Configuration is hard so make it somewhat easier by adding a layer of abstraction into your stack.
This provides a simple restful interface to create and update configurations.
A simple flask server running in your favorite wsgi container for the connections and a RethinkDB database in the backend.
Very cool. We currently provide the creation of configuration files based on some user_id and some config_name. These configs contain the values as well as a simple audit trail that can be used to follow changes throughout a configurations life time.
Input:
curl -XPOST http://localhost:5000/values/1/test -d'{"test": 1}' -H "Content-Type: application/json"Output:
{
"audit_trail": [
{
"changes": [
{
"action": "added",
"key": "test",
"value": 1
}
],
"created_at": "Sun Mar 08 2015 05:41:10 GMT+00:00",
"version": 1
}
],
"config_id": "1-test",
"id": "c51b1c98-5d1c-4657-86d2-ca19020192aa",
"values": {
"test": 1
},
"version": 1
}As you can see there is a lot of information. But, the simple is every config is versioned, contains an audit trail, as well as the latest configuration values. Now, let's remove the test value and add a hello value.
Intput:
curl -XPOST http://localhost:5000/values/1/test -d'{"hello": "world"}' -H "Content-Type: application/json"Output:
{
"audit_trail": [
{
"changes": [
{
"action": "added",
"key": "test"
}
],
"created_at": "Sun Mar 08 2015 05:41:10 GMT+00:00",
"version": 1
},
{
"changes": [
{
"action": "removed",
"key": "test",
"value": 1
},
{
"action": "added",
"key": "hello",
"value": "world"
}
],
"created_at": "Sun Mar 08 2015 05:46:14 GMT+00:00",
"version": 2
}
],
"config_id": "1-test",
"id": "c51b1c98-5d1c-4657-86d2-ca19020192aa",
"values": {
"hello": "world"
},
"version": 2
}As you can see the audit log was updated with the changes and the values reflect the updated configuration.
- authentication
- streaming changes
- verification of configurations through schemas
We recommend running this service within something similar to guinicorn. The other consideration is that this service could live on many different nodes being that all of the heavy lifting is performed inside of RethinkDB currently.