You can view this tutorial in presentation mode.
Here is installed SlapOS itself.
This is the directory containing all Software Releases installed on this node (wordpress, erp5, ...).
Each Software Release is self-contained in its own subdirectory, so that no Software Release can pollute another one. The name of this directory is the URL of the Software Release, hashed by MD5. So for example wordpress version "slapos-0.90", which URL is
http://git.erp5.org/gitweb/slapos.git/blob_plain/refs/tags/slapos-0.90:/software/wordpress/software.cfgwill be installed in
/opt/slapgrid/e884cf83212b2caf90bbb5cba481f58b
This directory contains all Computer Partitions.
This directory contains all configuration files for SlapOS.
Note: location of this directory is compatible with FHS 2.3.
SlapOS uses Buildout as its build system. Imagine slapgrid as a simple wrapper of buildout. When slapgrid is run, it will run buildout to install Software Releases and to deploy instances.
SlapOS uses Buildout to install software and deploy instances.
In SlapOS, a Software Release is defined by its URL. This URL is in fact a Buildout Profile. When SlapOS installs a Software Release, it launches Buildout with this URL. For example, to install wordpress, slapgrid will run something like (simplified):
buildout -c http://git.erp5.org/gitweb/slapos.git/blob_plain/refs/tags/slapos-0.90:/software/wordpress/software.cfgin the correct directory.
Once a Software Release has successfully been installed, it will be frozen, i.e SlapOS Node will just check for its existence but won't process it anymore.
Deployment of an instance:
While installing a Software Release, as a convention, buildout generates an "instance.cfg" file, describing how to deploy the service.
When deploying an instance of a Software Release, slapgrid will use this file, running something like (simplified):
cd /srv/slapgrid/slappart12/ cp /opt/slapgrid/e884cf83212b2caf90bbb5cba481f58b/instance.cfg . buildout
SlapOS will process (i.e run Buildout) each instance at least once per day by default. This periodicity can be overwritten by the Software Release by creating a "$SOFTWARE_RELEASE_INSTALL_DIRECTORY/periodicity" text file containing the periodicity in seconds.
Example of recipes:
slapos.git, the official git repository containing all the needed profiles and recipes, is separated into several directories.
Example: Wordpress
You can find here all Software Releases, which can be considered as "end user services". Each Software Release represents a full service (like Wordpress, or ERP5).
In each Software Release directory, there is at least a "software.cfg" Buildout profile.
Because it is a Buildout profile, it can reuse other Buildout profiles from "component" or "stack" directory using the Buildout "extends" mechanism.
Example: Apache, MySQL, OpenSSL
Here are all atomic components (all libraries or binaries, like Apache, MariaDB, CouchDB, cURL, OpenSSL, libxml, ncurses, ...) used in the Software Releases.
Those profiles describe how to install a component: how to compile it, how to fetch it, what are the dependencies...
Example: LAMP
If you need, you can also use what is called "stack". A stack is a group of components working together, in a reusable way. Most stack provide an instance.cfg file.
For example, when creating a PHP Sofware Release, you can extend the "LAMP" stack, containing all what is needed to run PHP applications (Apache, MySQL, etc) so that you can define only what is specific to your software in a few lines. Your software.cfg extending lamp stack will only contain a few lines.
Note: a special stack is stack/slapos.cfg, defining what is usually needed in every Software Releases, like using a special version of Buildout or fetching the slapos.cookbook egg. It should be always used to ease development of Software Release.
Example: Apache deployment description
Here are all recipes used for deployment of instances, called from instance.cfg file of a Software Release. This directory forms content of the "slapos.cookbook" egg.
All the SlapOS Node processes are run automatically via simple cron entries:
Of course, you can manually run those commands if you want to.
For more information, please contact Jean-Paul, CEO of Nexedi (+33 629 02 44 25).