glenux/vagrant-cachier-ng
glenux/vagrant-cachier-ng
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Improve dev docs

Browse source
This commit is contained in:
Fabio Rehm 2013-12-07 12:43:49 -02:00
parent c4ba0c9445
commit 54407fd2da
1 changed files with 24 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

31
docs/development.md
View file

@ -15,22 +15,39 @@ vagrant plugin install pkg/vagrant-cachier-VERSION.gem
## Sanity checks ## Sanity checks
There are also some [Bats](https://github.com/sstephenson/bats) tests that basically While we don't get to implement some proper unit testing, there are some basic [Bats](https://github.com/sstephenson/bats)
acts as a [sanity check](https://github.com/fgrehm/vagrant-cachier/blob/master/spec/acceptance/sanity_check.bats) tests that basically acts as a [sanity check](https://github.com/fgrehm/vagrant-cachier/blob/master/spec/acceptance/sanity_check.bats)
that you can run with `bats spec/acceptance` in case you are planning to submit a that you can run with `bats spec/acceptance` in case you are planning to submit a
Pull Request :) Just keep in mind that it might take a while to run if you are Pull Request. Just keep in mind that it might take a while to run if you are
using the default VirtualBox provider. using the default VirtualBox provider.
## How to create a new bucket? ## How to create a new bucket?
The concept of a cache _bucket_ is pretty simple to understand, we basically symlink a folder under the guest's `/tmp/vagrant-cache` into the folder where other tools keep downloaded packages. For example, in the case of the [apt bucket](https://github.com/fgrehm/vagrant-cachier#apt), we symlink `/var/cache/apt/archives` to `/tmp/vagrant-cache/apt` so that `.deb` packages downloaded with `apt-get install` can be reused when bringing machines up from scratch. The concept of a cache _bucket_ is pretty easy to understand, we basically
symlink a folder under the guest's `/tmp/vagrant-cache` into the folder where
other tools keep downloaded packages. For example, in the case of the
[apt bucket](buckets/apt), we symlink `/var/cache/apt/archives` to
`/tmp/vagrant-cache/apt` so that `.deb` packages downloaded with `apt-get install`
can be reused when bringing machines up from scratch.
If you want to see some other package manager supported, you'll first need to know if it has some sort of caching in place, where does it get stored and if it needs some extra configuration. For some managers that path will be fixed (like the canonical apt bucket), for others you might need to read some configuration by [running a command](https://github.com/fgrehm/vagrant-cachier/blob/master/lib/vagrant-cachier/cap/linux/rvm_path.rb#L10) on the guest VM (like [rvm](https://github.com/fgrehm/vagrant-cachier#rvm)) and you might even to [tweak some configs](https://github.com/fgrehm/vagrant-cachier/blob/master/lib/vagrant-cachier/bucket/yum.rb#L20) on the guest (like the [yum](https://github.com/fgrehm/vagrant-cachier#yum) bucket) If you want to see some other package manager supported, you'll first need to
know if it has some sort of caching in place, where does it get stored and if
it needs some extra configuration. For some managers that path will be fixed
(like the canonical apt bucket), for others you might need to read some
configuration by [running a command](https://github.com/fgrehm/vagrant-cachier/blob/master/lib/vagrant-cachier/cap/linux/rvm_path.rb#L10)
on the guest VM (like [rvm](buckets/rvm)) and you might even need to [tweak some configs](https://github.com/fgrehm/vagrant-cachier/blob/master/lib/vagrant-cachier/bucket/yum.rb#L20)
on the guest (like the [yum](buckets/yum) bucket).
There's currently a lot of duplication around the "installation" of cache buckets so there's plenty of source code for you to read in order to understand how things work under the hood but if you don't feel comfortable reading / writing Ruby code you can provide a high level overview of how to do things using plain old bash. There's currently a lot of duplication around the "installation" of cache buckets
so there's plenty of source code for you to read in order to understand how
things work under the hood but if you don't feel comfortable reading / writing
Ruby code you can provide a high level overview of how to do things using plain
old bash.
For example, if you were to explain how to set up the rvm bucket, the script below should give vagrant-cachier maintainers an overview of how to set things up: For example, if you were to explain how to set up the rvm bucket, the script
below should give vagrant-cachier maintainers an overview of how to set things
up:
```bash ```bash
# Check is rvm is installed # Check is rvm is installed