Update vendor of Buildah and imagebuilder

Fixes the testing issues we are hitting.

Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This commit is contained in:
Daniel J Walsh
2019-03-13 08:01:28 -04:00
parent 8b637bd78c
commit adad93342c
16 changed files with 118 additions and 59 deletions

View File

@@ -1,4 +1,4 @@
Docker / OCI Image Builder
OCI Image Builder
==========================
[![Go Report Card](https://goreportcard.com/badge/github.com/openshift/imagebuilder)](https://goreportcard.com/report/github.com/openshift/imagebuilder)
@@ -6,22 +6,22 @@ Docker / OCI Image Builder
[![Travis](https://travis-ci.org/openshift/imagebuilder.svg?branch=master)](https://travis-ci.org/openshift/imagebuilder)
[![Join the chat at freenode:openshift-dev](https://img.shields.io/badge/irc-freenode%3A%20%23openshift--dev-blue.svg)](http://webchat.freenode.net/?channels=%23openshift-dev)
Note: this library is beta and may contain bugs that prevent images from being identical to Docker build. Test your images (and add to our conformance suite)!
Please test your images (and add to our conformance suite)!
This library supports using the Dockerfile syntax to build Docker
compatible images, without invoking Docker build. It is intended to give
clients more control over how a Docker build is run, including:
This library supports using the Dockerfile syntax to build OCI & Docker
compatible images, without invoking a container build command such as `buildah bud` or `docker build`. It is intended to give
clients more control over how they build container images, including:
* Instead of building one layer per line, run all instructions in the
same container
* Set Docker HostConfig settings like network and memory controls that
are not available when running Docker builds
* Set HostConfig settings like network and memory controls that
are not available when running container builds
* Mount external files into the build that are not persisted as part of
the final image (i.e. "secrets")
* If there are no RUN commands in the Dockerfile, the container is created
and committed, but never started.
The final image should be 99.9% compatible with regular docker builds,
The final image should be 99.9% compatible with regular container builds,
but bugs are always possible.
Future goals include:
@@ -54,9 +54,6 @@ $ imagebuilder --mount ~/secrets/private.key:/etc/keys/private.key path/to/my/co
Any processes in the Dockerfile will have access to `/etc/keys/private.key`, but that file will not be part of the committed image.
Running `--mount` requires Docker 1.10 or newer, as it uses a Docker volume to hold the mounted files and the volume API was not
available in earlier versions.
You can also customize which Dockerfile is run, or run multiple Dockerfiles in sequence (the FROM is ignored on
later files):