mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andrew Schwartzmeyer <and...@schwartzmeyer.com>
Subject Re: Review Request 67931: Support Image Manifest Version 2 Schema 2.
Date Mon, 30 Jul 2018 23:39:40 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/67931/#review206622
-----------------------------------------------------------




3rdparty/CMakeLists.txt
Lines 721-736 (original), 721-736 (patched)
<https://reviews.apache.org/r/67931/#comment289655>

    Note for others: this was forced by an upstream build change in curl.



include/mesos/docker/spec.hpp
Lines 40-42 (patched)
<https://reviews.apache.org/r/67931/#comment289656>

    +1



include/mesos/uri/fetcher.hpp
Lines 101 (patched)
<https://reviews.apache.org/r/67931/#comment289657>

    I am not convinced we need to be passing this as a shared_ptr, wouldn't const-ref be just
fine?



src/Makefile.am
Lines 692-697 (original), 695-703 (patched)
<https://reviews.apache.org/r/67931/#comment289658>

    Whitespace I'll fix up when committing...



src/slave/containerizer/mesos/provisioner/docker/registry_puller.cpp
Lines 467-505 (patched)
<https://reviews.apache.org/r/67931/#comment289661>

    I believe you could accomplish what you're doing here with `goto fail/succeed` by refactoring
this into a function (or even an in-place lambda) where the `goto`s would turn into `return`s.
But I'm not sure if it'd be any cleaner. I have nothing against `goto` when used right, and
this seems reasonable.



src/uri/fetcher.hpp
Line 46 (original), 43 (patched)
<https://reviews.apache.org/r/67931/#comment289654>

    This can go.



src/uri/fetchers/docker.cpp
Lines 791-818 (patched)
<https://reviews.apache.org/r/67931/#comment289665>

    Hey that worked out nicely.



src/uri/fetchers/docker.cpp
Lines 1016-1017 (patched)
<https://reviews.apache.org/r/67931/#comment289666>

    Are we just skipping a failed blob here and trying to process the rest? Does that make
sense to do?


- Andrew Schwartzmeyer


On July 30, 2018, 3:40 p.m., Liangyu Zhao wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/67931/
> -----------------------------------------------------------
> 
> (Updated July 30, 2018, 3:40 p.m.)
> 
> 
> Review request for mesos, Akash Gupta, Andrew Schwartzmeyer, and Joseph Wu.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Support parsing schema 2 and fetching blob with multiple URLs as
> specified in schema 2. Update `curl` version to 7.61.0 because of bug
> encountered in version 7.57.0.
> 
> 
> Diffs
> -----
> 
>   3rdparty/CMakeLists.txt b58996d2ed7521e42d6298d174cc8c494b84eb8f 
>   3rdparty/cmake/Versions.cmake 0a897d808cd9e05ac0d1a4e1ca61b8f3538f0c4a 
>   include/mesos/docker/spec.hpp 2879414dc42ffe633ac74b51e1bb116698c41162 
>   include/mesos/docker/v2_2.hpp PRE-CREATION 
>   include/mesos/docker/v2_2.proto PRE-CREATION 
>   include/mesos/uri/fetcher.hpp 760d6b33234d8efdc533c0c6f05e83a5c7d1f56b 
>   src/CMakeLists.txt 10b0946d6f49c7e9c201bad6f9f1b41cc8460fe5 
>   src/Makefile.am 228e168c22f3fd0367f029c506171c6979b31c07 
>   src/docker/spec.cpp 96fbf1f9cf1c2c4b2383607a97990f3a9156e6d9 
>   src/slave/containerizer/mesos/containerizer.cpp 98129d006cda9b65804b518619b6addc8990410a

>   src/slave/containerizer/mesos/provisioner/docker/registry_puller.cpp a5683e3fe15dd35596122fcc0c580ae9d3adf7f2

>   src/uri/fetcher.hpp fbf64c6767dea3aa0798f68db8322ce47cd8ad36 
>   src/uri/fetcher.cpp 489c7ce0198ee6803dcc8eb9710b292fa743a0e8 
>   src/uri/fetchers/docker.hpp cdbab9a5684a68a729be12bc06d331acca137da5 
>   src/uri/fetchers/docker.cpp 55ca118660872a933a2dc186723bec6a39ee80f7 
> 
> 
> Diff: https://reviews.apache.org/r/67931/diff/3/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Liangyu Zhao
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message