使用bazel编译工具编译时提示“An error occurred during the fetch of repository 'io_bazel_rules_docker'”，编译失败

问题描述

使用bazel build --config=monolithic //tensorflow/compiler/aot:xlacompilebazel build命令编译过程中，提示“ERROR: An error occurred during the fetch of repository 'io_bazel_rules_docker'”，检查服务器，能够连接网络，但仍旧提示如下错误信息：

ERROR: An error occurred during the fetch of repository 'io_bazel_rules_docker':
   java.io.IOException: Error downloading [https://github.com/bazelbuild/rules_docker/releases/download/v0.14.3/rules_docker-v0.14.3.tar.gz] to /home/test/.cache/bazel/_bazel_test/abd37aaac8a380ca5a3f13938322fcb2/external/io_bazel_rules_docker/rules_docker-v0.14.3.tar.gz: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target
ERROR: no such package '@io_bazel_rules_docker//repositories': java.io.IOException: Error downloading [https://github.com/bazelbuild/rules_docker/releases/download/v0.14.3/rules_docker-v0.14.3.tar.gz] to /home/test/.cache/bazel/_bazel_test/abd37aaac8a380ca5a3f13938322fcb2/external/io_bazel_rules_docker/rules_docker-v0.14.3.tar.gz: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target
INFO: Elapsed time: 24.586s
INFO: 0 processes.
FAILED: Build did NOT complete successfully (0 packages loaded)

解决方案

如果连网情况下，仍旧提示上述无法下载压缩包的问题，则请参见如下方法解决：

将如下链接中的附件下载到本地，然后上传到linux服务器任意路径，例如上传到$HOME/bazel_tools路径。

https://github.com/bazelbuild/rules_docker/releases/download/v0.14.3/rules_docker-v0.14.3.tar.gz
https://github.com/bazelbuild/bazel-skylib/releases/download/0.8.0/bazel-skylib.0.8.0.tar.gz
https://github.com/bazelbuild/rules_swift/releases/download/0.11.1/rules_swift.0.11.1.tar.gz
https://github.com/llvm-mirror/llvm/archive/7a7e03f906aada0cf4b749b51213fe5784eeff84.tar.gz

则上述文件在linux服务器绝对路径为：

/home/test/bazel_tools/rules_docker-v0.14.3.tar.gz
/home/test/bazel_tools/bazel-skylib.0.8.0.tar.gz
/home/test/bazel_tools/rules_swift.0.11.1.tar.gz
/home/test/bazel_tools/llvm-7a7e03f906aada0cf4b749b51213fe5784eeff84.tar.gz

修改bazel编译工具相关文件中的下载链接：

切换到tensorflow-1.15.0目录，使用vi WORKSPACE命令打开WORKSPACE，修改该文件中的下载链接，将下载链接修改为linux服务器绝对路径地址：

bazel_toolchains_repositories()
http_archive(
    name = "io_bazel_rules_docker",
    sha256 = "6287241e033d247e9da5ff705dd6ef526bac39ae82f3d17de1b69f8cb313f9cd",
    strip_prefix = "rules_docker-0.14.3",
    urls = ["file:///home/test/bazel_tools/rules_docker-v0.14.3.tar.gz"],
)

load(
    "@io_bazel_rules_docker//repositories:repositories.bzl",
    container_repositories = "repositories",
)
container_repositories()

load("//third_party/toolchains/preconfig/generate:workspace.bzl",
     "remote_config_workspace")
remote_config_workspace()

# Apple and Swift rules.
http_archive(
    name = "build_bazel_rules_apple",
    sha256 = "6efdde60c91724a2be7f89b0c0a64f01138a45e63ba5add2dca2645d981d23a1",
    urls = ["https://github.com/bazelbuild/rules_apple/releases/download/0.17.2/rules_apple.0.17.2.tar.gz"],
)  # https://github.com/bazelbuild/rules_apple/releases
http_archive(
    name = "build_bazel_rules_swift",
    sha256 = "96a86afcbdab215f8363e65a10cf023b752e90b23abf02272c4fc668fcb70311",
    urls = ["file:///home/test/bazel_tools/rules_swift.0.11.1.tar.gz"],
)  # https://github.com/bazelbuild/rules_swift/releases

执行wq!保存退出。

修改tensorflow-1.15.0/tensorflow路径下workspace.bzl文件中llvm对应的链接：

    # TODO(phawkins): currently, this rule uses an unofficial LLVM mirror.
    # Switch to an official source of snapshots if/when possible.
    tf_http_archive(
        name = "llvm",
        build_file = clean_dep("//third_party/llvm:llvm.autogenerated.BUILD"),
        sha256 = "599b89411df88b9e2be40b019e7ab0f7c9c10dd5ab1c948cd22e678cc8f8f352",
        strip_prefix = "llvm-7a7e03f906aada0cf4b749b51213fe5784eeff84",
        urls = [
            "https://mirror.bazel.build/github.com/llvm-mirror/llvm/archive/7a7e03f906aada0cf4b749b51213fe5784eeff84.tar.gz",
            "file:///home/test/bazel_tools/llvm-7a7e03f906aada0cf4b749b51213fe5784eeff84.tar.gz",
        ],
    )

修改完成后，重新切换到tensorflow-1.15.0目录，执行如下命令编译xlacompile工具：
```
bazel build --config=monolithic //tensorflow/compiler/aot:xlacompile
```
若提示“ERROR: An error occurred during the fetch of repository 'bazel_skylib':”，则进入下一步继续修改bazel_skylib相关的文件。
修改.cache目录中的相关链接：
切换到$HOME目录，使用grep -r bazel-skylib.0.8.0 .cache/命令查看.cache目录哪个文件引用bazel-skylib.0.8.0.tar.gz的url，根据返回信息，进入引用该url文件所在的目录，例如.cache/bazel/_bazel_test/abd37aaac8a380ca5a3f13938322fcb2/external/io_bazel_rules_closure/closure/repositories.bzl（该路径只是样例，请以用户实际编译后的为准）

打开该文件，将其中的url改为：
```
file:///home/test/bazel_tools/bazel-skylib.0.8.0.tar.gz
```
执行wq!保存退出。

父主题： FAQ