CentCom is a suite of software used for implementing a data warehouse of bans for Space Station 13 from a variety of public sources. It is capable of ingesting data from several public ban lists, and provides an API through which to search these collected records.
Considering contributing your server's bans to CentCom? Check out our onboarding documentation to help inform you and answer questions you may have.
There is a central instance of CentCom running at https://centcom.melonmesa.com, where you can also find the compiled Swagger documentation to read about and experiment with the API.
Looking to run the standardized ban exporter? See the documentation here.
To run your own local instance of CentCom, you must have the following...
- A supported database to store the ban data within (Postgres, MySql, MariaDB). You will need to create a database for the CentCom data, and users for the parsing server and API server.
- A machine to run the CentCom Parsing Server
- A machine to run the CentCom API Server (this could be the same machine as the parsing server)
Once you have collected these things, you must then:
- Get the latest CentCom release
- Edit the
hostsettings.jsonof the API server to configure the URLs that the server will bind to. This should be an HTTPS url. By default, it will bind to any IP on port 6658. - Edit the appconfig.json of both the parsing server and API server to be configured for your database. You must provide
a connection string, as well as the type of the database. This can be one of
postgres,mysql, ormariadb. I would strongly recommend having two different accounts/roles on your database for CentCom, one for the parsing server which has write access to the CentCom database, and one for the API server which can only read. This helps to isolate any possible security concerns./
Finally:
- Run the parsing server first (
CentCom.Server) to create the database schema and apply any necessary migrations to the database. This will also begin to populate the database based on the parsing schedule, which by default will parse all ban sources for new bans every 5 minutes except for at 00 and 30 minutes of every hour, at which time a full ban pass will occur. - AFTER the migration/database setup has occurred successfully, you can now start the API server (
CentCom.API) without any concerns. This server will now take API requests, the documentation of which you can view at the/swaggerpages.
I will likely make this configurable in the future, but at the moment it will run at every 5 minute interval in the
hour (XX:05, XX:10...) for a latest refresh (getting all new bans), except for XX:00 and XX:30, which
will be full refreshes. At this point the entire source set is taken from each ban source and compared with what is
stored in the database, at which point any differences will be resolved.
Thanks to Sentry for sponsoring CentCom! Their platform helps to support error logging and improve performance on the primary hosted instance of CentCom.
PRs can be opened on this repository to propose changes to the codebase. There are a few things to note...
If you are going to add an additional ban source, you typically have to add two new objects: a subclass
of CentCom.Server.BanSources.BanParser, which is required, and optionally a BanService
in CentCom.Server.Services which helps to isolate the code used for parsing web pages and other forms of online
resources.
When you sub-class BanParser, you must override the Sources field to provide a definition of the ban sources
that are for that server. As well as this, you optionally may override the SourceSupportsBanIDs field if the server
exposes unique Ban IDs from their own database. These are ideal, as it removes any ambiguity as to if two bans that are
found from a ban source are the same ban, so use them when provided. If you use a ban source with IDs and flag it as
such, you MUST set the BanID property on the Ban itself.
As well as this, you must provide implementations for two different methods, FetchNewBansAsync(),
and FetchAllBansAsync(). FetchNewBansAsync() should only return bans that are new as-of the last time that bans
were fetched from this source, typically by getting the last bans for that source from the database and getting new bans
from the source until you find one of the pre-existing bans. FetchAllBansAsync() should not have this behaviour, and
instead should provide all bans from the ban source. This method is used on a much less common schedule to ensure that
the ban database stores any bans that have updated reasons, or have been changed in general. It will also help to
identify bans that have been deleted from the source so that we may do the same in our database.
Once you provide these methods, the base implementation of BanParser will handle the rest of the work of sorting
through and storing bans. You essentially just have to provide it a means of getting those bans.
If you make any changes to database objects (typically types in CentCom.Common.Models), you will need to generate a
migration. You can do this by installing dotnet-ef tools using dotnet tool install --global dotnet-ef in a
terminal, and then by running add-migration.bat with a single argument which is the name of the
migration to generate.
Ex: ./add-migration AddCoolPR will generate a migration for all data sources called AddCoolPR.
Note: You must add a migration for each database type supported, these types should be in CentCom.Common.Data,
subclassed from DatabaseContext. Any new DatabaseContext subclasses should be added to add-migration.bat.
If you wish to support an additional database backend, you will need to create a new subclass
of CentCom.Common.Data.DatabaseContext. If necessary, you can overload how model types are stored in the new
database backend by providing an override for the OnModelCreating method of DatabaseContext, which you can see
an example of in CentCom.Common.Data.MySqlDbContext. As well as this, you will need to add a migration for your new
database backend. You can do so using the same format as found in 'Making Changes to Database Objects' above.
