | 01:50 |
| fungi | we don't make every project declare that there's a ci job which will need twine installed | 01:51 |
| fungi | similarly we should make every single project indicate that to run unit tests under python 3.7 you need python3.7 installed | 01:51 |
| fungi | er, shouldn't | 01:51 |
| pabelanger | if we don't want to mass update bindep.txt file, i can understand that, we could then just installed python3.7 via DIB elements, like we do for python2.7 and python3.6 today. | 01:52 |
| fungi | distros provide packages for things like tox. why do we have an ensure-tox role instead of just making all projects put tox in their bindep.txt? | 01:52 |
| pabelanger | The issue I have with the patch, is we now have 2 places where somebody can define a python package to be used. For reasons I thought, bindep.txt was for the CI system to manage OS package dependencies, not expanding out roles in also include that | 01:53 |
| *** nhicher has quit IRC | 01:54 | |
| fungi | our ci jobs take advantage of the fact that projects often declare additional distro packages they need in a bindep.txt file, but any time we're adding things to bindep.txt just to get a ci job working chances are we're doing it wrong | 01:54 |
| fungi | if someone puts tox in their bindep.txt, then there are two places they can define a tox package to use | 01:55 |
| pabelanger | I actually disagree with that statement, bindep.txt at least for how we are using it today in ansible-network, is a good place to get CI depends and application dependencies installed. Otherwise, I am not sure why we have python3-dev in bindep.txt today. | 01:56 |
| fungi | i saw ensure-python as a potential future reusable role where we can provide ways to get interesting versions of python onto multiple distros (not just via distro packages, that one happens to be the initial case only) | 01:56 |
| fungi | having python3-dev, which is not installed by default when you install a python3 interpreter, is needed if the project is using any python dependencies built from sdist and linked against libpython3. it's not necessary for every project | 01:58 |
| fungi | projects with only pure python dependencies shouldn't need it, for example | 01:59 |
| fungi | also, this is proposed to zuul's stdlib, so #zuul is likely a better place to be having this debate. it's not an openstack-only use case | 02:00 |
| clarkb | the reason to have python3-dev in bindep is to say this needs python3 to run. To test with a specific version of python to ensure coverage of that version of pythin the job should handle that | 02:01 |
| clarkb | they are different concerns | 02:01 |
| *** rh-jelabarre has joined #openstack-infra | 02:02 | |
| *** armax has quit IRC | 02:03 | |
| ianw | i'd always considered it binary dependencies of the projects in requirements.txt | 02:04 |
| pabelanger | replied to orginal patchset | 02:06 |
| pabelanger | ianw: yes, that is how I understood it too. | 02:07 |
| fungi | it also seems strange to me, taken from the main bindep use case which is communicating distro package dependencies to the developer, that we would list that they need python3.7 but not any other python versions just because there happens to be some platform somewhere which preinstalls one of those other versions | 02:07 |
| pabelanger | Did we discuss updating our images to provide python3.7? | 02:08 |
| fungi | so a python3.7 bindep entry for ubuntu bionic does tell them that they might want python3.7 installed to run py37 tox tests, but doesn't tell them that to run py35 tox tests they need to be running on ubuntu xenial | 02:08 |
| clarkb | they are complimentary things. If I install zuul I want python3 installed via bindep for me. If I test zuul and I want to test python3.7 the job should verify and ensure that | 02:08 |
| clarkb | pabelanger: we dont need ti its available on bionic | 02:09 |
| *** hongbin has joined #openstack-infra | 02:09 | |
| pabelanger | I'm okay to discuss it more in #zuul or ML, I'm not really a fan of growing out ensure-python to support all the possible distros to be honest. But agree, if updating all the projects bindep.txt doesn't scale well, maybe going back to centralized bindep is an option | 02:14 |
| clarkb | but centralized bindeo doesnt really solve this either? | 02:14 |
| clarkb | the problem is there are 3 pythons on bionic | 02:14 |
| clarkb | 2.7, 3.6 and 3.7 | 02:14 |
| pabelanger | just noticed we have a new task in ansible-network jobs, and at first glance looked to duplicate bindep functionality | 02:14 |
| fungi | and as i said earlier, ensure-python can certainly be extended to do other things like install unusual python versions via pyenv or make&&make install, the first case for it just happens to be providing a way to declare a nonstandard python interpreter distro package for jobs which need thaht version | 02:14 |
| clarkb | to run unittests in a specific version you want to ensure that version is the one being used | 02:15 |
| clarkb | that seems out of the scope of bindep which is instead: give me a python3 | 02:15 |
| *** rh-jelabarre has quit IRC | 02:15 | |
| fungi | i had hoped that something along the lines of ensure-python was already existing in ansible galaxy, but that turns out to not be the case | 02:16 |
| pabelanger | confused why you could be more specific in bindep profile, for python36 environment: eg: bindep_profile: test python36' | 02:17 |
| *** diablo_rojo has quit IRC | 02:17 | |
| clarkb | pabelanger: because then you have to standardize arbitrary stirngs across hundreds of repos | 02:17 |
| clarkb | and then mordred makes a change that conflicts with non standard strings and things break | 02:17 |
| clarkb | see the "build" profile | 02:18 |
| pabelanger | clarkb: sure, but wasn't that the idea of pushing bindep or jobs into projects, we'd make those changes across all the repos vs central location | 02:18 |
| fungi | and that gets back to abusing bindep to drive the ci system rather than taking advantage of the fact that projects are using bindep and installing the things they tell their developers they should install | 02:18 |
| *** jamesmcarthur has quit IRC | 02:18 | |
| clarkb | pabelanger: its still entirely that way with ensure-python | 02:19 |
| clarkb | its config per project that asserts a value | 02:19 |
| *** rkukura has quit IRC | 02:19 | |
| pabelanger | I guess, I just don't see putting CI requirement into bindep as abusing bindep. We do think it abuse, we need to document that correctly. | 02:21 |
| clarkb | Im not sure its abause as much as impossible to police as we've already proven with "build" | 02:22 |
| promehteanfire | someone mind taking a second look at this for loop, getting a perm issue | 02:22 |
| promehteanfire | http://logs.openstack.org/68/604268/15/check/dib-dsvm-functests-python2-ubuntu-trusty-extras/6fe000c/job-output.txt.gz#_2018-10-16_01_19_53_069707 | 02:23 |
| clarkb | and so if a jobs reason to exist is to test with a specific version of python it is reasonable to assert that in the job | 02:23 |
| promehteanfire | ah, redirects don't work in sudo (not that way) | 02:23 |
| ianw | so setting max-servers to 1 in gra, it already seems to have decide to launch about 5 nodes | 02:24 |
| fungi | when i say abusing, i mean relying on bindep to communicate things to our ci system without considering what it's communicating to the primary audience of the tool and the bindep.txt file in a repo | 02:26 |
| *** openstackgerrit has joined #openstack-infra | 02:26 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder master: enable caching for gentoo builds https://review.openstack.org/604268 | 02:26 |
| promehteanfire | one of these days | 02:26 |
| *** annp has joined #openstack-infra | 02:26 | |
| *** jamesmcarthur has joined #openstack-infra | 02:27 | |
| fungi | but also, when there's an entry you expect to be present in the bindep.txt file of, say, every single python-based project, chances are it's not one you need to communicate to developers and so only noise we could be solving by making jobs do what is obvious | 02:28 |
| fungi | (the ensure-twine role as an example) | 02:29 |
| promehteanfire | question, if I can support all versions of python, should I? (in the image)? | 02:29 |
| promehteanfire | atm, I think it's hard coded to py35(only), but I could change it to anything else | 02:30 |
| clarkb | promehteanfire: I think we'd like to avoid baking that into the image but have some python2 stuff for historical reasons | 02:30 |
| clarkb | so making it possible to install arbitrary pythons without installing them all might be a good thing? | 02:30 |
| clarkb | pabelanger: is there a specific concern with allowing jobs ti specify their needs directly? | 02:31 |
| clarkb | one I could see is not testing that bindep works generally (but functional or integrarion tests might be better for that?) | 02:32 |
| promehteanfire | this is how I install the pythons people want | 02:32 |
| promehteanfire | https://github.com/openstack/diskimage-builder/blob/master/diskimage_builder/elements/gentoo/pre-install.d/02-gentoo-04-install-desired-python | 02:32 |
| promehteanfire | it's not quite so easy as just use bindep for me at least | 02:32 |
| pabelanger | clarkb: just bindep concern, so far on ansible-net side were looking to leverage bindep.txt for all dependencies. But doesn't break anything, was just surprised by the new task | 02:33 |
| pabelanger | I think I was just more surprised to see the change, I missed the discussion | 02:34 |
| promehteanfire | it's actually a little bit better than that now, though py37 will need similiar workarounds | 02:36 |
| promehteanfire | https://review.openstack.org/#/c/608576/4/diskimage_builder/elements/gentoo/pre-install.d/02-gentoo-04-install-desired-python | 02:36 |
| clarkb | pabelanger: and using python specific bindep envs for a complete distros x release x python ver matrix? | 02:36 |
| clarkb | I mostly want to avoid making the decsion to add a new python version to the test matrix complicated for individual projects and require them to have more than a generic python3 dep | 02:37 |
| clarkb | if you want to add more you can but not sure it should be required | 02:37 |
| pabelanger | clarkb: not yet, we are still on python2.7 and python3.6, we skipped python3.5. However, agree once next fedora comes out, we too have to do the python3.7 dance | 02:38 |
| pabelanger | for that, i was planning more specific bindep_profiles | 02:38 |
| *** felipemonteiro has joined #openstack-infra | 02:42 | |
| fungi | pabelanger: the ensure-twine role is in the zuul-jobs repo, not the project-config repo | 02:43 |
| fungi | we also have ensure-babel, ensure-sphinx... | 02:43 |
| pabelanger | fungi: yah, was trying to say we don't have a shared job today that is using them in zuul-jobs, only in ozj | 02:44 |
| pabelanger | err, project-config | 02:44 |
| fungi | (these are things which _could_ be installed via listing them in bindep.txt or test-requirements.txt or whatever, but makes more sense for jobs which need those to make sure they get installed rather than relying on all the projects to dtrt) | 02:45 |
| fungi | pabelanger: i expect the only reason jobs in zuul-jobs don't use ensure-twine is because the credentials have to be provided in the same repo as the playbook using them | 02:47 |
| fungi | due to zuul's safety measures around secrets handling | 02:47 |
| pabelanger | Yup! Think that too. This is an interesting topic, because the tox jobs really are our first jobs that we have shared between 3 zuul systems, that I know of, openstack-infra, rdoproject and ansible-network. And really a first for sharing job content in this CI context :) | 02:48 |
| pabelanger | just to throw another angle into this topic, this could also have been handled in tools/test-setup.sh also | 02:50 |
| fungi | yep | 02:50 |
| clarkb | I actually like that the job that says I use python 3.7 actually ensures that | 02:51 |
| clarkb | it isnt left to some external system to make that true | 02:51 |
| clarkb | and the change as written shouldve worked for everyone else? | 02:51 |
| clarkb | thus didnt break anyone right? it only did the ensuring on platforms it knew how to do that for | 02:52 |
| fungi | but again, i think any logic or information which is applicable to a broad class of projects is not something we should be duplicating in every one of them. tools/test-setup.sh as a pattern was meant for performing unusual steps unique to a particular project or small set of projects | 02:52 |
| clarkb | ++ | 02:52 |
| pabelanger | clarkb: we are projected because python_version isn't set however, once we do set it, it will break for us, because that package syntax isn't valid for fedora or centos | 02:53 |
| fungi | tools/test-setup.sh was, at least initially, for prepopulating test database permissions and adding nonstandard filesystems | 02:54 |
| clarkb | pabelanger: but doesnt it guard against unsupported distros? I thought it did | 02:54 |
| pabelanger | clarkb: no | 02:54 |
| clarkb | in any case it isnt going to break anyone running tox already | 02:54 |
| clarkb | it is opt in | 02:54 |
| pabelanger | it will actually fail if we set python_version with fedora | 02:54 |
| fungi | it guards by expecting you to test the change defining a python_version value | 02:55 |
| pabelanger | it should be updated to have when statement for specific distro it supports, ubuntu according to docs | 02:55 |
| fungi | that seems like a reasonable addition, i agree | 02:55 |
| clarkb | or do thr first found include and have a default that noops | 02:56 |
| *** carl_cai has quit IRC | 02:56 | |
| fungi | keep in mind it will also break similarly if you declare a nonexistent python_version for the distro on which you're testing | 02:56 |
| clarkb | I find that easier to understand when its distro specific behavior | 02:56 |
| fungi | like, set python_version=3.5 in a job for an ubuntu-xenial node and it will break | 02:57 |
| clarkb | looks like fedora just has "python3" | 02:57 |
| clarkb | so you avoid this class of problem entirely | 02:57 |
| fungi | er, i mean set python_version=3.5 in a job for an ubuntu-bionic node | 02:57 |
| fungi | clarkb: zaneb was saying that on fedora he has packages available for 3.4, 3.5, 3.6 and 3.7 | 02:58 |
| clarkb | fungi: those dont appear to be in the main repo according to a quick package search | 02:58 |
| clarkb | but Im on a phone and dont have dnf available to give a concrete answer | 02:58 |
| fungi | also possible i misinterpreted him | 02:59 |
| fungi | on debian i can get a few via packages, but find it easier to just compile them all from source instead and have a symlink farm in ~/bin i use | 03:00 |
| clarkb | aha found them | 03:01 |
| clarkb | python36 and python37 with no punctuation | 03:01 |
| pabelanger | python37, python3 on fedora-28 | 03:01 |
| clarkb | ya 36 is only a package if python3 isnt 3.6 | 03:02 |
| *** ramishra has joined #openstack-infra | 03:02 | |
| fungi | so we could use some string mangling to strip out the . in the fedora logic branch | 03:02 |
| *** jamesmcarthur has quit IRC | 03:02 | |
| clarkb | so its actually more complicated on fedora :/ | 03:02 |
| fungi | but is the python3 package installed by default on fedora? | 03:03 |
| promehteanfire | iirc it's been the default for a few releases in fedora | 03:03 |
| *** zhangfei has joined #openstack-infra | 03:03 | |
| pabelanger | fungi: it only is because of DIB python dependency | 03:03 |
| pabelanger | fedora-cloud images, is it missing IIRC | 03:03 |
| pabelanger | promehteanfire: oh, I need to double check then | 03:04 |
| fungi | again, considering this from an opt-in perspective, you'd only need to set python_version if calling the job on a platform for which that interpreter is provided from a nonstandard package | 03:04 |
| fungi | or rather one not already installed by default | 03:04 |
| promehteanfire | I'm not sure if they went py3-only like I'm doing, but I think they can | 03:05 |
| pabelanger | last I looked, we need to update pip-and-virtualenv element to support single python version. It is 2 by default | 03:05 |
| fungi | one addition which would probably be sane is to check first for the existence of a pythonX.Y executable in the environment's patyh | 03:06 |
| promehteanfire | it'd be nice if we could support the distro's version of pip at least for init (pip can upgrade itself to a constraint if needed) | 03:07 |
| promehteanfire | that'd also allow us to constrain pip | 03:07 |
| *** bhavikdbavishi has joined #openstack-infra | 03:07 | |
| promehteanfire | fungi++ | 03:07 |
| pabelanger | promehteanfire: we're actually planing on remove pip from our images shortly for a use case. To make sure everything is packaged properly for testing in ansible. Not looking forward to that fun | 03:08 |
| fungi | well, pip upgrading itself from a distro package means pip and the distro package manager fighting over some of the same files | 03:08 |
| *** jamesmcarthur has joined #openstack-infra | 03:08 | |
| promehteanfire | fungi: true, but that generally happens anyway if using pip systemwide | 03:09 |
| * promehteanfire wishes that wasn't the case and venvs were used for everything | 03:09 | |
| fungi | yeah, i shudder at every case of that we're propagating | 03:09 |
| promehteanfire | I was told ymmv, expect breakage, when I showed the gentoo python team what openstack does with pip (system wide) | 03:10 |
| promehteanfire | there's a reason we disabled system-wide pip by default for our pip package | 03:10 |
| clarkb | the debian split with /usr/local addresses the vast majority if problems though | 03:11 |
| clarkb | unfortunately this sanity was not embraced more globally | 03:11 |
| *** slaweq has joined #openstack-infra | 03:11 | |
| promehteanfire | iirc we have site and dist packages in $PYTHON_LIBDIR | 03:11 |
| fungi | well, and then as a result pip grew the desire to uninstall packages outside its install path | 03:11 |
| pabelanger | I thought we fixed most of our system-wide pip usage when pip10 came out | 03:11 |
| pabelanger | or was that just for devstack | 03:12 |
| clarkb | pabelanger: no interest for that died out | 03:12 |
| clarkb | we whack a moled the problems | 03:12 |
| pabelanger | clarkb: Oh, what did we end up doing? | 03:12 |
| clarkb | and addressed individual issues | 03:12 |
| clarkb | we just stopped installing system and pip packages that conflicted aiuu | 03:12 |
| pabelanger | ah | 03:12 |
| clarkb | new conflicts could show up tomorrow | 03:13 |
| pabelanger | ack | 03:13 |
| promehteanfire | s/could/will | 03:13 |
| pabelanger | promehteanfire: ++ | 03:13 |
| clarkb | on debian it doesnt even fight over the bin entries | 03:14 |
| fungi | basically the split between distro packages in usr and pip using usr/local on debian mostly solved the problems, but because red hat stuck to having pip and distro packages use the same tree and files were constantly overwritten when mixing the package managers, pip maintainers decided they needed to assume all python packages on the system were fair game for pip to try to uninstall | 03:14 |
| *** bobh has joined #openstack-infra | 03:14 | |
| *** rlandy has quit IRC | 03:15 | |
| fungi | regardless of distro | 03:15 |
| *** slaweq has quit IRC | 03:16 | |
| * fungi really needs to get some sleep. more cleaning awaits tomorrow. night all! | 03:16 | |
| promehteanfire | heh, ya, pip freeze shows all the distro stuff | 03:16 |
| promehteanfire | fungi: nn | 03:16 |
| promehteanfire | don't remove this one portage==2.3.49 | 03:16 |
| *** jamesmcarthur has quit IRC | 03:19 | |
| clarkb | https://review.openstack.org/#/c/610755/ is the fix for etherpad-dev puppeting if anyone hastime to review | 03:23 |
| clarkb | Im off to enjoy my evening now | 03:23 |
| *** bobh has quit IRC | 03:31 | |
| *** zhangfei has quit IRC | 03:32 | |
| *** jamesmcarthur has joined #openstack-infra | 03:45 | |
| *** hongbin has quit IRC | 03:48 | |
| *** Bhujay has joined #openstack-infra | 03:51 | |
| *** Bhujay has quit IRC | 03:52 | |
| *** lpetrut has joined #openstack-infra | 03:52 | |
| *** Bhujay has joined #openstack-infra | 03:52 | |
| *** bhavikdbavishi has quit IRC | 03:54 | |
| *** zhangfei has joined #openstack-infra | 03:56 | |
| *** zhangfei has quit IRC | 03:58 | |
| *** zhangfei has joined #openstack-infra | 03:58 | |
| *** felipemonteiro has quit IRC | 04:04 | |
| *** Bhujay has quit IRC | 04:06 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder master: enable caching for gentoo builds https://review.openstack.org/604268 | 04:17 |
| *** agopi_ is now known as agopi | 04:18 | |
| *** ykarel has joined #openstack-infra | 04:22 | |
| *** zhangfei has quit IRC | 04:28 | |
| *** lpetrut has quit IRC | 04:34 | |
| *** bhavikdbavishi has joined #openstack-infra | 04:35 | |
| *** jamesmcarthur has quit IRC | 04:44 | |
| *** bhavikdbavishi has quit IRC | 04:52 | |
| *** bhavikdbavishi has joined #openstack-infra | 04:52 | |
| promehteanfire | whoa, zuul got updated | 05:09 |
| *** Douhet has quit IRC | 05:09 | |
| *** rkukura has joined #openstack-infra | 05:21 | |
| *** apetrich has quit IRC | 05:40 | |
| promehteanfire | finally PASS: Element gentoo, test-element: build-succeeds | 05:41 |
| promehteanfire | debian minimal fails, but who uses that anyway :P | 05:42 |
| *** Douhet has joined #openstack-infra | 05:44 | |
| *** janki has joined #openstack-infra | 05:47 | |
| *** dpawlik has joined #openstack-infra | 05:55 | |
| *** dpawlik has quit IRC | 06:00 | |
| *** apetrich has joined #openstack-infra | 06:00 | |
| *** chkumar|off is now known as chandankumar | 06:05 | |
| *** dpawlik has joined #openstack-infra | 06:11 | |
| *** slaweq has joined #openstack-infra | 06:11 | |
| *** dpawlik has quit IRC | 06:12 | |
| *** dpawlik has joined #openstack-infra | 06:12 | |
| AJaeger | config-core, could you reveiw https://review.openstack.org/610385, please? dependency is approved but not merged yet - I'll +A once ready. | 06:16 |
| *** slaweq has quit IRC | 06:16 | |
| openstackgerrit | Merged openstack-infra/project-config master: Reduce gate window floor to 10 https://review.openstack.org/610724 | 06:18 |
| *** apetrich has quit IRC | 06:23 | |
| *** noama has joined #openstack-infra | 06:24 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/project-config master: Disable ovh-gra1 https://review.openstack.org/610885 | 06:35 |
| dpawlik | AJaeger: do you still have problems with gra1? | 06:37 |
| dpawlik | ianw: you still have problems with the ports? | 06:42 |
| *** apetrich has joined #openstack-infra | 06:44 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul-jobs master: run-test-command: support list in test_command variable https://review.openstack.org/610888 | 06:46 |
| *** e0ne has joined #openstack-infra | 06:48 | |
| AJaeger | dpawlik: we had last night | 06:52 |
| AJaeger | dpawlik: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2018-10-15.log.html#t2018-10-15T05:08:39 | 06:52 |
| AJaeger | dpawlik: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2018-10-15.log.html#t2018-10-15T20:11:59 | 06:52 |
| dpawlik | AJaeger: ack, I will ask run team | 06:55 |
| *** slaweq has joined #openstack-infra | 06:56 | |
| AJaeger | dpawlik: thanks! | 06:57 |
| dpawlik | AJaeger: just ping me when are some problems with infra ;) | 06:57 |
| *** aojea has joined #openstack-infra | 06:57 | |
| AJaeger | dpawlik: will try - let's inform ianw, since he's an infra-root ^, I expect ianw to be a better contact | 06:58 |
| ianw | dpawlik: yeah, gra1 is not happy | 06:59 |
| ianw | dpawlik: i don't really know the best way to use storyboard for this, but there's tasks for gra1 and bhs1 | 07:00 |
| dpawlik | ianw: | 07:00 |
| ianw | https://storyboard.openstack.org/#!/story/2004090 | 07:00 |
| dpawlik | ianw: that regions have been updated to newton. They are little big so they are also problematic | 07:01 |
| *** rcernin has quit IRC | 07:01 | |
| dpawlik | ianw: but we still working on it. It should be better, but if its still not working, I will talk with other teams | 07:01 |
| ianw | dpawlik: yeah, it's been ongoing. gra1 -- i set us to just launch one node in there and it constantly failed, and then leaked ports | 07:02 |
| ianw | there are id's in that storyboard which might help if someone wants to look at logs | 07:02 |
| ianw | oh the bhs1 side, we have a lot of hosts in ERROR with "No valid host was found. There are not enough hosts available." | 07:04 |
| *** hashar has joined #openstack-infra | 07:05 | |
| *** jtomasek has joined #openstack-infra | 07:05 | |
| *** ginopc has joined #openstack-infra | 07:08 | |
| *** bhavikdbavishi1 has joined #openstack-infra | 07:14 | |
| *** bhavikdbavishi has quit IRC | 07:16 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 07:16 | |
| *** jpena|off is now known as jpena | 07:20 | |
| *** lpetrut has joined #openstack-infra | 07:20 | |
| *** quiquell|off is now known as quiquell | 07:24 | |
| *** lpetrut has quit IRC | 07:25 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: ubuntu: Add option to ignore missing mirror components https://review.openstack.org/610430 | 07:28 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Set EPEL mirror during openstack-ci-mirrors https://review.openstack.org/609169 | 07:28 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Provide for an EPEL mirror during build https://review.openstack.org/608834 | 07:28 |
| *** markvoelker has quit IRC | 07:29 | |
| *** markvoelker has joined #openstack-infra | 07:29 | |
| *** bhavikdbavishi has quit IRC | 07:32 | |
| *** markvoelker has quit IRC | 07:34 | |
| *** aojea has quit IRC | 07:42 | |
| *** gfidente has joined #openstack-infra | 07:45 | |
| *** tosky has joined #openstack-infra | 07:52 | |
| *** icey has quit IRC | 07:52 | |
| *** icey has joined #openstack-infra | 07:52 | |
| *** bhavikdbavishi has joined #openstack-infra | 07:53 | |
| *** kjackal has joined #openstack-infra | 07:53 | |
| *** roman_g has joined #openstack-infra | 07:55 | |
| openstackgerrit | Merged openstack-infra/project-config master: Disable ovh-gra1 https://review.openstack.org/610885 | 07:59 |
| *** jpich has joined #openstack-infra | 08:00 | |
| *** dtantsur|afk is now known as dtantsur | 08:03 | |
| *** gnuoy has quit IRC | 08:10 | |
| *** bhavikdbavishi has quit IRC | 08:14 | |
| *** bhavikdbavishi has joined #openstack-infra | 08:15 | |
| quiquell | pabelanger: Have fedora28 nodes at zuul run have python3 ? | 08:17 |
| *** bhavikdbavishi has quit IRC | 08:19 | |
| *** bhavikdbavishi has joined #openstack-infra | 08:27 | |
| *** markvoelker has joined #openstack-infra | 08:30 | |
| *** priteau has joined #openstack-infra | 08:31 | |
| *** zhangfei has joined #openstack-infra | 08:33 | |
| *** e0ne has quit IRC | 08:37 | |
| *** gnuoy has joined #openstack-infra | 08:41 | |
| *** e0ne has joined #openstack-infra | 08:44 | |
| *** ykarel is now known as ykarel|mtg | 08:51 | |
| *** studarus has quit IRC | 08:52 | |
| *** derekh has joined #openstack-infra | 08:52 | |
| openstackgerrit | Stephen Finucane proposed openstack-infra/git-review master: Remove auto-branch name https://review.openstack.org/610573 | 08:54 |
| stephenfin | mordred, fungi: What would one need to do to get +2 rights on openstack-infra/git-review? There's a rather large review queue built up there | 09:02 |
| *** ykarel|mtg is now known as ykarel | 09:03 | |
| *** markvoelker has quit IRC | 09:03 | |
| *** panda|off has quit IRC | 09:15 | |
| *** Dobroslaw has joined #openstack-infra | 09:16 | |
| *** panda has joined #openstack-infra | 09:18 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Cleanup node requests that are declined by all current providers https://review.openstack.org/610915 | 09:25 |
| *** persia has quit IRC | 09:27 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Use quota handling code for min-ready declines https://review.openstack.org/610495 | 09:28 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Remove unneeded todo comment https://review.openstack.org/610496 | 09:28 |
| *** persia has joined #openstack-infra | 09:28 | |
| ssbarnea | a bindep question https://ask.openstack.org/en/question/116630/how-to-use-aliases-with-bindep-requirements/ | 09:32 |
| *** sshnaidm|afk has joined #openstack-infra | 09:32 | |
| *** shardy has joined #openstack-infra | 09:37 | |
| stephenfin | ssbarnea: Pretty sure aliases is a corner case that isn't covered by bindep yet | 09:44 |
| ssbarnea | stephenfin: soon we will need something to cover for this, at least for cli (yum/dnf) i was able to get around with wildcards. | 09:45 |
| stephenfin | ssbarnea: I imagine it would be an easy feature to add, if you were so inclined? | 09:45 |
| *** imacdonn has quit IRC | 09:54 | |
| *** imacdonn has joined #openstack-infra | 09:55 | |
| ssbarnea | yeah, i am not afraid of adding a CR if i find no way to fix a build. at the moment i have impression that I am ok, but this may change soon. | 09:57 |
| *** longkb has quit IRC | 09:59 | |
| *** ykarel has quit IRC | 09:59 | |
| *** markvoelker has joined #openstack-infra | 10:00 | |
| *** sshnaidm|afk is now known as sshnaidm | 10:03 | |
| *** chrisyang_0660 has quit IRC | 10:04 | |
| *** ykarel has joined #openstack-infra | 10:08 | |
| *** ykarel is now known as ykarel|lunch | 10:09 | |
| *** jamielennox has quit IRC | 10:11 | |
| *** jamielennox has joined #openstack-infra | 10:13 | |
| *** felipemonteiro has joined #openstack-infra | 10:27 | |
| *** markvoelker has quit IRC | 10:34 | |
| *** ykarel_ has joined #openstack-infra | 10:48 | |
| *** snapiri has quit IRC | 10:50 | |
| *** ykarel|lunch has quit IRC | 10:50 | |
| *** jamesmcarthur has joined #openstack-infra | 10:56 | |
| *** ykarel_ is now known as ykarel | 10:56 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Cleanup node requests that are declined by all current providers https://review.openstack.org/610915 | 10:58 |
| *** sshnaidm has quit IRC | 11:00 | |
| *** sshnaidm has joined #openstack-infra | 11:01 | |
| *** quiquell is now known as quiquell|lunch | 11:01 | |
| *** shardy has quit IRC | 11:02 | |
| *** electrofelix has joined #openstack-infra | 11:04 | |
| *** dave-mccowan has joined #openstack-infra | 11:09 | |
| *** ansmith has quit IRC | 11:16 | |
| *** e0ne has quit IRC | 11:18 | |
| *** felipemonteiro has quit IRC | 11:20 | |
| *** roman_g has quit IRC | 11:21 | |
| *** roman_g has joined #openstack-infra | 11:22 | |
| *** bhavikdbavishi has quit IRC | 11:23 | |
| *** jpena is now known as jpena|lunch | 11:29 | |
| *** markvoelker has joined #openstack-infra | 11:31 | |
| *** e0ne has joined #openstack-infra | 11:32 | |
| *** quiquell|lunch is now known as quiquell | 11:37 | |
| *** sshnaidm has quit IRC | 11:38 | |
| *** sshnaidm has joined #openstack-infra | 11:38 | |
| *** jistr is now known as jistr|afk | 11:47 | |
| *** nhicher has joined #openstack-infra | 11:58 | |
| hwoarang | dirk: have you noticed all the RETRY_LIMIT failures on opensuse-150 images? Error is "No viable v4 or v6 route found to git.openstack.org. The build node is assumed to be invalid" | 12:00 |
| *** trown|outtypewww is now known as trown | 12:00 | |
| *** rh-jelabarre has joined #openstack-infra | 12:03 | |
| *** markvoelker has quit IRC | 12:04 | |
| *** yamamoto has quit IRC | 12:04 | |
| *** smarcet has joined #openstack-infra | 12:04 | |
| *** yamamoto has joined #openstack-infra | 12:05 | |
| *** yamamoto has quit IRC | 12:05 | |
| *** yamamoto has joined #openstack-infra | 12:05 | |
| frickler | hwoarang: do you have a pointer to a log with that handy? | 12:06 |
| frickler | hwoarang: also, there is still a held node for opensuse-150 from 11d ago, do you still need that? | 12:07 |
| hwoarang | frickler: http://logs.openstack.org/56/606056/2/check/openstack-ansible-functional-opensuse-150/48f015a/job-output.txt.gz | 12:07 |
| hwoarang | frickler: sorry i was on paternity leave and did not have the chance to look more into that. Please release the node and I will let you know if I need one again. Thank you | 12:07 |
| *** quiquell is now known as quiquell|off | 12:11 | |
| frickler | hwoarang: so your failure might be related to the ipv6 address not being set up on the node, http://logs.openstack.org/56/606056/2/check/openstack-ansible-functional-opensuse-150/48f015a/zuul-info/inventory.yaml has a public v6 address while it is missing here http://logs.openstack.org/56/606056/2/check/openstack-ansible-functional-opensuse-150/48f015a/zuul-info/host-info.opensuse-150.yaml | 12:11 |
| frickler | hwoarang: I just did a test on the held node and see unbound failing there, too. do you remember whether it worked when you looked at it earlier | 12:15 |
| frickler | hwoarang: I'm guessing that this might be related to the ksk rollover, unbound.log says: Oct 16 12:12:23 unbound[2689:0] info: validation failure openstack.org. A IN | 12:16 |
| *** bhavikdbavishi has joined #openstack-infra | 12:17 | |
| frickler | hwoarang: indeed, /var/lib/unbound/root.key shows the new key as "ADDPEND" instead of "VALID". see https://www.unbound.net/root-11sep-11oct.html | 12:18 |
| *** electrofelix has quit IRC | 12:18 | |
| *** smarcet has quit IRC | 12:19 | |
| *** boden has joined #openstack-infra | 12:20 | |
| *** e0ne has quit IRC | 12:24 | |
| *** electrofelix has joined #openstack-infra | 12:24 | |
| hwoarang | hmm | 12:25 |
| hwoarang | not sure when that regressed | 12:26 |
| hwoarang | or why | 12:26 |
| hwoarang | frickler: the node i had reserved was for a different issue but it was not failing at this point | 12:27 |
| *** e0ne has joined #openstack-infra | 12:27 | |
| hwoarang | jobs from at least a week ago seem to work fine. Need to investigate i guess but I was hoping dirk may have the magic answer :) | 12:28 |
| frickler | the rollover was last thursday, so it would have regressed at that date probably. I tried regenerating the root.key file as mentioned in the unbound advisory above and that seems to resolve the issue. not sure why it would occur in the first place, though, when you have 1.6.8 installed | 12:28 |
| frickler | hwoarang: could you recheck your job above? I'd hold another node for that, then, in order to crosscheck | 12:29 |
| dirk | hwoarang: 42 is the magic answer | 12:29 |
| *** Emine has joined #openstack-infra | 12:30 | |
| hwoarang | tell me more | 12:31 |
| dirk | frickler: do you have any pointer ? I'm happy to help debugging just don't know what / where to look | 12:31 |
| *** felipemonteiro has joined #openstack-infra | 12:32 | |
| frickler | dirk: you could try to set up an instance locally with unbound as forwarding resolver and check whether dnssec works in that case. then we'd have something to crosscheck against our soon-to-be-held instance | 12:34 |
| *** rlandy has joined #openstack-infra | 12:35 | |
| frickler | dirk: https://www.icann.org/dns-resolvers-updating-latest-trust-anchor has a bit more information about what happened | 12:36 |
| *** sshnaidm has quit IRC | 12:36 | |
| dirk | frickler: ah, yeah, I heard about the dns key switch. didn't quite connect this | 12:37 |
| dirk | frickler: so it means the unbound package in leap 15.0 is missing the new root key? | 12:37 |
| frickler | dirk: not sure yet how this comes about. the new key is present, but in state "ADDPEND", which means that it was added only recently added and will become valid in 30 days. | 12:39 |
| *** priteau has quit IRC | 12:39 | |
| *** sshnaidm has joined #openstack-infra | 12:39 | |
| frickler | dirk: that issue should only happen with unbound < 1.6.4, though, while you have 1.6.8 installed. hoping to see more on a new held node or maybe if you try locally | 12:40 |
| *** priteau has joined #openstack-infra | 12:40 | |
| dirk | frickler: I see a root.key in the package, that looks suspicous. can we maybe fix that by releasing an update that includes only the new key? | 12:41 |
| frickler | dirk: that would be an option probably, yes. | 12:42 |
| dirk | frickler: I'll try to find somebody for this | 12:43 |
| tosky | did anyone else hit errors when getting SQLAlchemy_Utils-0.33.6-py2.py3-none-any.whl from the internal mirrors? | 12:45 |
| tosky | ("Bad magic number for central directory", which I guess means a corrupted file) | 12:45 |
| *** bhavikdbavishi has quit IRC | 12:49 | |
| openstackgerrit | Corey Bryant proposed openstack-infra/zuul-jobs master: Limit ensure-python to Debian/Ubuntu use https://review.openstack.org/610948 | 12:53 |
| dirk | frickler: hwoarang : I submitted a maintenance update: https://build.opensuse.org/request/show/642274 | 12:53 |
| coreycb | pabelanger: fungi: clarkb: I saw some comments limiting ensure-python to debian/ubuntu so I added some logic ^^. thanks for the discussion last night, sorry i missed it. | 12:54 |
| fungi | coreycb: great, thanks! | 12:57 |
| hwoarang | thank you dirk | 12:58 |
| fungi | stephenfin: clarkb is the ptl, so the one to ask. ssbarnea has helped a bunch recently with getting some of the more obvious git-review fixes triaged and reviewed for the 1.27.0 release just under a month ago but i agree there are still a number that need to be considered and either merged or rejected as out of scope (git-review gets a lot of proposed "features" which aren't really a good fit) | 13:00 |
| *** kgiusti has joined #openstack-infra | 13:01 | |
| *** bobh has joined #openstack-infra | 13:04 | |
| *** sshnaidm is now known as sshnaidm|afk | 13:05 | |
| mordred | stephenfin: I left a comment on that patch | 13:06 |
| stephenfin | ta | 13:06 |
| *** ansmith has joined #openstack-infra | 13:08 | |
| *** sshnaidm|afk has quit IRC | 13:09 | |
| *** Emine has quit IRC | 13:09 | |
| *** efried has joined #openstack-infra | 13:10 | |
| coreycb | fungi: clarkb: infra: if you have a sec, can you comment on this please? https://review.openstack.org/#/c/610708/ personally i really would prefer to drop 3.5 in favor of 3.7 unit tests as we originally discussed. but this governance change seems like the current path of least resistance (keep 3.5 and add 3.6 *or* 3.7). | 13:10 |
| *** yamamoto has quit IRC | 13:10 | |
| *** yamamoto has joined #openstack-infra | 13:11 | |
| *** e0ne has quit IRC | 13:12 | |
| *** dpawlik has quit IRC | 13:13 | |
| *** agopi has quit IRC | 13:15 | |
| *** yamamoto has quit IRC | 13:15 | |
| *** e0ne has joined #openstack-infra | 13:16 | |
| *** dpawlik has joined #openstack-infra | 13:18 | |
| frickler | infra-root: I looked at the logs regarding my doubly held node yesterday. seem zuul saw an error for the first hold request although nodepool still did hold the node successfully. not sure how to continue with this, though, probably will ignore it unless it happens regularly http://paste.openstack.org/show/732244/ | 13:18 |
| frickler | tosky: I just saw a gate failure due to this. maybe some other infra-root can dig into that? http://logs.openstack.org/34/610734/1/gate/requirements-integration/63dc003/job-output.txt.gz#_2018-10-16_13_13_47_687378 | 13:20 |
| *** jistr|afk is now known as jistr | 13:20 | |
| *** mriedem has joined #openstack-infra | 13:23 | |
| fungi | coreycb: i think you're caught in the middle of some disagreements on what we should be testing. my position is that we need to complete our transition from xenial to bionic (hence from 3.5 jobs to 3.6 jobs) which people seem to have forgotten we were in the middle of | 13:24 |
| frickler | fungi: mnaser: there is an old held node for you, too, also for fuentess who doesn't seem to be here, does anybody know the status of that? | 13:24 |
| fungi | coreycb: the intent was never to test 3.5 and 3.6 for a long period of time, only long enough for projects which had been using py35 jobs to make sure py36 jobs were working and set them voting so we didn't have a gap nor a major disruption | 13:25 |
| coreycb | fungi: yes it seems that way. there are a lot of competing thoughts. | 13:25 |
| fungi | adding py37 testing is nice, but perhaps confusing people by happening in the same patches as the py35 removal, as they're separate efforts | 13:26 |
| coreycb | fungi: i agree. if only we could just add py37 and not drop py35, but i don't think infra has resources to support that. | 13:27 |
| fungi | i think we've also given the projects who had py35 working long enough to get py36 working, so should be looking to drop py35 asap | 13:27 |
| Shrews | frickler: that appears to be a race between when get the child nodes in a tree and when we iterate through those children, getting each individually. We've seen that before in many other scenarios | 13:27 |
| *** Serhii_Rusin has joined #openstack-infra | 13:27 | |
| *** felipemonteiro has quit IRC | 13:27 | |
| coreycb | fungi: do you think temporarily running 3.5, 3.6, and 3.7 unit tests is a no-go? | 13:29 |
| frickler | Shrews: sounds plausible, yes, but not like it could be easily fixed | 13:29 |
| *** Serhii_Rusin has left #openstack-infra | 13:30 | |
| AJaeger | coreycb: we're missing some clouds and are running at production limit - any extra job means slowdown for all of us. | 13:30 |
| *** dpawlik has quit IRC | 13:30 | |
| fungi | coreycb: i don't think it's a no-go, but i think the priorities are the bionic transition (we say in our pti we test on the latest lts), then getting python3 jobs working for teams which haven't switched to python3 yet (it's a cycle goal), then adding py37 jobs (a nice-to-have) | 13:31 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul master: Handle missing node during hold check https://review.openstack.org/610969 | 13:32 |
| Shrews | frickler: i think that ^ will handle it a bit better | 13:32 |
| coreycb | fungi: that's reasonable that the original python3-first goals are priority over py37 jobs | 13:33 |
| *** Serhii_Rusin has joined #openstack-infra | 13:33 | |
| Shrews | frickler: *very* easily fixed :) | 13:34 |
| coreycb | fungi: but if it's not a no-go to support all 3 py3 unit test jobs, then that makes me think i can start adding py37 jobs. and is contrary to AJaeger's comment. | 13:34 |
| *** agopi has joined #openstack-infra | 13:36 | |
| fungi | coreycb: i'm not sure what percentage increase of overall ci resource consumption we're talking about, but i'm guessing not huge since unit tests tend to be fairly quick (compared to functional/integration type jobs) and would only likely make sense for projects which already have voting py36 jobs | 13:36 |
| *** ociuhandu has joined #openstack-infra | 13:37 | |
| *** jpena|lunch is now known as jpena | 13:37 | |
| *** ociuhandu has quit IRC | 13:38 | |
| *** agopi_ has joined #openstack-infra | 13:38 | |
| coreycb | fungi: yes in my experience most unit tests are fairly quick particularly for dependencies. some larger core projects like neutron/nova are not as quick. | 13:38 |
| *** agopi has quit IRC | 13:40 | |
| *** dpawlik has joined #openstack-infra | 13:40 | |
| coreycb | fungi: so perhaps a strategy could be to add py37 jobs (on top of py35 and py36) if and only if py36 unit test jobs are already voting for the project | 13:40 |
| coreycb | of course that would need +1 from infra | 13:41 |
| frickler | Shrews: looks easy indeed, thanks | 13:41 |
| slaweq | frickler: hi | 13:41 |
| slaweq | frickler: I have a question, for jobs which runs on Xenial (e.g. neutron-functional) we are using nodeset: devstack-single-node | 13:42 |
| slaweq | frickler: if I would like to run this job on Bionic, should I use openstack-single-node-bionic instead? | 13:43 |
| frickler | slaweq: for short term testing, yes. I was just thinking how to move devstack towards bionic in general, though, and one solution for that might be to make devstack-single-node be bionic instead, possibly having devstack-single-node-xenial for jobs that do need xenial explicitly | 13:45 |
| *** sthussey has joined #openstack-infra | 13:45 | |
| *** eharney has joined #openstack-infra | 13:45 | |
| slaweq | frickler: ok, thx | 13:46 |
| *** ykarel is now known as ykarel|afk | 13:50 | |
| slaweq | frickler: and what about multinode jobs? e.g job which now uses ubuntu-xenial-2-node - I don't see ubuntu-nionic-2-node nodeset | 13:51 |
| *** e0ne has quit IRC | 13:53 | |
| *** janki has quit IRC | 13:58 | |
| openstackgerrit | Stephen Finucane proposed openstack-infra/git-review master: DNM: Test patch https://review.openstack.org/610976 | 13:58 |
| openstackgerrit | Stephen Finucane proposed openstack-infra/git-review master: DNM: Test patch https://review.openstack.org/610976 | 13:59 |
| frickler | slaweq: see #openstack-qa | 13:59 |
| mordred | coreycb, fungi: we could also investigate a combined job - so one that ran "tox -epy35,py36,py37" | 14:00 |
| *** yamamoto has joined #openstack-infra | 14:00 | |
| mordred | which would consume fewer overal nodes - and because of shared setup and wheel caching would also take less overall cloud CPU time than three different distinct jobs | 14:01 |
| tosky | a unified unit-test job? That would not be bad | 14:01 |
| slaweq | frickler: thx | 14:01 |
| *** e0ne has joined #openstack-infra | 14:01 | |
| tosky | mordred: do you think it would make sense to fold the pep8 job there too? | 14:01 |
| mordred | I'm not sure the subunit2html post-processor would handle that today - so it would be a little bit of work | 14:01 |
| coreycb | mordred: true. for ubuntu it would need to still be 2 jobs with 3.5 on xenial and 3.6/3.7 on bionic but still would save on spin up/down time. | 14:02 |
| mordred | tosky: pep8 reports its errors a bit differently, so it might be a weirder user experience ... but I believe that would be possible if someone chose if we can sort out the error reporting for more than one subunit stream | 14:02 |
| mordred | coreycb: yah. | 14:03 |
| *** agopi_ is now known as agopi | 14:04 | |
| openstackgerrit | David Moreau Simard proposed openstack-infra/zuul master: web: Increase height and padding of zuul-job-result https://review.openstack.org/610980 | 14:04 |
| mordred | tosky: WAY back in the day there was a suggestion to write a dynamic unittest class which ran the pep8 stuff as if it were unittests | 14:04 |
| mordred | tosky: nobody ever got around to writing the code to do so though | 14:04 |
| mordred | tosky: otoh - we've got a patch coming to zuul-jobs to process the pep8 output and leave in-line comments about the failures which we'd lose if we folded pep8 in to the subunit stream (most of the subunit failures don't make a ton of sense as in-line comments) | 14:05 |
| tosky | oh | 14:06 |
| tosky | I see | 14:06 |
| mordred | tosky: I should never talk to people when I'm still on my first coffee | 14:07 |
| tosky | O.O | 14:08 |
| * tosky didn't notice anything weird | 14:08 | |
| *** dpawlik has quit IRC | 14:11 | |
| openstackgerrit | Stephen Finucane proposed openstack-infra/git-review master: Don't set topic when submitting no-topic patches https://review.openstack.org/610988 | 14:13 |
| *** rkukura has quit IRC | 14:14 | |
| *** munimeha1 has joined #openstack-infra | 14:19 | |
| *** eernst has joined #openstack-infra | 14:20 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Initialize label statistics to zero https://review.openstack.org/610993 | 14:23 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Initialize label statistics to zero https://review.openstack.org/610993 | 14:28 |
| *** ykarel|afk has quit IRC | 14:29 | |
| *** ykarel|afk has joined #openstack-infra | 14:30 | |
| *** eharney has quit IRC | 14:33 | |
| *** eernst has quit IRC | 14:34 | |
| *** Serhii_Rusin has quit IRC | 14:39 | |
| clarkb | stephenfin: fungi: after conversations with ssbarnea I'm happy to add a few new cores (maybe stephenfin and ssbarnea to start) if we can agree on some rough starter rules like bug fixes and testing improvements first. Then if we get reliable testing again that we are happy with consider feature changes at that point. | 14:41 |
| clarkb | stephenfin: ssbarnea fungi even then I'm not sure we want to add a ton of features to git review | 14:41 |
| clarkb | its a fairly small predictable and understandable tool today, if needing an update or three to keep up with new gerrit changes | 14:42 |
| stephenfin | clarkb: Yup, Unix philosophy ftw | 14:43 |
| clarkb | ssbarnea: "With wilcards both, yum and dnf behave the same: returning an error only if at the end none of the packages is installed. If one of the packages is installed, they will return success." I wish dnf hadn't preserved that behavior from yum :( | 14:43 |
| *** armax has joined #openstack-infra | 14:43 | |
| clarkb | makes the lives of config management tools difficult because they must install one package at a time to check if all packages install correctly | 14:43 |
| clarkb | (among other confusing results) | 14:44 |
| *** dpawlik has joined #openstack-infra | 14:46 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/puppet-etherpad_lite master: Bump eplite version in rspec tests https://review.openstack.org/611000 | 14:47 |
| clarkb | anyone willing to review ^ and its parent? THe parent change fixes a bug getting etherpad-dev redeployed on xenial | 14:47 |
| fungi | or they have to check back at the end to make sure the packages which were meant to be installed actually are | 14:47 |
| *** eharney has joined #openstack-infra | 14:48 | |
| *** ykarel|afk has quit IRC | 14:48 | |
| *** ykarel|afk has joined #openstack-infra | 14:48 | |
| *** dpawlik has quit IRC | 14:50 | |
| *** rkukura has joined #openstack-infra | 14:51 | |
| *** fuentess_ has joined #openstack-infra | 14:51 | |
| *** fuentess_ is now known as fuentess | 14:52 | |
| openstackgerrit | Daniel Lublin proposed openstack-infra/git-review master: Allow choosing which field to use as author when naming branch https://review.openstack.org/444574 | 14:53 |
| *** felipemonteiro has joined #openstack-infra | 14:55 | |
| *** Emine has joined #openstack-infra | 14:56 | |
| ssbarnea | clarkb: (dnf) hehe, well one's bug is someone else's feature | 14:57 |
| tosky | infra core, as mentioned earlier, I'm hitting an error while getting one file from the mirrors | 14:57 |
| clarkb | ssbarnea: indeed | 14:57 |
| tosky | SQLAlchemy_Utils-0.33.6-py2.py3-none-any.whl | 14:57 |
| tosky | see http://logs.openstack.org/27/610527/1/gate/sahara-tests-scenario/73bcec5/job-output.txt.gz#_2018-10-16_13_38_57_196915 | 14:57 |
| *** ccamacho has quit IRC | 14:57 | |
| ssbarnea | clarkb: but i am happy to hear about better ways of coping with the same issue. | 14:58 |
| tosky | and also http://logs.openstack.org/27/610527/1/gate/sahara-tests-tempest/8adcdff/job-output.txt.gz#_2018-10-16_13_32_16_326018 (different provider) | 14:58 |
| clarkb | tosky: pulling the file locally and running `unzip` on it reports the same error. The files are served from the same AFS filesystem regardless of provider so should be identical across the board | 15:00 |
| clarkb | likely is a corrupt file in this case | 15:00 |
| *** ccamacho has joined #openstack-infra | 15:00 | |
| clarkb | let me see about deleting it | 15:00 |
| tosky | thanks | 15:01 |
| tosky | ! | 15:01 |
| *** ramishra has quit IRC | 15:01 | |
| ssbarnea | clarkb: regarding git review, i got your point which I would translate as "preserving compatibility is no1 prio". git-review is one of the most used tools around here, so I can imagine what would happen if a new release starts to break our workflows. once approach would be to publish all merges to master as prereleases on pypa. this would make it much easier for (more) people to use "bleading edge" version before we would do a new release. | 15:01 |
| fungi | definitely easy. we can just make sure to push frequent prerelease tags and our jobs should do the rest | 15:03 |
| fungi | if there's agreement we want to publish pre-releases of git-review i can push a 1.27.1.0a1 tag now | 15:05 |
| *** rkukura has quit IRC | 15:05 | |
| *** kaiokmo has quit IRC | 15:08 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid-resources master: Added new my affiliations endpoints https://review.openstack.org/611006 | 15:09 |
| *** kaiokmo has joined #openstack-infra | 15:10 | |
| clarkb | I'm getting permission denied rm'ing the sqlalchemy file using my clarkb/admin kerberos account | 15:10 |
| clarkb | what am I missing here | 15:10 |
| clarkb | aha aklog | 15:10 |
| clarkb | yay for docs I wrote :) | 15:10 |
| fungi | heh | 15:10 |
| *** kjackal has quit IRC | 15:11 | |
| clarkb | https://review.openstack.org/#/c/601855/1/doc/source/afs.rst not merged yet. Debating if I should change the account there to the admin one and maybe I should | 15:11 |
| *** kjackal has joined #openstack-infra | 15:12 | |
| clarkb | I am going to /usr/sbin/vos release -cell openstack.org mirror.wheel.xenialx64 | 15:14 |
| clarkb | fungi: ^ that look right to you? and any idea if I need to lock that volume somehow so that our publish jobs don't conflict with me? | 15:15 |
| openstackgerrit | Merged openstack-infra/openstackid-resources master: Added new my affiliations endpoints https://review.openstack.org/611006 | 15:15 |
| * dtantsur sees that the BadZipFile problem is already known | 15:15 | |
| clarkb | that job seems to run in periodic which happens at 0600 so I should be safe doing a release now | 15:16 |
| clarkb | going to try it | 15:16 |
| ssbarnea | fungi: small nit https://review.openstack.org/#/c/456929/ | 15:16 |
| clarkb | "Released volume mirror.wheel.xenialx64 successfully" tosky dtantsur for timestamps after 1517UTC you should see this work now I expect | 15:17 |
| clarkb | we'll have to see if sqlalchemy utils wheel building is generally corrupted | 15:17 |
| tosky | I remember that something like this happened few days (weeks?) ago | 15:18 |
| clarkb | I'm going to double check quotas and disk usage too | 15:19 |
| clarkb | though if the write failed due to lack fo disk then we shouldn't release the volume so I don't expect htat is a problem | 15:19 |
| clarkb | when we run out of disk the behavior is to stop udpating at all | 15:19 |
| *** zhangfei has quit IRC | 15:19 | |
| *** chandankumar is now known as chkumar|off | 15:19 | |
| tosky | clarkb: one month ago: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2018-09-17.log.html#t2018-09-17T09:14:57 | 15:19 |
| clarkb | ya we use 5% of the current quota on that fs | 15:20 |
| clarkb | and grafana says we still have disk space on vicepa | 15:20 |
| tosky | interesting enough, it was the exact same wheel (two minor versions ago) | 15:21 |
| clarkb | tosky: hrm we did recently merge the change to not consume our own wheels to build wheels | 15:22 |
| tosky | this one? https://review.openstack.org/#/c/603064/ | 15:22 |
| clarkb | but yes, suspicious that it is the same package again. My hunch is something about wheel builds for that package are flaky | 15:22 |
| clarkb | tosky: yes that one | 15:23 |
| tosky | ianw wrote that immediately after the issue with that wheel | 15:23 |
| fungi | clarkb: lgtm | 15:25 |
| openstackgerrit | Felipe Monteiro proposed openstack-infra/project-config master: Remove airship-armada jobs, as they are all in project https://review.openstack.org/611013 | 15:25 |
| clarkb | tosky: http://logs.openstack.org/periodic/git.openstack.org/openstack/requirements/master/publish-wheel-mirror-ubuntu-xenial/25125c0/python2/master-job.log shows where we built this wheel I think | 15:27 |
| clarkb | tosky: and the retval column says 0 | 15:27 |
| clarkb | imo this is broken packaging | 15:27 |
| openstackgerrit | Graham Hayes proposed openstack-infra/project-config master: Install Pygments before checking README syntax for publish jobs https://review.openstack.org/611014 | 15:27 |
| clarkb | sorry exitval | 15:27 |
| clarkb | tosky: https://github.com/kvesteri/sqlalchemy-utils/blob/master/setup.py#L74 is suspicious to me | 15:34 |
| clarkb | however they've set it to false which is what I'd expect it to be if not safe so propbably unrelated | 15:36 |
| *** sshnaidm has joined #openstack-infra | 15:37 | |
| dtroyer | Does anyone know if we (community) already do any checking of website URLs? I'm looking for something to do really basic "yup the right page is there" or "the redirect is still working" kinds of things periodically against published docs sites. | 15:39 |
| clarkb | dtroyer: external monitoring? we don't currently run that against any of the sites that I know of. The one check we do with reminders is ssl cert verification and expiry notifications | 15:41 |
| *** hamzy has quit IRC | 15:42 | |
| tosky | dtroyer: there is a report of 404 errors for doc pages, AJaeger knows more about it | 15:42 |
| fungi | there is also the whereto checker for docs changes to make sure they don't break relevant redirects | 15:43 |
| dtroyer | tosky: ok, thanks. clarkb: external from the standpoint that it's not on the web server is enough. I've got 30+ job/repo combinations and don't want to check it by hand forever… | 15:45 |
| dtroyer | fungi: I think that is next on our list of things to understand, we're not doing the versioned repos yet and will need to get deeper there before March. | 15:45 |
| dtroyer | er, versioned docs trees | 15:46 |
| dtroyer | thanks guys | 15:46 |
| openstackgerrit | weizj proposed openstack/os-testr master: Update the hacking to latst https://review.openstack.org/611048 | 15:47 |
| *** felipemonteiro has quit IRC | 15:49 | |
| *** e0ne has quit IRC | 15:50 | |
| *** promehteanfire is now known as prometheanfire | 16:01 | |
| *** ginopc has quit IRC | 16:02 | |
| openstackgerrit | Merged openstack-dev/pbr master: Add an option to skip generating RELEASENOTES.rst https://review.openstack.org/587204 | 16:08 |
| *** noama has quit IRC | 16:10 | |
| openstackgerrit | sebastian marcet proposed openstack-infra/openstackid-resources master: Fixed my speaker profile https://review.openstack.org/611069 | 16:11 |
| prometheanfire | dnssec stuff for opensuse-150, is that handled? | 16:13 |
| *** sshnaidm has quit IRC | 16:13 | |
| prometheanfire | dirk: dnssec stuff for opensuse-150, is that handled? https://data.iana.org/root-anchors/ | 16:14 |
| *** hashar is now known as hasharAway | 16:14 | |
| *** quiquell|off has quit IRC | 16:15 | |
| openstackgerrit | weizj proposed openstack/os-testr master: Update the hacking to latst https://review.openstack.org/611048 | 16:16 |
| openstackgerrit | Merged openstack-infra/zuul master: Provide some accounting of node usage in logs https://review.openstack.org/605856 | 16:17 |
| *** ykarel|afk is now known as ykarel | 16:18 | |
| *** apetrich has quit IRC | 16:18 | |
| *** bhavikdbavishi has joined #openstack-infra | 16:20 | |
| *** d0ugal has quit IRC | 16:21 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config master: Check wheels built are valid zip files https://review.openstack.org/611075 | 16:22 |
| clarkb | tosky: dtantsur ^ fyi I think that may catch it in the future? | 16:22 |
| *** d0ugal has joined #openstack-infra | 16:23 | |
| tosky | I don't know really much about wheels and the process, but I concur that this change should prevent bad zip files | 16:23 |
| clarkb | infra-root review on https://review.openstack.org/#/c/610755/ much appreciated so I can continue to make progress with the etherpad upgrades | 16:24 |
| * clarkb finds breakfast | 16:25 | |
| *** gfidente is now known as gfidente|afk | 16:25 | |
| *** jpich has quit IRC | 16:29 | |
| *** apetrich has joined #openstack-infra | 16:30 | |
| *** rlandy is now known as rlandy|biab | 16:32 | |
| *** diablo_rojo has joined #openstack-infra | 16:33 | |
| *** rpioso|afk is now known as rpioso | 16:36 | |
| *** ykarel is now known as ykarel|away | 16:37 | |
| *** eernst has joined #openstack-infra | 16:42 | |
| clarkb | interesting data, the check queue backlog has already at 2 hours and this is after the gate window floor change (which has limited the impact of some integrated gate resets this morning from what I have seen) | 16:42 |
| *** jpena is now known as jpena|off | 16:45 | |
| *** ykarel|away has quit IRC | 16:46 | |
| corvus | we changed the floor? | 16:47 |
| clarkb | corvus: yes, pabelanger had suggested it and watching the gate reset many times yesterday I went ahead and pushed the change to half it to 10 yesterday. Looks like it merged overnight | 16:47 |
| *** panda is now known as panda|off | 16:48 | |
| clarkb | considering the backlog isn't appreciably better I'm not sure it has had the impact pabelanger thought it would have | 16:48 |
| openstackgerrit | Merged openstack-infra/puppet-etherpad_lite master: Ensure the custom js dir is created https://review.openstack.org/610755 | 16:50 |
| pabelanger | clarkb: corvus: looks like AJaeger approved it. But just now also noticing it merged | 16:54 |
| *** chkumar|off is now known as chandankumar | 16:57 | |
| *** eernst has quit IRC | 16:58 | |
| *** jamesmcarthur has quit IRC | 17:00 | |
| *** jamesmcarthur has joined #openstack-infra | 17:00 | |
| *** jesusaur has quit IRC | 17:02 | |
| clarkb | anyway I think we knew it would be a short term experiment. So far data indicates it doesn't change much and we can probably revert | 17:02 |
| *** dtantsur is now known as dtantsur|afk | 17:03 | |
| *** kopecmartin|ruck is now known as kopecmartin|off | 17:05 | |
| *** hamzy has joined #openstack-infra | 17:05 | |
| *** e0ne has joined #openstack-infra | 17:08 | |
| *** munimeha1 has quit IRC | 17:10 | |
| *** kaiokmo has quit IRC | 17:17 | |
| *** kaiokmo has joined #openstack-infra | 17:18 | |
| *** panda|off has quit IRC | 17:24 | |
| *** panda has joined #openstack-infra | 17:27 | |
| dirk | prometheanfire: in general yes, except that we forgot unbound as noone uses that in opensuse | 17:31 |
| dirk | prometheanfire: that update is in the works now | 17:31 |
| clarkb | yay we helped find a bug? | 17:31 |
| dirk | Unfortunately its one of the 'this would have been really helpful, last week' | 17:32 |
| dirk | clarkb: yep | 17:32 |
| AJaeger | pabelanger, clarkb I considered +8 a good time to try the experiment ;) | 17:32 |
| dirk | clarkb: and for your pleasure, it was fixed in tumbleweed weeks ago ;-) | 17:33 |
| AJaeger | config-core, could you review https://review.openstack.org/610385 and https://review.openstack.org/610386 , please? The job moved to tempest repo in periodic pipeline, so we can remove it. | 17:33 |
| openstackgerrit | Merged openstack-infra/openstackid-resources master: Fixed my speaker profile https://review.openstack.org/611069 | 17:33 |
| AJaeger | clarkb: but I saw quite a few reset in integrated-gate, so that lower number might have helped. | 17:36 |
| clarkb | AJaeger: ya there have been some resets there | 17:36 |
| AJaeger | clarkb: since we prioritize gate pipeline over chck now, I think limiting it might be fine - the gate is too flaky to merge 10 or even 20 changes at once... | 17:36 |
| AJaeger | and we give priority with nodes to it - so, I would not revert directly | 17:37 |
| *** trown is now known as trown|lunch | 17:37 | |
| clarkb | AJaeger: ya I think we can continue to experiment with it. Its just that current data doesn't show it helping much so maybe we should look at other things | 17:38 |
| fungi | also if the gate queues stabilize then the floor becomes irrelevant | 17:39 |
| AJaeger | fungi: yeah... | 17:40 |
| *** electrofelix has quit IRC | 17:41 | |
| AJaeger | tosky, dtroyer, http://files.openstack.org/docs-404s/ is the list of 404s that people hit when going to docs.o.o | 17:41 |
| *** rlandy|biab is now known as rlandy | 17:46 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Update governance-uc docs publishing https://review.openstack.org/611114 | 17:51 |
| *** gfidente|afk has quit IRC | 17:52 | |
| AJaeger | fungi, could you review https://review.openstack.org/611014 to unblock designate, please? | 17:55 |
| openstackgerrit | Merged openstack-infra/project-config master: Remove periodic-tempest-dsvm-oslo-latest-full-master job https://review.openstack.org/610385 | 17:57 |
| clarkb | new etherpad-dev seems to be working. I am updating dns now as a result | 17:59 |
| *** e0ne has quit IRC | 17:59 | |
| openstackgerrit | Merged openstack-infra/zuul-website master: Add three more case studies https://review.openstack.org/610757 | 18:00 |
| *** olivierb has quit IRC | 18:02 | |
| clarkb | https://etherpad-dev.openstack.org/p/clarkb-test2 seems to be working from here | 18:03 |
| *** diablo_rojo has quit IRC | 18:03 | |
| clarkb | I'm going to prep for meeting and then look at upgrading etherpad.o.o after the meeting | 18:05 |
| clarkb | I'll delete the old hosts in a few days (to allow local db backups to roll over onto new hosts) | 18:06 |
| *** diablo_rojo has joined #openstack-infra | 18:10 | |
| *** bhavikdbavishi has quit IRC | 18:11 | |
| AJaeger | clarkb, fungi, the security SIG leads proposed 5 weeks ago to abandon anchor, see https://review.openstack.org/#/c/597022/ - but neither of those two leads are in the core group. what about adding them to https://review.openstack.org/#/admin/groups/498,members to allow them to finish the retirement, please? | 18:11 |
| prometheanfire | is it possible to have the certificates api thing to not show up in https://storyboard.openstack.org/#!/project/openstack/requirements (the requirements task is done)? | 18:13 |
| prometheanfire | same for the kafka-python one | 18:14 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add mysql to quick-start https://review.openstack.org/610697 | 18:14 |
| fungi | AJaeger: no objection from me | 18:14 |
| *** Emine has quit IRC | 18:15 | |
| *** e0ne has joined #openstack-infra | 18:15 | |
| AJaeger | gagehugo, see above my comment about anchor, please | 18:15 |
| AJaeger | fungi: some infra-root would still need to do it - but can wait until after the meeting ;) | 18:16 |
| openstackgerrit | Merged openstack-infra/project-config master: Install Pygments before checking README syntax for publish jobs https://review.openstack.org/611014 | 18:18 |
| *** gyee has joined #openstack-infra | 18:19 | |
| gagehugo | AJaeger: that works | 18:19 |
| AJaeger | gagehugo: ok, let's hope an admin adds either you or Luke so that you can approve that change and abandon all open reviews and we can finish retirement | 18:21 |
| clarkb | gagehugo: you have two accounts do you want the gmail or att one added to that group? | 18:21 |
| gagehugo | clarkb: gmail | 18:22 |
| *** e0ne has quit IRC | 18:22 | |
| clarkb | done | 18:22 |
| *** hamzy has quit IRC | 18:23 | |
| clarkb | prometheanfire: I think its waiting for the magnum todo to be set to soem closed state | 18:23 |
| clarkb | I'm nto aware of any other way to indicate a story is no longer active than to update the task state | 18:23 |
| prometheanfire | boo, imo the default view for a project should only have stories with active tasks for that project | 18:24 |
| AJaeger | thanks, clarkb | 18:24 |
| AJaeger | gagehugo: if you need help, feel free to ask me... | 18:25 |
| gagehugo | clarkb: thanks | 18:25 |
| gagehugo | AJaeger: so the next step is to abandon the open reviews then? | 18:26 |
| AJaeger | gagehugo: yes, exactly get this list empty: https://review.openstack.org/#/q/project:openstack/anchor+is:open | 18:26 |
| AJaeger | gagehugo: shall I push the remaining changes for you? | 18:26 |
| AJaeger | gagehugo: or want to follow https://docs.openstack.org/infra/manual/drivers.html#retiring-a-project yourself? | 18:26 |
| gagehugo | AJaeger: I can follow that guide, it will be good practice as I've not done this before | 18:27 |
| gagehugo | I'll def ping you if I have questions | 18:28 |
| AJaeger | gagehugo: so, steps 1-2 are done, you're at step 3 now - and thus need step 4 plus a governance change to finish retirement. | 18:28 |
| AJaeger | gagehugo: ok | 18:29 |
| gagehugo | ok | 18:29 |
| gagehugo | AJaeger: thanks! | 18:29 |
| *** rkukura has joined #openstack-infra | 18:35 | |
| *** jamesmcarthur has quit IRC | 18:38 | |
| fungi | prometheanfire: it's been proposed to rework the project and project group views so that only stories with active tasks for those projects are considered active | 18:41 |
| fungi | or, stated another way, constrain the task query so that only tasks for those projects are taken into consideration when deciding what story state to indicate | 18:41 |
| prometheanfire | fungi: well, have another +1 from me on it :P | 18:42 |
| prometheanfire | or maybe make the query selectable from a dropdown | 18:42 |
| *** jamesmcarthur has joined #openstack-infra | 18:52 | |
| *** bobh has quit IRC | 18:54 | |
| *** bobh has joined #openstack-infra | 18:56 | |
| *** jamesmcarthur has quit IRC | 18:56 | |
| *** ssbarnea_ has joined #openstack-infra | 18:57 | |
| clarkb | infra meeting over in #openstack-meeting in just a few minutes | 18:57 |
| *** bobh has quit IRC | 19:01 | |
| *** rkukura has quit IRC | 19:04 | |
| *** dave-mccowan has quit IRC | 19:07 | |
| *** trown|lunch is now known as trown | 19:07 | |
| *** rkukura has joined #openstack-infra | 19:07 | |
| *** dave-mccowan has joined #openstack-infra | 19:08 | |
| *** auristor has quit IRC | 19:19 | |
| *** auristor has joined #openstack-infra | 19:20 | |
| *** dkehn has quit IRC | 19:25 | |
| *** jamesmcarthur has joined #openstack-infra | 19:25 | |
| *** bobh has joined #openstack-infra | 19:32 | |
| *** dkehn has joined #openstack-infra | 19:33 | |
| *** auristor has quit IRC | 19:34 | |
| *** jamesmcarthur has quit IRC | 19:35 | |
| *** bobh has quit IRC | 19:37 | |
| openstackgerrit | Merged openstack-infra/puppet-cgit master: Filter out link-local ipv6 address in loadbalancer https://review.openstack.org/576262 | 19:44 |
| *** rkukura has quit IRC | 19:45 | |
| *** rkukura has joined #openstack-infra | 19:48 | |
| openstackgerrit | Merged openstack-infra/puppet-cgit master: Update Gemfile for Zuulv3 https://review.openstack.org/571987 | 19:49 |
| *** auristor has joined #openstack-infra | 19:52 | |
| *** icey has quit IRC | 19:52 | |
| *** jamesmcarthur has joined #openstack-infra | 19:54 | |
| *** hamzy has joined #openstack-infra | 19:54 | |
| *** jamesmcarthur has quit IRC | 19:56 | |
| *** jamesmcarthur has joined #openstack-infra | 19:56 | |
| *** auristor has quit IRC | 19:57 | |
| *** eernst has joined #openstack-infra | 19:59 | |
| *** olivierb has joined #openstack-infra | 19:59 | |
| *** eernst has quit IRC | 20:01 | |
| *** auristor has joined #openstack-infra | 20:02 | |
| corvus | fungi, clarkb: i was just looking into the super-weird bounce message we got for openstack-infra@lists and i think we broke verp with https://review.openstack.org/484989 | 20:02 |
| *** eernst has joined #openstack-infra | 20:02 | |
| corvus | i think i can fix it | 20:02 |
| *** eernst has joined #openstack-infra | 20:03 | |
| fungi | corvus: the gmail deferrals for catchrobot at mirantis? | 20:04 |
| *** eernst_ has joined #openstack-infra | 20:04 | |
| corvus | fungi: yep | 20:04 |
| corvus | catchrobot is not subscribed to the list; i don't know what address we sent to that goes there, but it was one of a batch of something like 30 addrs at mirantis. verp generates unique bounce addresses to prevent this. | 20:05 |
| *** bobh has joined #openstack-infra | 20:06 | |
| clarkb | etherpad01.openstack.org is booting now | 20:07 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: lists: run mailman verp router before dnslookup https://review.openstack.org/611155 | 20:07 |
| corvus | fungi, clarkb: ^ | 20:07 |
| *** eernst has quit IRC | 20:08 | |
| *** bobh has quit IRC | 20:12 | |
| *** jamesmcarthur has quit IRC | 20:12 | |
| fungi | how did moving the mailman_router entry have that impact? | 20:13 |
| mordred | corvus: 2x+2 - but no +A bcause you said "merge with caution" | 20:13 |
| fungi | ahh, yep, it was the verp router | 20:14 |
| corvus | fungi: yeah, we only needed to move the mailman router, but we moved both | 20:15 |
| clarkb | I'm running kick.sh on new etherpad01 server now | 20:15 |
| corvus | mordred: thx, i approved, i should be around for a while... depending on check queue. | 20:15 |
| clarkb | if this finishes successfully my next step will be to stop apache and nodejs on the old server then update dns | 20:15 |
| clarkb | you may notice a short etherpad service interruption while that happens | 20:15 |
| corvus | if i leave before it merges, i'll either make sure someone is around or WIP it till tomorrow | 20:16 |
| fungi | corvus: thanks, my ability to monitor it is limited at the moment, but lgtm! | 20:16 |
| corvus | this won't cause those bounces to stop, since they are for an already delivered message; we'll have to wait for gmail to time out on that. | 20:18 |
| mriedem | clarkb: when you get some leisure time, i think you might want to get in on this hot hot placement d-g action https://review.openstack.org/#/c/606853/ | 20:18 |
| clarkb | mriedem: ya reviewing cdents list of changes was on my todo list. do we also need to add placement to the enabled services list or is it already there? | 20:19 |
| *** jamesmcarthur has joined #openstack-infra | 20:19 | |
| mriedem | it's been in d-g for awhile now | 20:19 |
| mriedem | since ocata | 20:19 |
| rpioso | In gertty, is there a way to follow up on comments left on a patch set other than the most recent one? I appear to have successfully added the comments; however, when I submitted the review via 'r', they don't appear in Gerrit's web interface or the change's front screen in gertty. | 20:23 |
| openstackgerrit | Ian Wienand proposed openstack-infra/nodepool master: Use zuul-sphinx for configuration layout https://review.openstack.org/604274 | 20:25 |
| openstackgerrit | Ian Wienand proposed openstack-infra/nodepool master: Add overview of config options https://review.openstack.org/604984 | 20:25 |
| clarkb | just waiting for the dns timeout to run out on etherpad.openstack.org so that you end up talking to the new server | 20:26 |
| clarkb | once I've got that running I'll make sure bup is happy | 20:26 |
| clarkb | https://etherpad.openstack.org/p/clarkb-test is my long running test etherpad | 20:29 |
| clarkb | if you see the new lines in pad and in chat about post xenail upgrade I think we are set | 20:29 |
| clarkb | rpioso: is it posible you haven't synced yet? | 20:30 |
| clarkb | rpioso: there should be a little sync todo counter in the top right | 20:30 |
| clarkb | if that is greater than 0 then that could epxlain it | 20:30 |
| *** rkukura has quit IRC | 20:30 | |
| *** rkukura_ has joined #openstack-infra | 20:30 | |
| *** rkukura_ has quit IRC | 20:31 | |
| rpioso | clarkb: I don't believe that's the issue. The first time I tried, comments for the most recent patch set were successfully submitted. And those were added after the ones on the older patch set. | 20:31 |
| rpioso | I was hoping the comments on both patch sets would be submitted. | 20:32 |
| *** rkukura has joined #openstack-infra | 20:32 | |
| *** kgiusti has left #openstack-infra | 20:33 | |
| *** slaweq has quit IRC | 20:33 | |
| clarkb | backups have been disabled on old etherpad.o.o server and I am running a start backup on etherpad01 now to make sure it works | 20:37 |
| *** bobh has joined #openstack-infra | 20:38 | |
| clarkb | ok I think etherpad server upgrades are all done except for deleting the old servers which I wanted to keep around for a day or so just in case we have to revert | 20:41 |
| *** ssbarnea has quit IRC | 20:42 | |
| *** ssbarne__ has joined #openstack-infra | 20:42 | |
| *** rkukura has quit IRC | 20:42 | |
| *** ssbarnea has joined #openstack-infra | 20:42 | |
| *** hamzy has quit IRC | 20:44 | |
| *** ssbarnea_ has quit IRC | 20:44 | |
| *** ansmith has quit IRC | 20:45 | |
| clarkb | infra-root https://review.openstack.org/#/c/529875/ and its parent will simplify things a bit to make the logstash.o.o upgrade to xenial easier. If you are able to review those two changes I can approve them when able to monitor them | 20:47 |
| fungi | rpioso: expand the patch set details for the earlier patch and there should be a review button for it | 20:48 |
| *** jesusaur has joined #openstack-infra | 20:48 | |
| clarkb | ianw: https://review.openstack.org/#/c/611075/ is a change that may interest you. We had another corrupt sqlalchemy utils whel file. This change adds a zip validity checker to our build/publish jobs for wheels | 20:49 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Simplify Builds page https://review.openstack.org/610137 | 20:49 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add more information to build page https://review.openstack.org/610138 | 20:49 |
| rpioso | fungi: Sure enough. Beware! Student driver! | 20:50 |
| rpioso | clarkb, fungi: Thank you. | 20:51 |
| fungi | np | 20:51 |
| ianw | clarkb: hrm ... did you find in the logs what went wrong? | 20:51 |
| *** bobh has quit IRC | 20:52 | |
| clarkb | ianw: I did not, the wheel build fine according to the log I could see | 20:52 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add opendev nameservers (1/2) https://review.openstack.org/605092 | 20:52 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add opendev nameservers (2/2) https://review.openstack.org/610066 | 20:52 |
| rpioso | fungi: No CodeReview{-1,0,+1} option is provided on that screen. | 20:54 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Indicate whether a build is voting in the logs https://review.openstack.org/605857 | 20:54 |
| *** eernst_ has quit IRC | 20:54 | |
| fungi | rpioso: correct, gerrit won't allow you to vote on labels for non-current patchsets. this mimics the behavior of the webui | 20:56 |
| corvus | but you can still leave comments, and clicking save there should do it | 20:56 |
| rpioso | corvus, fungi: Got it! Thanks, again :) | 20:56 |
| rpioso | I'll take an additional action to vote. | 20:57 |
| *** priteau has quit IRC | 20:57 | |
| *** ssbarne__ has quit IRC | 20:57 | |
| *** eernst has joined #openstack-infra | 20:58 | |
| fungi | rpioso: yes, you can (with separate actions either in gertty or the gerrit webui) comment on an earlier patch set and vote on the current patch set | 20:59 |
| *** priteau has joined #openstack-infra | 20:59 | |
| *** sshnaidm has joined #openstack-infra | 21:01 | |
| *** sshnaidm is now known as sshnaidm|off | 21:01 | |
| *** mriedem is now known as mriedem_away | 21:02 | |
| rpioso | I expanded that old patch set with the new comments, selected Review, entered a message, and saved. The comments have not been submitted, yet. Does that sound right? | 21:03 |
| clarkb | ianw http://logs.openstack.org/periodic/git.openstack.org/openstack/requirements/master/publish-wheel-mirror-ubuntu-xenial/25125c0/ is the log fwiw, you may know how to navigate it better | 21:03 |
| *** eharney has quit IRC | 21:03 | |
| clarkb | but the failed.txt for python2 and 3 are both empty | 21:03 |
| *** rkukura has joined #openstack-infra | 21:04 | |
| *** trown is now known as trown|outtypewww | 21:04 | |
| *** munimeha1 has joined #openstack-infra | 21:06 | |
| rpioso | Hrm It looks like the draft messages have vanished. | 21:06 |
| fungi | rpioso: they should be submitted on the next sync | 21:06 |
| fungi | at which point they'll no longer be draft comments | 21:07 |
| fungi | well, as soon as they're saved they should no longer be draft comments from gertty's perspective | 21:07 |
| openstackgerrit | Merged openstack-infra/zuul master: web: expand status panels only when the header is clicked https://review.openstack.org/609348 | 21:09 |
| openstackgerrit | Merged openstack-infra/zuul master: web: fix stream url from status json https://review.openstack.org/609365 | 21:13 |
| *** kaiokmo has quit IRC | 21:17 | |
| *** rkukura has quit IRC | 21:17 | |
| *** rkukura has joined #openstack-infra | 21:19 | |
| *** kaiokmo has joined #openstack-infra | 21:19 | |
| *** jamesmcarthur has quit IRC | 21:20 | |
| *** jamesmcarthur has joined #openstack-infra | 21:22 | |
| *** bobh has joined #openstack-infra | 21:28 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Extract pep8 messages for inline comments https://review.openstack.org/589634 | 21:29 |
| corvus | mordred, AJaeger: ^ that should pass now | 21:29 |
| *** rkukura has quit IRC | 21:30 | |
| *** rh-jelabarre has quit IRC | 21:32 | |
| mordred | corvus: ooh! | 21:33 |
| *** bobh has quit IRC | 21:33 | |
| ianw | gary_perkins: thanks for credentials; that's all setup. I've put some comments on ssl in https://review.openstack.org/#/c/602436 | 21:35 |
| smcginnis | Hah, Duncan on the Cinder team had suggested commenting on reviews with the pep8 errors and was told it was a horrible idea. | 21:35 |
| smcginnis | It all takes time. | 21:35 |
| smcginnis | Oh, I see that's slightly different than I assumed with the summary line. :) | 21:36 |
| mordred | corvus: I pushed up https://review.openstack.org/610744 DNM Testing inline pep8 with a depends-on | 21:37 |
| mordred | corvus: and added several juicy pep8 errors | 21:37 |
| mordred | corvus: including but not limited to trailing whitespace, incorrect indentation - and SHOCK, no space after a comma! | 21:37 |
| mordred | smcginnis: what did you think it was going to be? | 21:38 |
| smcginnis | mordred: Adding comments to the review to point out the pep8 errors. Or is it? | 21:39 |
| mordred | yup. leaving in-line source code comments with the pep8 error message | 21:40 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config master: Add opendev nameservers (1/2) https://review.openstack.org/605092 | 21:41 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config master: Add opendev nameservers (2/2) https://review.openstack.org/610066 | 21:41 |
| clarkb | I think that may make puppet happy in testing | 21:41 |
| smcginnis | mordred: OK, then it is the same as suggested as before. Personally, I think that will be very useful. | 21:41 |
| mordred | me too | 21:41 |
| mordred | WAY less clicky-clicky to figure out what the error was | 21:42 |
| *** lbragstad is now known as lbragstad-503 | 21:42 | |
| smcginnis | Especially for those where clicking on the job link and then clicking on the log link is one too many clicks for the day. | 21:42 |
| mordred | smcginnis: out of curiosity, what was the concern? | 21:42 |
| corvus | i don't remember a conversation about that, but perhaps it was before gerrit had in-line comment api support? | 21:42 |
| smcginnis | mordred: This is like 2-3 years ago now, but from what I remember there was an unconvincing but strong opinion from someone that there was no need to go to such lengths when it's all in the log file. | 21:43 |
| smcginnis | corvus: That could very well be too. | 21:43 |
| mordred | smcginnis: ah. so it was more of "that seems like a waste of effort" rather than "this will physically impede progress towards joy" | 21:43 |
| clarkb | fwiw I think learning to read logs is a valuable skill/tool, but if computers can reduce overhead they should | 21:44 |
| mordred | to be fair - it did take us 2-3 years to implement the feature :) | 21:44 |
| smcginnis | mordred: The former with the attitude of the latter. | 21:44 |
| openstackgerrit | Gage Hugo proposed openstack-infra/project-config master: Retire project Anchor https://review.openstack.org/611181 | 21:44 |
| mordred | clarkb: oh - also, per our conversaiton the other day about stackviz - it works _perfectly_ fine with other tox runs | 21:44 |
| clarkb | neat | 21:44 |
| *** hasharAway has quit IRC | 21:44 | |
| *** munimeha1 has quit IRC | 21:45 | |
| mordred | http://logs.openstack.org/67/610167/3/check/openstacksdk-functional-devstack/4c5c356/controller/logs/stackviz/#/stdin <-- there is an example of using it with openstacksdk functional tests | 21:45 |
| mordred | clarkb: SO - given an extremely small sample size - I think we _could_ add a dstat role to the base tox job pre and add the stackviz role to the base tox job post ... and get stackviz reports for every tox-based thing with subunit files | 21:47 |
| rpioso | fungi: I'm good, and learned more about gertty :) | 21:47 |
| mordred | deciding we should is a whole other thing - but in terms of could - it should totally work | 21:47 |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config master: Revert "Reduce gate window floor to 10" https://review.openstack.org/611183 | 21:47 |
| openstackgerrit | Gage Hugo proposed openstack-infra/project-config master: Retire project Anchor https://review.openstack.org/611181 | 21:48 |
| *** sshnaidm|off has quit IRC | 21:48 | |
| clarkb | mordred: your stackviz report doesn't seem to have the dstat info, is that because you don't have dstat info in that job? | 21:48 |
| mordred | clarkb: I think we noticed the other day that the dstat info seems to be missing even from normal ones? | 21:48 |
| clarkb | oh maybe | 21:49 |
| clarkb | mtreinish: may know why | 21:49 |
| mordred | http://logs.openstack.org/91/608091/6/check/tempest-full/9f77a01/controller/logs/stackviz/#/stdin/timeline?test=tempest.api.compute.admin.test_servers.ServersAdminTestJSON.test_rebuild_server_in_error_state | 21:49 |
| mordred | clarkb: ^^ that's from tempest-full in the nova repo | 21:50 |
| * mordred assumes mtreinish knows all things | 21:50 | |
| mordred | especially related to subunit streams | 21:51 |
| *** slaweq has joined #openstack-infra | 21:53 | |
| openstackgerrit | Gage Hugo proposed openstack-infra/project-config master: Retire project Anchor - step 4 https://review.openstack.org/611181 | 21:53 |
| openstackgerrit | Ben Nemec proposed openstack-infra/project-config master: Fix Pygments install for release check https://review.openstack.org/611184 | 21:54 |
| clarkb | ianw: gary_perkins left this in gerrit as a comment too but I think we should try ot setup https even if it is a self signed cert or snakeoil cert. We can give that cert to oscc clouds.yaml directly to have it verify the connection that way | 21:54 |
| clarkb | its a bit hacky but should still prevent mitm and encrypt the data path | 21:55 |
| *** mriedem_away has quit IRC | 21:55 | |
| *** smcginnis is now known as smcginnis_vaca | 21:55 | |
| clarkb | other clouds have used letsencrypt | 21:55 |
| *** priteau has quit IRC | 22:03 | |
| *** slaweq has quit IRC | 22:09 | |
| gagehugo | AJaeger: Is there any other change in governance needed for anchor? https://review.openstack.org/#/c/611187/ | 22:09 |
| clarkb | mordred: browser debugger says invalid dstat csv | 22:11 |
| mordred | oh! | 22:11 |
| clarkb | mordred: if I had to guess its the header causing it | 22:11 |
| mordred | well, that's at least something | 22:11 |
| *** slaweq has joined #openstack-infra | 22:11 | |
| clarkb | I'm going to move the dstat csv file aside for that tempest one and edit the csv file to remove the header and see if that does it | 22:12 |
| mordred | clarkb: ++ | 22:12 |
| clarkb | no that isn't helping it | 22:15 |
| clarkb | I also tried unzipping it | 22:15 |
| mordred | hrm | 22:15 |
| mordred | clarkb: my browser does not show that you edited the file | 22:15 |
| clarkb | oh maybe there is a cache acting on it | 22:16 |
| clarkb | libreoffice is fine with the csv parsing | 22:18 |
| *** boden has quit IRC | 22:18 | |
| clarkb | reading the js it almost seems like its looking for Dstat to lead every row? | 22:19 |
| clarkb | it is actually looking for that header that is good to know | 22:20 |
| mordred | clarkb: oh - it pulls the file from stackviz/dstat.csv | 22:24 |
| clarkb | the gzip part is fine I think | 22:25 |
| mordred | sorry - when I was refreshing I was looking at the one in the logs dir | 22:25 |
| clarkb | there are quota in the header that are not expected by stackviz | 22:25 |
| *** rkukura has joined #openstack-infra | 22:25 | |
| clarkb | ya I moved it around I'm about to gzip it again to test it works if zipped | 22:26 |
| clarkb | it does not | 22:26 |
| *** jamesmcarthur has quit IRC | 22:26 | |
| clarkb | ok the quotes are fine its the gzipped file it doesn't like | 22:26 |
| clarkb | I have unzipped it and it works, gzipped it and it doesn't | 22:27 |
| clarkb | I think because the config.json says the dstat file is not gizipped | 22:27 |
| clarkb | I think we solve this if we have stackviz do the gzipping which was buggy when I was fixing things on python3 | 22:27 |
| mordred | and I guess the javascript isn't sending deflate headers? | 22:28 |
| clarkb | I think the config.json file in the data dir is what it uses to detminer stuff | 22:28 |
| clarkb | and that says text/csv | 22:28 |
| mordred | nod | 22:28 |
| clarkb | let me gzip again and see the server response headers | 22:28 |
| clarkb | they are application/g-zip on return | 22:29 |
| clarkb | I'm guessing it doesnt' check what the server says and jsut goes off of its config | 22:29 |
| mordred | where is the javascript? | 22:30 |
| *** timothyb89 has joined #openstack-infra | 22:30 | |
| clarkb | in the openstack/stackviz repo | 22:30 |
| mordred | yah - have you found where it does the fetch yet? | 22:31 |
| clarkb | or do you mean the stuff we serve? the served stuff is all minimized and impossible to read | 22:31 |
| mordred | oh - I meant the source | 22:31 |
| mordred | also - we can serve map files with this to make it debuggable in the browser | 22:31 |
| mordred | but that'll take me slightly longer to sort out than right now | 22:31 |
| *** eernst has joined #openstack-infra | 22:31 | |
| clarkb | mordred: app/js/directives/timeline.js: look for dstatLoaded | 22:32 |
| mordred | thanks | 22:32 |
| clarkb | that broadcasts the dstatloaded even then we trigger an action on that event elesewhere but I think its the stuff in that file that is failing | 22:32 |
| *** pbourke has quit IRC | 22:35 | |
| *** pbourke has joined #openstack-infra | 22:35 | |
| *** jamesmcarthur has joined #openstack-infra | 22:38 | |
| clarkb | I'm quickly lost in what I'm assuming are angular magic things for fetching and decoding resources | 22:38 |
| clarkb | dataset.js seems to do the actual GET | 22:38 |
| mordred | clarkb: yes - and from what I can read the browser should do the gzip handling before the data gets to javascript | 22:39 |
| mordred | but maybe we need to send an accept-encoding: gzip header? | 22:40 |
| mordred | clarkb: or - jeez, this hurts my brain | 22:41 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config master: Add afs client docs for non Debuntu https://review.openstack.org/601855 | 22:41 |
| clarkb | fungi: ianw ^ how about that for specifying how to kinit as admin? | 22:42 |
| clarkb | mordred: my browser does send accept-encoding gzip | 22:42 |
| clarkb | or maybe it was just deflate | 22:42 |
| *** jamesmcarthur has quit IRC | 22:43 | |
| *** slaweq has quit IRC | 22:44 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/storyboard master: Build containers using pbrx https://review.openstack.org/611191 | 22:46 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config master: Add afs client docs for non Debuntu https://review.openstack.org/601855 | 22:46 |
| *** rcernin has joined #openstack-infra | 22:49 | |
| *** agopi is now known as agopi|brb | 22:52 | |
| *** jamesmcarthur has joined #openstack-infra | 22:54 | |
| *** agopi|brb has quit IRC | 22:56 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/storyboard master: Build containers using pbrx https://review.openstack.org/611191 | 22:57 |
| openstackgerrit | Merged openstack-infra/system-config master: lists: run mailman verp router before dnslookup https://review.openstack.org/611155 | 23:01 |
| openstackgerrit | Monty Taylor proposed openstack-infra/storyboard master: Build containers using pbrx https://review.openstack.org/611191 | 23:01 |
| openstackgerrit | Monty Taylor proposed openstack-infra/storyboard master: Fix up a few requirements https://review.openstack.org/611194 | 23:01 |
| *** jamesmcarthur has quit IRC | 23:07 | |
| *** jamesmcarthur has joined #openstack-infra | 23:07 | |
| *** eernst has quit IRC | 23:09 | |
| *** eernst has joined #openstack-infra | 23:09 | |
| *** tosky has quit IRC | 23:10 | |
| *** eernst has quit IRC | 23:10 | |
| *** slaweq has joined #openstack-infra | 23:11 | |
| *** jamesmcarthur has quit IRC | 23:18 | |
| *** jamesmcarthur has joined #openstack-infra | 23:21 | |
| *** dave-mccowan has quit IRC | 23:24 | |
| *** hamzy has joined #openstack-infra | 23:24 | |
| *** jamesmcarthur has quit IRC | 23:32 | |
| *** rkukura has quit IRC | 23:41 | |
| *** slaweq has quit IRC | 23:44 | |
| *** jamesmcarthur has joined #openstack-infra | 23:44 | |
| clarkb | corvus: https://review.openstack.org/#/c/605092/5 passes now for opendev dns | 23:45 |
| *** ssbarnea|bkp2 has joined #openstack-infra | 23:45 | |
| *** ssbarnea has quit IRC | 23:46 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!