This repository has been archived by the owner on Jan 22, 2020. It is now read-only.
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Updated {linuxkit/linuxkit 30b8670a98bab7a5bd3d343ad2613e3ed783e8e5:ci/circleci[pending] url=https://circleci.com/gh/linuxkit/linuxkit/1324?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link descr=Your tests are queued behind your running builds} {linuxkit/linuxkit 3168[1d1c8fcee99d58781b90bc646398046d22e286ef] master kkroo open "Use containerd restart manager to monitor services" [|0 kkroo: "<!--\r\nPlease make sure you've read and understood our contributing guidelines;\r\n"; 416053776 GordonTheTurtle: "<!-- AUTOMATED:POULE:DCO-EXPLANATION -->\nPlease sign your commits following thes"; 416566165 justincormack: "Hi, we can't merge PRs unless they are signed-off-by as per instructions above.\r"; 417360415 ijc: "containerd/containerd#2318 shows it using `container.Upd"; 442205572 kkroo: "Addressing comments regarding race conditions. Regarding the use of this interfa"; 442222157 medic15: "This needs to be configurable for each service. Not all those who wander are lo"; 442319928 kkroo: "> This needs to be configurable for each service. Not all those who wander are l"; 442400886 ijc: "@medic15 and @justincormack are correct, this needs to be configurable per servi"; 442401015 justincormack: "At present a few of the services we run may terminate (eg `rngd` if no rng is su"; 442472675 medic15: "My use case involves populating a Redis in-memory database using data from sever"; 442473416 justincormack: "@medic15 yes that seems a reasonable use case where it should be configurable."; 442621437 kkroo: "@medic15 @justincormack I still don't understand the use case and why such a one"; 442622363 justincormack: "Services are not started until all `onboot` containers have exited."|]} {linuxkit/linuxkit 2648[bfaa9e7fc599a8b9edba4429ae513f0736411fce] master w9n open "linuxkit bake" [|0 w9n: "When writing a moby yml I do not care about the container version and hashes (on"; 340192010 GordonTheTurtle: "<!-- AUTOMATED:POULE:DCO-EXPLANATION -->\nPlease sign your commits following thes"; 340204660 rn: "I don't understand the explanation for this PR (maybe I'm missing some context)."; 340224169 w9n: "added some context to the description. "; 340241991 deitch: "> When writing a moby yml I do not care about the container version and hashes ("; 340268573 w9n: "i basically thought of a not buildable config\r\n```\r\nkernel:\r\n image: linuxkit/k"; 340269095 deitch: "So, is this basically a way of saying, \"for some packages, I want the latest ava"; 340269511 w9n: "yep, but still create the immutable ymls so its reproducable what has been build"; 340269588 deitch: "So rather than the usual moby `.yml`, it kind of is a moby \"template\" `.yml`, th"; 340269816 w9n: "exactly. It would enable other possibilities as well as described above but ther"; 340269924 deitch: "OK, I get it now; want to think it over somewhat. @rn had commented earlier; let"; 340270518 rn: "This would be better done in the `moby` tool as it handles the build YAML files."; 340271020 deitch: "Ah yes, the `.mobytags` and overrides proposals.\r\n\r\n@w9n does that address the q"; 340273985 w9n: "yes, it pretty much goes is in the right direction. I will read and think about "; 341074173 ijc: "I'm still catching up on my post PTO backlog(s) so I haven't fully grokked this "; 341552432 w9n: "i didnt choose ```foo:latest``` because it can conflict with other contexts. But"; 352289887 w9n: "Since moby became a dependency this works now depending on moby/tool#194 `linuxk"; 355176335 w9n: "@deitch @rn @ijc Please recheck, adopting this could quit manually updating hash"; 355233916 deitch: "I remain torn on this. I think I still like the idea of having the ability to ta"; 355287913 w9n: "> If I understand the implementation correctly, pkg collect would then look in l"; 355293548 ijc: "> if a container has `<latest>` tag\r\n\r\nAre the angle-brackets (`<>`) literally p"; 355298378 w9n: "> Are the angle-brackets (<>) literally part of the tag here or just \"syntax\"?\r\n"; 355304180 deitch: "> Yes, you would need atleast the linuxkit/linuxkit repo and a setup pkgroot par"; 355307237 w9n: "> Are there 2 distinct behaviours?\r\n\r\nThere is `linuxkit pkg collect --pkgroot` "; 355310745 deitch: "> there is linuxkit build -collect which pre-processes and directly outputs the "; 355311332 rn: "`linuxkit pkg collect` seems a misnomer as it is not an operation on a package. "; 355313071 deitch: "> linuxkit pkg collect seems a misnomer as it is not an operation on a package\r\n"; 355314861 rn: "> Yes, but I could see how you might want a convenience of a single step.\r\n\r\nYou"; 355316644 deitch: "> I really do not want linuxkit build to do too much magic. It should just buil"; 355316963 deitch: "On second thought, I guess I can see it. `lkt` already contains all of the code "; 355318030 rn: "sure, this can be done with a shell script. `grep` image patterns and `linuxkit "; 355321845 ijc: "> You get a \"single\" step with linuxkit tranform | linuxkit build -. I really do"; 355347992 w9n: "dropping `linuxkit build -collect` and `--pkgroot <org>:<path>` sounds good to m"; 355446448 rn: "`linuxkit transform -o outputfile <filename> | linuxkit build -` does not seem t"; 355491579 deitch: "> I think generate is my favourite\r\n\r\nI like it too. My only concern is that, t"; 355491607 deitch: "Or perhaps `lkt template`?"; 355492212 w9n: "generate sounds good but dont mind changing.\r\n\r\nChanged to `linuxkit generate --"; 355530363 rn: "@w9n thanks for the quick change, I have a few comments on the implementation it"; 355535990 ijc: "> then a new `linuxkit build` command \r\n\r\nITYM `linuxkit generate`? Not mad enam"; 355538317 ijc: "> My only concern is whether or not there is some existing YAML templating synta"; 355540809 deitch: "> Couldn't find anything\r\n \r\nYeah, I had that issue a while ago. Ended up writi"; 355540982 deitch: "> ITYM linuxkit generate? Not mad enamoured on the name, but no better ideas (ba"; 355567882 w9n: "> you shouldn't need to add anything to pkglib. This is a new top level command "; 355569715 w9n: "bake & build would kind of short and precisely describe whats going on and actua"; 355570963 ijc: "> the git methods are not public \r\n\r\nYou appear to be using them to reimplement "; 355572091 w9n: "Yes `New` in pkglib.Pkg would be awesome - but i think this needs some more work"; 355574550 ijc: "Where/why does this code need to deal with old hashes?\r\n\r\nGenerating the hash co"; 355577199 w9n: "if there is\r\n```\r\ntags:\r\n - name: <2.1>\r\n hash: bb0c6ae2f12a1b55df24ebce2067"; 355577929 rn: "Can we address one issue at a time. templating of the YAML file is one thing and"; 355579842 ijc: "You can already get the package hash from a git tag today already, it's currentl"; 355656944 w9n: "I made `pkgInfo` public and moved it into the config namespace I put GlobalConfi"; 355967540 ijc: "> I made pkgInfo public \r\n\r\nI don't like this, sorry. This struct should remain "; 356205605 w9n: "@ijc I did what you suggested and exposed a `SetCommit` `GetHash` and `Image` fu"; 356250551 ijc: "You shouldn't need `SetCommit` since the only caller outside `pkglib` passes a h"; 356256042 w9n: "done"; 356556345 w9n: "so i rethought this and the whole templating problem might just not fit into lin"; 356567368 ijc: "I don't think there are objections to templating in general, my specific objecti"; 356571088 w9n: "If git tags or parameter like `foo=abcdef` are used than this would require e.g."; 356572786 ijc: "We intend to release linuxkit as a whole, with a single consistent set of packag"; 356576666 rn: "I agree with @ijc that we do want some templating. We had several discussions in"; 356579465 justincormack: "I don't understand what the output of `linuxkit bake example/linuxkit_template.y"; 356581808 ijc: "It's a new file which you can pass to `linuxkit build` (after archiving it for r"; 356582385 justincormack: "Does it output to stdout or what? This needs to be documented properly."; 356582717 w9n: "Okay, so I could live with this somewhere:\r\n```\r\ntags:\r\n - name: v1.0\r\n "; 356585190 rn: "@w9n please read my comment linuxkit/linuxkit#2648 (comment)"; 356587899 w9n: "> Do you have an example for a baked YAML file. It would be useful if it would c"; 356589814 w9n: "i dont really trust the test time, something must be wrong...\r\n```\r\n[+ 24m 58s] "; 356647166 w9n: "@ijc \r\n```\r\nlinuxkit bake -D foo=abcdef foo.yml\r\n```\r\nThis can actually be done "; 356922250 w9n: "updated the last comment, otherwise i guess i should revert setting global confi"; 357164648 w9n: "Sorry for all these changes\r\n - dropped putting GlobalConfig in its own namespac"; 357425907 w9n: "```\r\n[+ 24m 27s] [PASS ] linuxkit.build.examples.bake 32.57s\r\n```\r\n\r\nfixed te"|]} Removed {linuxkit/linuxkit 0748457296dda072557f13358346261a8c1948b8} {linuxkit/linuxkit 13bfb5c238b453bdf578b7c11c3055ef4074505d} {linuxkit/linuxkit 1616b18fb07ba39d04ddd616593769b2d0384fef} {linuxkit/linuxkit 1721ea26718943ca036f036fad894b0f05032e77} {linuxkit/linuxkit 38b563a11443db7e81211a9e7fd8717c941178d0} {linuxkit/linuxkit 4b66ef2e2a65a6c8bb2c03fd95fc09527b3f5c5f} {linuxkit/linuxkit 4e634e3b2975c36450be70d550080b76f71a9c06} {linuxkit/linuxkit 618c8675e364d936ee7561f411365e2df217806c} {linuxkit/linuxkit 89a7fbebcc83c2b1b2a4ffce29a8f65d11fd540b} {linuxkit/linuxkit a36e1149a9b6daffeb8778e87b0b5f490752858a} {linuxkit/linuxkit bbde423c6f9190c7b32ff2974422de0daa59cbd1} {linuxkit/linuxkit d45f905adb8623c6c7ca06b3b4c5e3c66f4882e7} {linuxkit/linuxkit eef5d1a9051a7db6d08e70d4d3c8202d977b36ff} {linuxkit/linuxkit ef9302bc01f45e937eef52f0a68d3b9f073e0747}
- Loading branch information