and our snapshot optimizations.
If a previous base image has a volume, the directory is added to the
list of files to snapshot. That directory may not actually exist in the image.
This change refactors the build loop a bit to make cache optimization easier in the future. Some notable changes:
The special casing around volume snapshots is removed. Every volume is added to the snapshotFiles list for every command that will snapshot anyway.
Snapshot saving was extracted to a sub-function
The decision on whether or not to snapshot was extracted
* Rework cache key generation a bit.
Cache keys are now based on the previous commands, rather than the previous state
of the filesystem.
* Refactor command interface a bit, only cache the context for commands that use it.
CacheCommand returns true if the command should be cached. Currently,
it's only true for RUN but can be added to ADD/COPY later on (these are
different since the contents of files for ADD/COPY need to be included
in the cache key generation).
I also changed CreatedBy to String so that we can log each command
before cache extraction or regular execution takes place.
Kaniko uses mtime (as well as file contents and other attributes) to
determine if files have changed. COPY and ADD commands should _always_
update the mtime, because they actually overwrite the files. However it
turns out that the mtime can lag, so kaniko would sometimes add a new
layer when using COPY or ADD on a file, and sometimes would not. This
leads to a non-deterministic number of layers.
To fix this, we have updated the kaniko commands to be more
authoritative in declaring when they have changed a file (e.g. WORKDIR
will now only create the directory when it doesn't exist) and we will
trust those files and _always_ add them, instead of only adding them if
they haven't changed.
It is possible for RUN commands to also change the filesystem, in which
case kaniko has no choice but to look at the filesystem to determine
what has changed. For this case we have added a call to `sync` however
we still cannot guarantee that sometimes the mtime will not lag, causing the
number of layers to be non-deterministic. However when I tried to cause
this behaviour with the RUN command, I couldn't.
This changes the snapshotting logic a bit; before this change, the last
command of the last stage in a Dockerfile would always scan the whole
file system and ignore the files returned by the kaniko command. Instead
we will now trust those files and assume that the snapshotting
performed by previous commands will be adequate.
Docker itself seems to rely on the storage driver to determine when
files have changed and so doesn't have to deal with these problems
directly.
An alternative implementation would use `inotify` to track which files
have changed. However that would mean watching every file in the
filesystem, and adding new watches as files are added. Not only is there
a limit on the number of files that can be watched, but according to the
man pages a) this can take a significant amount of time b) there is
complication around when events arrive (e.g. by the time they arrive,
the files may have changed) and lastly c) events can be lost, which
would mean we'd run into this non-deterministic behaviour again anyway.
Fixes#251
* Vendor changes for go-containerregistry switch.
* Manual changes for go-containerregistry switch.
The biggest change is refactoring the tarball unpacking.
* Pull more of container-diff out.
* More vendor removals.
* More unit tests.
* adding VOLUME command
* proper test project
* general fixes
* fixing project name
* fixing volume unit test
* fixing integration test
* adding tests
* adding util test
* fixing test
* actually create the volume mounted directory
* fix test
Tejal Desai
Cole Wippern
Taylor Barrella
dlorenc
peter-evans
Priya Wadhwa
Christie Wilson
Nick Kubala
Jon Johnson
Matt Rickard
Sharif Elgamal