| *** LinuxMe has joined #kata-dev | 00:15 | |
| *** LinuxMe has quit IRC | 00:21 | |
| *** zerocoolback has joined #kata-dev | 01:03 | |
| *** zerocoolback has quit IRC | 01:09 | |
| *** dlw has joined #kata-dev | 01:24 | |
| *** cdent has quit IRC | 01:32 | |
| *** LinuxMe has joined #kata-dev | 01:33 | |
| *** LinuxMe has quit IRC | 01:38 | |
| *** LinuxMe_ has joined #kata-dev | 02:07 | |
| *** zerocoolback has joined #kata-dev | 03:08 | |
| *** zerocoolback has quit IRC | 03:16 | |
| *** zerocoolback has joined #kata-dev | 03:32 | |
| *** zerocoolback has quit IRC | 03:38 | |
| *** zerocoolback has joined #kata-dev | 03:47 | |
| changcheng | Hi | 03:56 |
|---|---|---|
| changcheng | Can't sync kata-container/runtime by following the developer guide | 03:57 |
| changcheng | go get -d -u github.com/kata-containers/runtime/ | 03:57 |
| changcheng | Error: package github.com/kata-containers/runtime: no Go files in /home/ryoma/LINUX/go_repo/src/github.com/kata-containers/runtime | 03:57 |
| changcheng | How to resolve it? | 03:57 |
| kata-dev-irc-bot | <xu> networking issue? | 03:59 |
| changcheng | No | 04:00 |
| changcheng | Other 4 projects could be cloned by following the developer guide | 04:00 |
| changcheng | Same error when executing: go get -d -u github.com/kata-containers/osbuilder | 04:00 |
| *** LinuxMe_ has quit IRC | 04:08 | |
| *** sjas_ has joined #kata-dev | 04:30 | |
| *** sjas has quit IRC | 04:33 | |
| *** LinuxMe has joined #kata-dev | 04:43 | |
| *** LinuxMe has quit IRC | 04:48 | |
| *** LinuxMe has joined #kata-dev | 05:15 | |
| *** zerocoolback has quit IRC | 05:18 | |
| *** zerocoolback has joined #kata-dev | 05:18 | |
| *** zerocoolback has quit IRC | 05:19 | |
| *** zerocoolback has joined #kata-dev | 05:19 | |
| *** zerocoolback has joined #kata-dev | 05:20 | |
| *** zerocoolback has quit IRC | 05:20 | |
| *** LinuxMe has quit IRC | 05:21 | |
| *** zerocoolback has joined #kata-dev | 05:21 | |
| *** zerocoolback has joined #kata-dev | 05:22 | |
| *** zerocoolback has quit IRC | 05:22 | |
| *** dklyle has quit IRC | 05:59 | |
| *** jodh has joined #kata-dev | 06:19 | |
| *** jodh has quit IRC | 06:19 | |
| *** jodh has joined #kata-dev | 06:19 | |
| kata-dev-irc-bot | <xu> @eric.ernst as talked in the meeting, @fupan from hyper.sh team could join the infra team and serve for the asia timezones. who is leading the team in intel side now? | 07:55 |
| *** dklyle has joined #kata-dev | 07:57 | |
| *** gwhaley has joined #kata-dev | 08:00 | |
| *** davidgiluk has joined #kata-dev | 08:00 | |
| *** dklyle has quit IRC | 08:09 | |
| *** dklyle has joined #kata-dev | 08:12 | |
| *** LinuxMe has joined #kata-dev | 08:14 | |
| *** LinuxMe has quit IRC | 08:19 | |
| *** dklyle has quit IRC | 08:29 | |
| *** dklyle has joined #kata-dev | 08:29 | |
| changcheng | Do you mean the guys work in ASIA TZ at Intel? | 08:41 |
| changcheng | Does anyone build the osbuild/rootfs on ubuntu os host? | 08:44 |
| changcheng | Since the bare metal ubuntu doesn't support the build, I try to build it in docker. But I always meet issue which maybe network proxy issue. | 08:45 |
| changcheng | I execute below command and get the error: | 08:45 |
| changcheng | command: $ sudo -E ./rootfs-builder/rootfs.sh -r "${PWD}/fedora_rootfs" fedora | 08:46 |
| changcheng | Info: Step 2/6 : RUN dnf -y update && dnf install -y git redhat-release systemd pkgconfig gcc coreutils | 08:46 |
| changcheng | Error: he command '/bin/sh -c dnf -y update && dnf install -y git redhat-release systemd pkgconfig gcc coreutils' returned a non-zero code: 1 | 08:46 |
| kata-dev-irc-bot | <xu> no, the initial kata developers come from intel and hyper.sh, intel team are in US and EU, while the devs of hyper.sh team are in utc+0800. | 08:49 |
| kata-dev-irc-bot | <xu> and wei from huawei is in utc+08 as well | 08:49 |
| *** LinuxMe has joined #kata-dev | 08:51 | |
| gwhaley | Hi @changcheng - osbuilder has built in 'build in a docker image' option to allow building on a host that is not configured for native build. I use that on Fedora... let me grab a link | 08:52 |
| gwhaley | here: https://github.com/kata-containers/osbuilder#usage - see the 'USE_DOCKER' option - you set that on the host side, and it will do the full build in a docker image | 08:52 |
| changcheng | OK. Hope there's asian team from that company could also attend this project. | 08:53 |
| gwhaley | Note - you *must* have runc as the default docker runtime for this to work though | 08:53 |
| changcheng | My docker configuratin: /usr/bin/dockerd -D --add-runtime kata-runtime=/usr/bin/kata-runtime --default-runtime=kata-runtime | 08:53 |
| changcheng | So I have to use the default docker runtime on ubuntu, right? | 08:54 |
| *** LinuxMe has quit IRC | 08:56 | |
| gwhaley | @changcheng: you need runc to be the default on your host to build osbuilder in a USE_DOCKER image, yes. You can just change to '--default-runtime=runc', and reload/reset docker with systemctl and | 09:00 |
| gwhaley | that should do it. | 09:00 |
| gwhaley | All the kata tests etc. should still work, as they normally explicitly set --runtime=kata on the docker command line | 09:00 |
| gwhaley | you either need to 1) set runc as the default runtime and use USE_DOCKER=1, or 2) install all the required components for the native build onto your host | 09:01 |
| changcheng | Prefer to option 1), But | 09:02 |
| * gwhaley wonders if 'kata' is still coming out as 'j f r a z e l l' on the slack channel via the irc-bot - anybody? | 09:02 | |
| changcheng | In the document of "ubuntu-installation-guide.md", the runtime is replaced with kata-runtime. Will it be conflict if I install default docker runtime and kata-runtime on the same native host? | 09:05 |
| *** LinuxMe has joined #kata-dev | 09:05 | |
| kata-dev-irc-bot | <xu> it displays the mention link normally @graham.whaley | 09:05 |
| changcheng | Yes. I've read it and follow the guides. | 09:06 |
| changcheng | I want to know if there's conflict that I install the default runtime and kata-runtime on the same host. The "ubuntu-installation-guide.md" file asks to remove the other runtime and install kata-runtime instead. | 09:07 |
| kata-dev-irc-bot | <xu> normally they could co-exist on a same host, and you may choose one to launch | 09:08 |
| changcheng | OK. Feedback later. | 09:08 |
| gwhaley | @changcheng - I think the document is asking to remove old docker and install a new one - we never remove the docker default runtime from the system - we change which runtime is the default to use, but the other | 09:09 |
| gwhaley | (runc) runtime is still available. | 09:09 |
| gwhaley | If you do a 'docker info', you should see them both listed, and which one is the default | 09:09 |
| *** LinuxMe has quit IRC | 09:10 | |
| *** zerocoolback has joined #kata-dev | 09:23 | |
| changcheng | Bad luck. The runtime has replaced with runc, but It still meet with the same issue. Check the issue history later. | 09:40 |
| *** LinuxMe has joined #kata-dev | 09:41 | |
| gwhaley | @changcheng: as we thought maybe you were having proxy issues - it does look like osbuilder does try to pass the host proxy settings into the docker build: https://github.com/kata-containers/osbuilder/blob/a12618811dd715c2d4438ce3c64ce6bd9fe7c630/image-builder/image_builder.sh#L92 | 09:45 |
| *** LinuxMe has quit IRC | 09:45 | |
| gwhaley | maybe you can use a 'set -x' on the scripts to get more verbose output to check they are being passed in correctly (just a guess - maybe you are using HTTP_PROXY on the host and the script is looking for http_proxy) iyswim etc. | 09:46 |
| *** zerocoolback has quit IRC | 09:53 | |
| kata-dev-irc-bot | <kaly.xin> @eric.ernst thanks for the help from @xu, we can use one of their server on packet.net for CI on arm. To kick off the process, I'd like to ask a few basic questions: 1) If I'm right, there are 2 ways to create the connection between a master and a slave, one is SSH, the other is JNLP. Which way is used by the Jenkin master for kata? If it's the first way, do you need username and passwd or username and private key? 2) what kind of | 10:25 |
| kata-dev-irc-bot | software is needed to install on the slave in advance? Java, ant, maven? Or master can push the software to install on slave? 3) Will the tester on the jenkin slave have some kind of rights to configure the jobs from the master jenkins web page or all the jobs are configured by master administrator? 4) What's the information for the first step you need from us? (I guess it's our IP, username, passwd or private key?), any other information, | 10:25 |
| kata-dev-irc-bot | please let us know. Thanks! | 10:25 |
| * gwhaley just notes for @kaly.xin @xu that, so far, for the Jenkins for x86/amd64 we use the SSH methods, and we have our Azure slaves pre-configured with some minimum packages via a cloud-init | 10:29 | |
| gwhaley | then the CI scripts from the tests repository (https://github.com/kata-containers/tests/tree/master/.ci) install all the other deps they need | 10:30 |
| gwhaley | but - I suspect you already have something on packet.net you can model on - so there may be differences | 10:30 |
| gwhaley | @kaly.xin - there is a little information in this pending PR that may help: https://github.com/kata-containers/ci/pull/32 | 10:31 |
| gwhaley | https://github.com/grahamwhaley/kata-containers-ci/blob/0f97abb3ee911b8373252a59c8cbe605e000a692/Jenkins_setup.md | 10:31 |
| kata-dev-irc-bot | <xu> @graham.whaley do we create a new slave instance in Azure for every new PR update? | 10:31 |
| gwhaley | @xu - for x86, yes :-) Means we get a nice clean repeatable experience - but, at the cost of setup time of course :-( | 10:32 |
| gwhaley | we have seen (on a bare metal metrics CI system), that some PRs could corrupt the docker install etc. or leave things running (like proxy etc.), and that affected later builds/runs - so it is much safer and | 10:33 |
| gwhaley | more stable to have a fresh machine for each build | 10:33 |
| gwhaley | That is why I have been putting together the metric VM build scripts on the PR on the ci repo (github just broke on me - sorry, no link...) | 10:34 |
| kata-dev-irc-bot | <xu> hmm… so we need some work to get it works on packet.net | 10:37 |
| gwhaley | @xu - ah, you currently have long-running instances I guess. I see three options maybe: | 10:38 |
| gwhaley | 1 - use them as-is, and watch out for broken builds affecting later builds (not ideal...) | 10:39 |
| gwhaley | 2 - reboot machines between runs (this normally fixed any corruption - stuck processes or hung up docker etc.) | 10:39 |
| kata-dev-irc-bot | <xu> in the weekly meeting , we tend to run the tests in a container | 10:39 |
| gwhaley | 3 - maybe use the VM instances from the PR I posted | 10:40 |
| gwhaley | right - I consided if I could use docker-in-docker or docker under lxc or similar to get isolation for each run | 10:40 |
| kata-dev-irc-bot | <xu> the existing arm servers do not support nested virtualization yet | 10:41 |
| gwhaley | but it looked like running overlay-on-overlay etc. was not a good idea | 10:41 |
| gwhaley | so, there may be a solution where you can run 'a container in a container' - but, I went with the VM option | 10:41 |
| *** gwhaley has quit IRC | 10:53 | |
| *** dklyle has quit IRC | 11:02 | |
| *** dlw has quit IRC | 11:07 | |
| *** LinuxMe has joined #kata-dev | 11:25 | |
| *** LinuxMe has quit IRC | 11:30 | |
| *** dklyle has joined #kata-dev | 11:36 | |
| *** LinuxMe has joined #kata-dev | 12:00 | |
| *** LinuxMe has quit IRC | 12:04 | |
| *** gwhaley has joined #kata-dev | 12:08 | |
| *** zerocoolback has joined #kata-dev | 12:27 | |
| *** fuentess has joined #kata-dev | 12:29 | |
| *** LinuxMe has joined #kata-dev | 12:35 | |
| *** LinuxMe has quit IRC | 12:39 | |
| *** LinuxMe_ has joined #kata-dev | 12:48 | |
| kata-dev-irc-bot | <eric.ernst> @changcheng.liu this sounds like proxy connectivity within the container itself. | 12:55 |
| kata-dev-irc-bot | <eric.ernst> Run an Ubuntu container normally and check it you can run apt-get update, for example. | 12:57 |
| kata-dev-irc-bot | <eric.ernst> Usually there’s a separate proxy setting specific to docker... | 12:57 |
| *** LinuxMe_ has quit IRC | 12:59 | |
| *** sjas_ is now known as sjas | 13:16 | |
| *** LinuxMe has joined #kata-dev | 13:36 | |
| *** LinuxMe has quit IRC | 13:41 | |
| *** lamego has joined #kata-dev | 13:45 | |
| *** david-lyle has joined #kata-dev | 13:56 | |
| *** dklyle has quit IRC | 13:58 | |
| *** devimc has joined #kata-dev | 14:14 | |
| *** LinuxMe has joined #kata-dev | 14:16 | |
| *** LinuxMe has quit IRC | 14:21 | |
| *** annabelleB has joined #kata-dev | 14:28 | |
| annabelleB | gwhaley: to answer your other question…yes, it’s still happening :( i added a username between JF and k-a-t-a to see what happened; didn’t seem to make a difference | 14:30 |
| gwhaley | annabellB :-( A short while ago we had many slack and few irc folks, so it was not so bad (maybe) - but now we are gathering more irc folks as well :-) I guess we have somebody we can tickle on the irc bot integration to have a look? | 14:35 |
| * gwhaley resists typing k-a-t-a over and over :-) | 14:35 | |
| *** annabelleB has quit IRC | 14:41 | |
| *** LinuxMe has joined #kata-dev | 14:41 | |
| *** annabelleB has joined #kata-dev | 14:50 | |
| kata-dev-irc-bot | <anne> someone made a great joke that our tier 1 support is *very* good ;) | 14:54 |
| *** eernst has joined #kata-dev | 14:55 | |
| kata-dev-irc-bot | <anne> @verytired1 would be the man to dive into it AFAIK. I know he's pretty busy for the next little while, but we should get it resolved | 14:56 |
| gwhaley | @anne - is there somewhere we/you can open a ticket/Issue at least so it is in the queue and we don't forget? | 14:59 |
| kata-dev-irc-bot | <anne> gwhaley: I'm wondering if a github issue on community would be okay? It's not really a proper use of issues, but it would keep it in front of us until it gets resolved? | 15:02 |
| gwhaley | @anne, if there is no relevant internal OSF place, then yes, github. In community repo is probably the right place, as that is where we mention slack and irc: https://github.com/kata-containers/community/blob/ec7cac95744cc4df995591774ae6df30af2988ee/CONTRIBUTING.md#contact | 15:05 |
| kata-dev-irc-bot | <anne> yeah, our internal stuff won't bother the right people :slightly_smiling_face: I'll put an issue in now | 15:06 |
| kata-dev-irc-bot | <anne> actually, for info collecting...so sorry jess..about to do it again! | 15:07 |
| kata-dev-irc-bot | <anne> so all channels | 15:08 |
| *** sameo has joined #kata-dev | 15:08 | |
| kata-dev-irc-bot | <verytired1> Yeah I looked into it before and I think I tracked it down to the slack client library that the irc bridge integrates with, but I didn't find the fix previously. An issue would be good and you can attach me to it | 15:09 |
| kata-dev-irc-bot | <anne> perfecto! thanks JB! | 15:10 |
| *** jodh has quit IRC | 15:17 | |
| *** annabelleB has quit IRC | 15:25 | |
| *** annabelleB has joined #kata-dev | 15:28 | |
| *** annabelleB has quit IRC | 15:31 | |
| *** mcastelino has joined #kata-dev | 15:31 | |
| *** mcastelino_ has joined #kata-dev | 15:34 | |
| *** mcastelino has quit IRC | 15:34 | |
| *** annabelleB has joined #kata-dev | 15:35 | |
| *** annabelleB has quit IRC | 15:40 | |
| *** eernst has quit IRC | 15:41 | |
| *** eernst has joined #kata-dev | 15:42 | |
| *** annabelleB has joined #kata-dev | 15:47 | |
| kata-dev-irc-bot | <mnaser> ^ feel free to loop me in if you need help checking things out | 15:50 |
| *** kata-dev-irc-bot has quit IRC | 15:57 | |
| *** kata-dev-irc-bot has joined #kata-dev | 15:59 | |
| *** kata-dev-irc-bot has quit IRC | 16:00 | |
| *** kata-irc-bot has joined #kata-dev | 16:01 | |
| jbryce | testing kata from irc | 16:02 |
| kata-irc-bot | <verytired1> so it didn't do it in the working-committee channel but it still does it here... | 16:03 |
| *** jugs has quit IRC | 16:08 | |
| *** jugs has joined #kata-dev | 16:09 | |
| *** zerocoolback has quit IRC | 16:16 | |
| kata-irc-bot | <eric.ernst> @jose.carlos.venegas.m ping? | 16:23 |
| kata-irc-bot | <eric.ernst> Is there a reason we are installing CRI-containerd from sources for the CI? | 16:23 |
| kata-irc-bot | <eric.ernst> @salvador.fuentes ^ | 16:23 |
| kata-irc-bot | <eric.ernst> I've been reviewing our documentation on kubernetes, and just merged Sebastien's PR for clarifying some documentation. | 16:25 |
| kata-irc-bot | <eric.ernst> Basically, the hard part is getting k8s up with either containerd or crio. The installation of Kata is quite simple. I was thinking it would be easier to have self-contained 'scripts' as a reference, and generally it'd be a lot easier if we didn't have developers build from sources for these third party components... | 16:27 |
| kata-irc-bot | <eric.ernst> the existing CI ones are a bit complicated as a guide since they aren't self-contained. | 16:27 |
| kata-irc-bot | <eric.ernst> I don't think we can just point users to containerd/crio install directions if we are keeping to a particular version. | 16:33 |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst when I send the PR to test cri-containerd I got some comments about not looking good just get the tarball and put the content in root | 16:33 |
| kata-irc-bot | <eric.ernst> they have packages, don't they? | 16:33 |
| kata-irc-bot | <jose.carlos.venegas.m> but we can move to install the static binaries | 16:33 |
| kata-irc-bot | <eric.ernst> dnf install? apt-get install? | 16:33 |
| kata-irc-bot | <jose.carlos.venegas.m> I did not seen packages but a tarball | 16:34 |
| kata-irc-bot | <jose.carlos.venegas.m> https://github.com/containerd/cri/blob/v1.0.2/docs/installation.md | 16:35 |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst I can send a PR to install in that way, sounds good | 16:36 |
| kata-irc-bot | <jose.carlos.venegas.m> ? | 16:36 |
| kata-irc-bot | <eric.ernst> let me check... | 16:37 |
| kata-irc-bot | <eric.ernst> ah. | 16:39 |
| kata-irc-bot | <eric.ernst> you're right: VERSION="1.1.0" curl -OL \ https://storage.googleapis.com/cri-containerd-release/cri-containerd-${VERSION}.linux-amd64.tar.gz sudo tar -C / -xzf cri-containerd-${VERSION}.linux-amd64.tar.gz | 16:39 |
| kata-irc-bot | <jose.carlos.venegas.m> yes, those lines I used in the past | 16:40 |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst btw about static binaries | 16:41 |
| kata-irc-bot | <jose.carlos.venegas.m> https://gist.github.com/jcvenegas/f88546859b40f99f7ba6ff23ca55cdde | 16:41 |
| kata-irc-bot | <raravena80> I believe you can also get the binaries from here: https://github.com/containerd/containerd/releases/tag/v1.1.0 ^^ | 16:41 |
| kata-irc-bot | Action: eric.ernst wonders why they aren't providing packages. | 16:42 |
| kata-irc-bot | <eric.ernst> or if they assume you'd tarball binaries, drop into a container, and then deploy an install onto a running k8s cluster. | 16:42 |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst @raravena80 yep containerd releases should include the CRI pluging as well | 16:44 |
| kata-irc-bot | <jose.carlos.venegas.m> @eric.ernst I personally sometimes find useful have a tarball with static binaries, for a quick test in a VM is simple | 16:45 |
| kata-irc-bot | <jose.carlos.venegas.m> but yeah probably packages are prefered in some cases | 16:45 |
| kata-irc-bot | <eric.ernst> I tend to agree, since it is coming from the release page of that particular project, too | 16:49 |
| kata-irc-bot | <eric.ernst> we should have a self contained script for installing crio on top of k8s | 16:50 |
| kata-irc-bot | <eric.ernst> and for installing containerd on top of k8s | 16:50 |
| kata-irc-bot | <eric.ernst> then another set of selft contained scripts for configuration | 16:50 |
| kata-irc-bot | <eric.ernst> today its a mix of the two, and its a bit convulated as a guide reference. | 16:50 |
| *** eernst has quit IRC | 16:54 | |
| *** david-lyle has quit IRC | 16:59 | |
| *** gwhaley has quit IRC | 17:01 | |
| *** LinuxMe has quit IRC | 17:13 | |
| *** LinuxMe has joined #kata-dev | 17:26 | |
| *** lamego1 has joined #kata-dev | 17:36 | |
| *** lamego has quit IRC | 17:38 | |
| *** eernst has joined #kata-dev | 17:40 | |
| *** dklyle has joined #kata-dev | 17:40 | |
| *** manjeets has joined #kata-dev | 18:10 | |
| *** cdent has joined #kata-dev | 18:21 | |
| *** fiddletwix has quit IRC | 18:39 | |
| *** jdandrea has left #kata-dev | 18:49 | |
| *** davidgiluk has quit IRC | 19:00 | |
| *** cdent has quit IRC | 19:18 | |
| *** LinuxMe has quit IRC | 19:33 | |
| *** LinuxMe has joined #kata-dev | 19:37 | |
| *** annabelleB has quit IRC | 19:56 | |
| *** annabelleB has joined #kata-dev | 20:10 | |
| kata-irc-bot | <eric.ernst> hey @jonolson @tallclair | 20:26 |
| kata-irc-bot | <eric.ernst> I was thinking more on kata-deploy, and while I think it is fun, the hard part isn't kata, its getting containerd/crio | 20:26 |
| kata-irc-bot | <eric.ernst> I think you can recall this well @jonolson | 20:27 |
| kata-irc-bot | <eric.ernst> So... I was thinking about making the same for crio and or containerd | 20:28 |
| kata-irc-bot | <eric.ernst> I need to dig into how kubeadm would like having the CRI socket changed underneath it. | 20:28 |
| kata-irc-bot | <eric.ernst> but, was curious about your input | 20:28 |
| *** lamego has joined #kata-dev | 20:33 | |
| kata-irc-bot | <tallclair> Yeah, the ability to dynamically add runtimes is interesting, especially if we agree on the RuntimeClass route (which is the way the winds are blowing) | 20:33 |
| kata-irc-bot | <eric.ernst> in this case, its CRI runtime. | 20:33 |
| kata-irc-bot | <eric.ernst> I'm not sure if kubeadm cares after --init | 20:33 |
| kata-irc-bot | <tallclair> I think to really support that usecase natively in a portable way, we would need to have a multiplexed CRI in the kubelet layer. | 20:33 |
| kata-irc-bot | <eric.ernst> I would just restart kubelet after updatings its config | 20:34 |
| kata-irc-bot | <tallclair> Then you could have the RuntimeClass definition include the socket to talk to. | 20:34 |
| kata-irc-bot | <tallclair> er, sorry, interleaving 2 conversations | 20:34 |
| kata-irc-bot | <tallclair> I'm afraid I don't really no anything about kubeadm. I think it would be worth raising this discussion in sig-node though | 20:36 |
| kata-irc-bot | <tallclair> I'm sure folks would love a demo too, if you're able to make it one of these weeks | 20:36 |
| *** lamego has quit IRC | 20:36 | |
| *** lamego1 has quit IRC | 20:36 | |
| kata-irc-bot | <eric.ernst> Yeah. I'm working on it now to see "what happens" | 20:37 |
| kata-irc-bot | <raravena80> can you use the `--cri-socket string` option? | 20:37 |
| kata-irc-bot | <eric.ernst> essentially just: | 20:37 |
| kata-irc-bot | <eric.ernst> 1) get dockershim k8s running (easy) | 20:37 |
| kata-irc-bot | <eric.ernst> 2) kubeadm apply -f <i want crio>.yaml, which sould reset kubelet | 20:38 |
| kata-irc-bot | <eric.ernst> to use crio now | 20:38 |
| kata-irc-bot | <eric.ernst> 3) kata deploy (this part already works like a charm) | 20:38 |
| kata-irc-bot | <raravena80> seems like `kubeadm init` and `kubeadm join` support the --cri-socket option https://kubernetes.io/docs/reference/setup-tools/kubeadm/kubeadm-join | 20:38 |
| kata-irc-bot | <eric.ernst> they do -- i'm talking about a daemonset to install on the fly, @raravena80 | 20:41 |
| kata-irc-bot | <eric.ernst> ie, use dockershim to install crio | 20:42 |
| kata-irc-bot | <eric.ernst> @tallclair thanks for the feedback. I'll see what falls over and put some time into sig-node | 20:42 |
| *** lamego has joined #kata-dev | 20:43 | |
| *** mcastelino_ has quit IRC | 20:44 | |
| kata-irc-bot | <raravena80> oh I see, did you mean `kubectl apply -f` to create your daemonset instead of `kubeadm apply`? I don't see that as an option in the kubeadm options | 20:46 |
| *** lamego has quit IRC | 20:54 | |
| *** annabelleB has quit IRC | 20:56 | |
| *** eernst has quit IRC | 21:02 | |
| *** eernst has joined #kata-dev | 21:03 | |
| *** LinuxMe has quit IRC | 21:04 | |
| *** eernst has quit IRC | 21:06 | |
| *** LinuxMe has joined #kata-dev | 21:06 | |
| *** lamego has joined #kata-dev | 21:07 | |
| *** eernst has joined #kata-dev | 21:35 | |
| *** eernst_ has joined #kata-dev | 21:37 | |
| *** eernst has quit IRC | 21:37 | |
| *** annabelleB has joined #kata-dev | 21:46 | |
| *** devimc has quit IRC | 21:53 | |
| *** fuentess has quit IRC | 22:03 | |
| *** annabelleB has quit IRC | 22:29 | |
| *** LinuxMe has quit IRC | 22:30 | |
| *** kata-irc-bot has quit IRC | 22:30 | |
| *** kata-irc-bot has joined #kata-dev | 22:31 | |
| jbryce | testing again: kata | 22:32 |
| *** annabelleB has joined #kata-dev | 22:40 | |
| kata-irc-bot | <sgordon> \o/ | 22:42 |
| kata-irc-bot | <kata> yayyyy | 22:48 |
| kata-irc-bot | <verytired1> we should probably test one more time now that you're back... | 22:50 |
| jbryce | testing one more time: kata | 22:50 |
| kata-irc-bot | <verytired1> woohoo | 22:51 |
| kata-irc-bot | <kata> yayyyyy | 22:51 |
| *** lamego has left #kata-dev | 22:53 | |
| kata-irc-bot | <anne> RIP @kata's favorite github issue | 22:59 |
| kata-irc-bot | <verytired1> well...i don't know how to fix it for the general channel, so it may live on there | 23:01 |
| *** LinuxMe has joined #kata-dev | 23:02 | |
| *** LinuxMe has quit IRC | 23:07 | |
| kata-irc-bot | <sgordon> the github issue is dead, long live the github issue? | 23:33 |
| *** LinuxMe has joined #kata-dev | 23:34 | |
| kata-irc-bot | <kata> I'm so sad now hahaha | 23:36 |
| *** LinuxMe has quit IRC | 23:39 | |
| *** annabelleB has quit IRC | 23:53 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!