Tuesday, 2019-04-30

« Monday, 2019-04-29 Index Wednesday, 2019-05-01 »
*** igordc has quit IRC00:28
*** dklyle has joined #kata-dev01:07
*** dklyle has quit IRC01:12
*** jiapei has quit IRC01:18
*** changcheng has joined #kata-dev01:24
kata-irc-bot2<xu> Move the arch committee meeting this week to Thursday? align with PTG04:03
kata-irc-bot2<harshal.patil> @archana.m.shinde we are evaluating kata for some of the use cases for isolated/encrypted memory solution like IBM SMF, https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.ieag200/intro.htm. (Something like Intel MKTME). Unfortunately, I don't have the liberty to talk about it at this point. But I came across the lack of name resolution bug during such evaluation. If you anyone of you are attending the04:21
kata-irc-bot2upcoming kubecon we can briefly discuss it.  @archana.m.shinde Let me know if I should pursue the issue created by @bergwolf to address this.04:21
kata-irc-bot2<bergwolf> > We dont want the agent to be able to access the network, reduce the attack surface there @archana.m.shinde At the point of creating containers, the agent already has network capability as guest interface and route are already set up. Name resolution is just a missing piece for convenience.05:47
*** sgarzare has joined #kata-dev06:25
*** tmhoang has joined #kata-dev06:47
*** jodh has joined #kata-dev07:34
*** pcaruana has joined #kata-dev07:40
*** davidgiluk has joined #kata-dev07:56
*** gwhaley has joined #kata-dev08:02
*** changcheng has quit IRC08:05
kata-irc-bot2<graham.whaley> Hey Denver PTG folks - are you having any hack-fest sessions whilst there then? Dont' suppose you might consider fixing up all the kata k8s docs to be in one place, upto date, and complete would you? :slightly_smiling_face: I was looking for the 'how to do per-pod kernel configs', when I found we still have the old 'trusted' annotations in at least one (of three iirc) k8s docs :(08:52
kata-irc-bot2https://github.com/kata-containers/documentation/blob/master/how-to/how-to-use-k8s-with-cri-containerd-and-kata.md#create-an-untrusted-pod-using-kata-containers08:52
*** SteveCapper has joined #kata-dev08:56
*** gwhaley has quit IRC10:59
kata-irc-bot2<graham.whaley> hi (/me picks semi-random people...) @salvador.fuentes @julio.montes - do you know if there are any known circumstances where 'docker diff' does not work as expected with kata? https://github.com/kata-containers/tests/blob/master/integration/docker/diff_test.go#L35https://github.com/kata-containers/tests/blob/master/integration/docker/diff_test.go#L3512:18
kata-irc-bot2<graham.whaley> I have that test failing on one of my test machines - it is running devicemapper, and I would suspect that to be a key suspect. Maybe a sync back of a block device does not ripple through to whereever docker does the diff (on the host side I guess)?12:19
kata-irc-bot2<graham.whaley> @archana.m.shinde ^^^ as well12:19
kata-irc-bot2Action: graham.whaley was trying to test the kernel frags, but hit this at his desk instead :-(12:19
*** fuentess has joined #kata-dev12:22
kata-irc-bot2<salvador.fuentes> hi @graham.whaley not sure, haven't tested docker with devicemapper for a while, but yeah, that might be the reason12:26
*** devimc has joined #kata-dev12:30
kata-irc-bot2<graham.whaley> fuentess: ah, no devicemapper based docker testing CI job ?12:35
kata-irc-bot2<graham.whaley> I know, devicemapper is going the way of the dinosaurs....12:35
kata-irc-bot2<salvador.fuentes> yeah, no ci job for docker + devmapper, only for cri-o12:35
kata-irc-bot2<graham.whaley> I'll flip back to my overlay2 test machine then. np! ;)12:36
kata-irc-bot2<salvador.fuentes> cool12:36
*** altlogbot_1 has quit IRC12:44
*** altlogbot_2 has joined #kata-dev12:45
kata-irc-bot2<graham.whaley> So, we hear it is snowing in Denver? !!! @eric.ernst13:06
kata-irc-bot2<xu> Yep, very strong:snowflake:13:06
*** pcaruana has quit IRC14:15
*** jodh has quit IRC14:16
* davidgiluk heard it's a tradition for OpenStack to hold meetings in places snowed under14:17
kata-irc-bot2<graham.whaley> heh heh - so they can continue to party...14:20
kata-irc-bot2<eric.ernst> :slightly_smiling_face:14:22
kata-irc-bot2<eric.ernst> I got sunburn the morning before (and was hot in the sun)14:23
kata-irc-bot2<eric.ernst> impressive weather change.14:23
kashyapdavidgiluk: Was there in Dublin :D14:23
kashyapdavidgiluk: The "cabin fever" was real14:23
kata-irc-bot2<eric.ernst> :slightly_smiling_face:14:23
kashyapdavidgiluk: Tomorrow it's going to get better, apparently.  Let's see how it'll be when I arrive tomm14:24
*** dims has quit IRC14:31
*** dims has joined #kata-dev14:44
*** pcaruana has joined #kata-dev14:46
clarkbkashyap: it is nothing like dublin here. The snow only stuck to the grass and bushes and car tops14:58
*** devimc has quit IRC15:02
kata-irc-bot2<graham.whaley> @eric.ernst @xu @bergwolf @sebastien.boeuf - presume you are heading to the arch call today are you? just a couple of us waiting in the call right now15:02
*** devimc has joined #kata-dev15:03
*** sameo has joined #kata-dev15:05
*** dims has quit IRC15:35
*** tmhoang has quit IRC15:42
*** dims has joined #kata-dev15:42
*** pcaruana has quit IRC15:55
*** dklyle has joined #kata-dev15:57
*** dklyle has quit IRC16:09
kata-irc-bot2<gmmaharaj> @graham.whaley @salvador.fuentes @james.o.hunt it seems the recent merge to 1.6 stable branch while done to fix our last release apparently does not conform to our commit msg format https://github.com/kata-containers/runtime/commits/stable-1.616:33
kata-irc-bot2<gmmaharaj> we wil need to revert and force-push `stable-1.6` to 2cbbadb93b2a41103450992ff5e8fc2775ca8edb16:34
kata-irc-bot2<gmmaharaj> this for the runtime repositoru16:34
kata-irc-bot2<gmmaharaj> any chance the push to head option can be disabled for this stable branch temporarily and push this back16:35
kata-irc-bot2<graham.whaley> also /cc @jose.carlos.venegas.m :slightly_smiling_face:16:35
kata-irc-bot2<gmmaharaj> following which i can push another patch with the same change and the right format16:35
kata-irc-bot2<gmmaharaj> +1. thanks @graham.whaley16:35
kata-irc-bot2<gmmaharaj> do we need someone to approve this? i assume @graham.whaley @jose.carlos.venegas.m and one other person approving this is good enough. the ask is fairly trivial. just replacing the commit with one that has the right commit msg16:38
kata-irc-bot2<jose.carlos.venegas.m> @gmmaharaj I dont follow it , push force?16:38
kata-irc-bot2<gmmaharaj> @jose.carlos.venegas.m we need to reset the `stable-1.6` branch to sha 2cbbadb93b2a41103450992ff5e8fc2775ca8edb16:38
kata-irc-bot2<gmmaharaj> then i will push the same kernel revert patch but with the right commit msg string. so that all future patches to `stable-1.6` won't fail with an error that something in the tree fails to have the right msg.16:39
kata-irc-bot2<gmmaharaj> check this PR for what we are seeing. https://github.com/kata-containers/runtime/pull/160716:39
kata-irc-bot2<graham.whaley> gmmaha: a sideways thought - how comes the CI picked up that commit - it should only check the format of the new commits - maybe it was because the push (PR) had not been rebased to the HEAD of the stable branch, so it examined the fast forward as well - iyswim. just a thought16:39
kata-irc-bot2<gmmaharaj> i thought the branch i pushed was rebased to the latest stable-1.66.16:40
kata-irc-bot2<gmmaharaj> that is how i cherry-picked the patch. should have had all the chnages16:40
kata-irc-bot2<jose.carlos.venegas.m> I see, I am fine with it but I think the real issue is our script to check commit message16:41
kata-irc-bot2<jose.carlos.venegas.m> the range is HEAD and branch master16:41
kata-irc-bot2<gmmaharaj> aaah..16:41
kata-irc-bot2<jose.carlos.venegas.m> it should be HEAD and ${target_branch}16:41
kata-irc-bot2<gmmaharaj> then @jose.carlos.venegas.m it would be prudent to fix the script and not reset head..16:41
kata-irc-bot2<gmmaharaj> i HATE RE-TOUCHING HISTORY OF RELEASE & PRODUCTION BRANCHES16:42
kata-irc-bot2<jose.carlos.venegas.m> I feel that that should be the fix yep, @james.o.huntany suggestion16:42
kata-irc-bot2<salvador.fuentes> yeah, I think in travis we are not using that script for stable branches... let me check16:43
kata-irc-bot2<jose.carlos.venegas.m> @gmmaharaj agree, I feel better to not touch the history16:43
kata-irc-bot2<gmmaharaj> thanks @salvador.fuentes16:43
kata-irc-bot2<salvador.fuentes> we have an old issue about this16:45
kata-irc-bot2<salvador.fuentes> https://github.com/kata-containers/tests/issues/63216:45
kata-irc-bot2<salvador.fuentes> not sure why I changed the title to say only 'on travis'16:45
kata-irc-bot2<gmmaharaj> @salvador.fuentes does it have to be `origin/stable-1.1` instead of just `stable-1.1` in that issue?16:48
kata-irc-bot2<graham.whaley> fixing the script gets :100: from me :slightly_smiling_face:16:48
kata-irc-bot2<salvador.fuentes> @gmmaharaj it seems so16:49
kata-irc-bot2<gmmaharaj> yeah. not sure if you at all are checking out that branch when you are cloning the repository. that will make a big difference.16:50
kata-irc-bot2<salvador.fuentes> I dont remember how we solved on travis, or did we run a skip?16:50
*** dklyle has joined #kata-dev16:52
kata-irc-bot2<salvador.fuentes> yeah, we should be doing that, checkout to the correct branch at the beginning16:54
kata-irc-bot2<salvador.fuentes> but for some reason, checkcommits uses master as reference...16:54
kata-irc-bot2<gmmaharaj> hrm..16:57
kata-irc-bot2<gmmaharaj> @salvador.fuentes can i assume you are working out your magic on this now?16:57
kata-irc-bot2<gmmaharaj> i don't mind lending a hand if you are busy.16:57
kata-irc-bot2<salvador.fuentes> yes, checking it, but I am not sure how much time it will take, seems that the fix will be easy, but want to test it...16:58
kata-irc-bot2<jose.carlos.venegas.m> I dont see any comment in the code16:59
kata-irc-bot2<gmmaharaj> @jose.carlos.venegas.m not following?17:00
kata-irc-bot2<jose.carlos.venegas.m> https://github.com/kata-containers/tests/blob/master/cmd/checkcommits/checkcommits.go17:01
kata-irc-bot2<jose.carlos.venegas.m> I dont see any comment to not use the target branch17:01
*** dklyle has quit IRC17:03
kata-irc-bot2<salvador.fuentes> I think the issue comes directly from here https://github.com/kata-containers/tests/blob/master/cmd/checkcommits/checkcommits.go17:04
kata-irc-bot2<salvador.fuentes> as we are not specifying the branch name when we run the `checkcommits` command17:04
*** sgarzare has quit IRC17:05
kata-irc-bot2<gmmaharaj> aaah..17:05
*** sameo has quit IRC17:06
kata-irc-bot2<jose.carlos.venegas.m> @salvador.fuentes what is the name of the branch varaible in jenkins?17:12
kata-irc-bot2<salvador.fuentes> it is `ghprbPullId`17:13
kata-irc-bot2<jose.carlos.venegas.m> do we export the commit id of the HEAD of the PR ?17:15
kata-irc-bot2<salvador.fuentes> yeah, but `checkcommits` is only configured for travis and semaphore, but not jenkins, so it is taking default values, which is master17:20
kata-irc-bot2<jose.carlos.venegas.m> and finally do we have the name of the branch where the PR comes from?17:20
kata-irc-bot2<jose.carlos.venegas.m> @salvador.fuentes yeah I am adding that case to checkcommits17:20
kata-irc-bot2<salvador.fuentes> ahh, something like this: https://paste.fedoraproject.org/paste/p5~SSzgwPv8J7dOUQKqyLQ ?17:20
kata-irc-bot2<jose.carlos.venegas.m> what about if else CI env variable is set17:20
kata-irc-bot2<salvador.fuentes> yeah, `CI` will do the job17:21
kata-irc-bot2<jose.carlos.venegas.m> that looks good too17:21
kata-irc-bot2<jose.carlos.venegas.m> please send it :slightly_smiling_face:17:22
kata-irc-bot2<salvador.fuentes> ok, let me  send the pr17:22
kata-irc-bot2<salvador.fuentes> @gmmaharaj @jose.carlos.venegas.m https://github.com/kata-containers/tests/pull/151717:27
*** igordc has joined #kata-dev17:28
kata-irc-bot2<jose.carlos.venegas.m> @salvador.fuentes could you send it to stable-1.6 too ?17:28
kata-irc-bot2<salvador.fuentes> sure17:30
*** sameo has joined #kata-dev17:37
*** dklyle has joined #kata-dev17:54
*** dklyle has quit IRC18:02
*** igordc has quit IRC18:05
*** igordc has joined #kata-dev18:58
kata-irc-bot2<gmmaharaj> @salvador.fuentes seems like we need some test fixes too on this.19:01
kata-irc-bot2<gmmaharaj> https://github.com/kata-containers/tests/pull/151719:01
*** sameo has quit IRC19:03
kata-irc-bot2<gmmaharaj> @salvador.fuentes also this patch won't help when i run tests locally on the stable branch.19:05
kata-irc-bot2<gmmaharaj> is there a way we can fix that for both.. i am guessing we don't have `ghbr..` stuff locally19:05
*** pcaruana has joined #kata-dev19:10
kata-irc-bot2<graham.whaley> you don't, `ghprb...` is stuff injected by the Jenkins GitHubPullRequestBuilder plugin :slightly_smiling_face:19:10
kata-irc-bot2<graham.whaley> you could maybe fake it up if you want to test locally :slightly_smiling_face:19:13
kata-irc-bot2<gmmaharaj> right.19:13
kata-irc-bot2<gmmaharaj> but would that be something that an end user can do ?19:13
kata-irc-bot2<gmmaharaj> as long as we have to maintain stable-1.6 this will be biting us locally19:13
kata-irc-bot2<gmmaharaj> :(19:13
kata-irc-bot2<graham.whaley> if you need to see the format - if you go to the 'rebuild' on a PR on Jenkins master, I think you might see the fields it set in the env, maybe19:13
*** davidgiluk has quit IRC19:20
kata-irc-bot2<gmmaharaj> hacked it locally and was able to check.. but think how we can do this with the scripts19:21
kata-irc-bot2<gmmaharaj> this sounds like it would work well for the script.. `git for-each-ref --format='%(upstream:short)' $(git symbolic-ref -q HEAD)`19:33
kata-irc-bot2<gmmaharaj> that will tell you the first remote branch you are based off of19:33
kata-irc-bot2<gmmaharaj> @salvador.fuentes @jose.carlos.venegas.m ^^ what do you think?19:34
kata-irc-bot2<gmmaharaj> and we can pass that in the test script19:34
*** dklyle has joined #kata-dev19:34
*** david-lyle has joined #kata-dev19:40
*** dklyle has quit IRC19:41
kata-irc-bot2<gmmaharaj> @salvador.fuentes @jose.carlos.venegas.m ignore my comments. that will only work if the brnch was created  with a remote tracking branch.19:43
kata-irc-bot2<jose.carlos.venegas.m> @gmmaharaj :P sure19:44
kata-irc-bot2<jose.carlos.venegas.m> oh, I see @salvador.fuentes pr failing19:45
kata-irc-bot2<jose.carlos.venegas.m> @salvador.fuentes you will need to fix some unit tests :P19:46
kata-irc-bot2<gmmaharaj> yup.19:46
kata-irc-bot2<gmmaharaj> and we need a way for someone to be able to test this locally. unless we can live with it for now19:46
kata-irc-bot2<jose.carlos.venegas.m> I can test it locally19:50
kata-irc-bot2<jose.carlos.venegas.m> @gmmaharaj ``` Running checkcommits version  (commit ) Detected JenkinsCI - github pull request builder Environment Defaulting commit to HEAD Found 2 commits between commit HEAD and branch stable-1.6 ERROR: Commit 105d6489a77ad04841a9637755c9938a347301ce: Failed to find subsystem in subject: "readme test" ```19:59
kata-irc-bot2<jose.carlos.venegas.m> I tested with a local commit19:59
kata-irc-bot2<salvador.fuentes> hi, sorry20:05
kata-irc-bot2<salvador.fuentes> yeah, was checking the test, but not sure why it is failing, any idea?20:06
*** pcaruana has quit IRC20:23
kata-irc-bot2<gmmaharaj> @salvador.fuentes yeah.. that won't work.. cause the place where we test, env has `CI` and we will never hit the fallbackof  HEAD and master20:30
kata-irc-bot2<gmmaharaj> can this be done via the test script instead?20:30
*** david-lyle has quit IRC20:34
kata-irc-bot2<salvador.fuentes> I think I fixed the unit tests20:34
kata-irc-bot2<gmmaharaj> ohh cool20:35
kata-irc-bot2Action: gmmaharaj goes to check the PR then.20:36
kata-irc-bot2<salvador.fuentes> I spawned a single jenkins job to check if it is fixed, I tried locally and they passed, but lets see20:36
kata-irc-bot2<gmmaharaj> holy cow.. that is one interesting change20:36
kata-irc-bot2<gmmaharaj> :slightly_smiling_face:20:37
kata-irc-bot2<salvador.fuentes> hehe, yeah...20:37
*** david-lyle has joined #kata-dev20:40
*** sameo has joined #kata-dev20:42
kata-irc-bot2<gmmaharaj> @salvador.fuentes you want to do a PR to stable-1. and 1.5 too?20:43
kata-irc-bot2<salvador.fuentes> yes, for 1.6 it is already opened, I'll open for 1.5 once I see unit test passing20:53
*** igordc has quit IRC20:54
*** david-lyle has quit IRC20:56
kata-irc-bot2<gmmaharaj> thanks.21:02
kata-irc-bot2<gmmaharaj> @salvador.fuentes did you see this? https://github.com/kata-containers/tests/pull/1518#discussion_r27986400821:04
kata-irc-bot2<gmmaharaj> i think CI is fine, but i am wondering if there is a way to screen something else that says, I am on jenkins21:04
kata-irc-bot2<salvador.fuentes> yeah, I changed to look for ghpbPullId insted of CI21:05
kata-irc-bot2<salvador.fuentes> ohh, only changed on one of the PRs, let me change on the other21:06
*** igordc has joined #kata-dev21:07
kata-irc-bot2<salvador.fuentes> @gmmaharaj opened for 1.5 also, unit tests passed, so should be fine now21:10
kata-irc-bot2<gmmaharaj> cool..21:15
kata-irc-bot2<gmmaharaj> thanks a lot @salvador.fuentes21:15
kata-irc-bot2<gmmaharaj> once that lands, i can get the 1.6 PRs out21:16
kata-irc-bot2<gmmaharaj> and sorry for piling up on the CI.21:16
kata-irc-bot2<salvador.fuentes> dont worry :slightly_smiling_face:21:16
*** devimc has quit IRC21:17
*** david-lyle has joined #kata-dev21:24
*** david-lyle has quit IRC21:34
*** fuentess has quit IRC21:49
*** david-lyle has joined #kata-dev21:53
*** igordc has quit IRC21:58
*** david-lyle has quit IRC22:10
kata-irc-bot2<gmmaharaj> @salvador.fuentes ``` ERROR: Failed to run command /usr/bin/git rev-list --no-merges --reverse stable-1.6..01a60f26ff688dd122c31266ec425862235cca8d: exit status 128 (stdout: , stderr: fatal: ambiguous argument 'stable-1.6..01a60f26ff688dd122c31266ec425862235cca8d': unknown revision or path not in the working tree. 14:46:31 Use '--' to separate paths from revisions, like this: 1 ```22:14
kata-irc-bot2<gmmaharaj> shouldn't it be origin/stable-1.22:14
kata-irc-bot2<gmmaharaj> think it should be that.  won't have the branch otherwise. `refs/remotes/origin/stable-1.6`22:17
*** david-lyle has joined #kata-dev22:21
*** david-lyle has quit IRC22:43
*** sameo has quit IRC23:23
*** david-lyle has joined #kata-dev23:27
*** david-lyle has quit IRC23:33
kata-irc-bot2<gmmaharaj> @salvador.fuentes @gabriela.cervantes.te are you abble to make sense of why things are failing w.r.t CI..23:44
kata-irc-bot2<gmmaharaj> the 3 PRs that i am looking at are https://github.com/kata-containers/runtime/pull/1588 https://github.com/kata-containers/tests/pull/1517 and https://github.com/kata-containers/tests/pull/151823:45
« Monday, 2019-04-29 Index Wednesday, 2019-05-01 »

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!