Platform Team/Server Kit/Architecture

The purpose

The Sugar Server Kit initiative is an attempt to achieve the following major goals:

Provide a split between the community level project (Sugar Server Kit) and any number of downstream solutions based on the community project. This should stimulate the downstream community to contribute to this upstream community project, facilitating reuse of its experience in all other downstreams;
Treat the community project as a collection of useful tools, created and supported by community contributors, that might be composed into a final deployment solution on purpose, i.e., Sugar Server Kit is not an OS or a final solution, but rather a bunch of tools that might be launched on any major GNU/Linux distribution at the deployment level. And because some of these tools might be implemented in several ways, it should make the acceptance process of new features by upstream more flexible;
The whole system should be as reliable as possible. Thus, the community project will provide a decent testing environment (several levels of automatic and human driven tests at the top level), which might be used not only for Sugar Server Kit itself, but for deployment solutions as well;
Be GNU/Linux distribution agnostic, different deployments might decide to use different GNU/Linux distributions;
It is not only about supporting XO laptops, but about any Sugar based environments.

Components

There are several major Sugar Server Kit project components that will be mentioned on this page:

sugar-server provides basic sugar specific services,
sugar-server-templates provides configuration templates that might be reused in downstream solution,
mace process configuration sources, e.g., from sugar-server-templates,

The only relationship between them are:

sugar-server-templates contain basic configuration for sugar-servers.
sugar-server-templates makes sense only after processing its content by mace.

These are building blocks for the final solutions in downstream deployments.

Server functionality models

Sugar Server Kit components are being designed to accomplish different functionality models, when the particular model might include only limited number of ingredients. The following models might describe Sugar Server Kit design, the final model might be an intermediate variant of them.

Only sugar-server package

The model with minimal Sugar Server Kit design influence.

The key points:

Existing and configured out of Sugar Server Kit servers at schools.
School admins need to take care about setting up sugar-server specific requirement configuration for external environment.
School admins install the sugar-server package from upstream binary repository, and just launch it.
sugar-server starts to serve all sugar boxes around, providing only basic sugar specific functionality.
sugar-server doesn't break the system configuration (it touches nothing).

The maintaining process is the same as for any other service launched on a server.

Dumb school servers

The model where the server at school is entirely dependent on Sugar Server Kit design decisions. There are two types of machines:

Dumb servers at schools under Sugar Server Kit control. The explicit intention is to minimize maintaining intervention as much as possible.
The Mothership to minimal control of school servers and handling the rest of functionality that people at schools need.

The key points:

Functionality of school servers is simple, only basic services.
Complex, thus not trivial for maintenance, services are on the mothership and singular, i.e., to minimize maintaining costs and let the minimal number of skilled personal support them effectively.
A set of school server's services tends to be constant, at least it doesn't require regular intervention to add/remove services.
All school serves have exactly the same set of services.

This functionality model involves Mace and sugar-server-templates components. The entirely content of school servers is:

a set of packages, including the main one that contains:
- all services as dependencies,
- upstream Mace configuration as sugar-server-templates dependency,
- downstream Mace configuration;
Mace environment file, e.g., the one from sugar-server-templates examples;
pure data, like leases and content black lists, to fetch as-is from the Mothership.

After installing the main package, it will provide possibility to fetch pure data from the Mothersip, and finally running Mace to complete school server setup.

The maintaining process will be:

sugar-server-templates provides unattended packages update on school servers, packages to update come from:
- GNU/Linux official repositories with security updates,
- Sugar Server Kit upstream repository that follows the Statement of purpose for releases, i.e., declares that newly appeared updates should not break already deployed systems;
taking care that pure data on the Mothership is up-to-date, e.g., leases are properly created, content filtering blacklists are fresh, etc.;
do occasional changes in the main school server package to add new services or tune downstream Mace configuration and upload it to the repository that will be used for unattended updates on school servers.

Highly maintained school servers

Servers at schools might be not only simple like in the previous model. They might contain complex services like content management systems. This usecase might require more regular maintaining scenarios, e.g., using configuration tools like Puppet or CFEngine, having more detailed monitoring, etc.

It seems that the only useful Sugar Server Kit component here is the sugar-server.

Client functionality models

There is only one Sugar Server Kit component to work on a client side, sugar-client the implements Client API. Sugar-client works directly with sugar-server using its public API.

Distribution model

Sugar Server Kit is designed to avoid patching its sources in downstream. Reusing upstream binary packages as-is from repositories on http://download.sugarlabs.org is how Sugar Server Kit is being designed. Downstream might create new packages, that don't have file collisions with upstream packages, to have tweaks for local environment. It is being accomplished by:

sugar-server project has services formed as plugins, from downstream packages. Such services might be:
- enabled/disabled
- added new services, e.g., variants of existing services that are highly tuned for local needs
mace does not create any final configuration level logic and might be used to process any downstream configuration for services that it supports.

Core packages are per-component and might be reused as-is in downstream, in a way that is most practical for them.

Deployment model

It is an entirely downstream decision as to how to deploy Sugar Server Kit based solutions.

Development model

Having a decent testing infrastructure is the major intention for Sugar Server Kit components. Every component needs to have, at least, unit tests for its internals. Components like sugar-server have also integration tests to cover integration issues for its parts.

Except component specific tests, it is important to have system integration testing when all components are being tested in collaboration. In that case, the sugar-unit component, sugar client bot, will help.

The reliable testing environment should help with avoiding regressions while following Sugar Server Kit statement of purpose for releases.