VIT Servicing Station
VIT as a service (VaaS)
Building tips and tricks
We use diesel for database (sqlite3) integration. Please refer to the diesel_cli documentation to understand how to operate with migrations and setup.
Diesel generates rust code based on a SQL migration script (/migrations/*/up.sql) when running the migration with diesel_cli.
Diesel code generation is configured in the diesel.toml file. Right now it just contains the path on where the rust code should be generated.
Another file to look at is the .env file. This file holds the environment variables used by this project sql configuration.
diesel uses a DATABASE_URL variable to know where should he generate the database file.
Server settings
The server settings can be loaded via three options, environment variables, command line flags and a json file. These configurations follows some priority from low to high. Env variables are overwritten by command line flags if used and those before are overwritten by the json file if used too.
Env variables
DATABASE_URL->URLfor the database connectionTLS_CERT_FILE-> Path to server X.509 certificate chain file, must be PEM-encoded and contain at least 1 itemTLS_PRIVATE_KEY_FILE-> Path to server private key file, must be PKCS8 with single PEM-encoded, unencrypted keyCORS_ALLOWED_ORIGINS-> Semicolon separated list of allowedCORSorigins. For example:https://foo.test;https://test.foo:5050
Command line flags
The command line flags can be retrieved using the --help when running the server:
--address <address> Server binding address [default: 0.0.0.0:3030]
--allowed-origins <allowed-origins> If none provided, echos request origin [env: CORS_ALLOWED_ORIGINS=]
--block0-path <block0-path> block0 static file path [default: ./resources/v0/block0.bin]
--cert-file <cert-file>
Path to server X.509 certificate chain file, must be PEM-encoded and contain at least 1 item [env:
TLS_CERT_FILE=]
--db-url <db-url> Database url [env: DATABASE_URL=] [default: ./db/database.sqlite3]
--in-settings-file <in-settings-file> Load settings from file
--max-age-secs <max-age-secs> If none provided, CORS responses won't be cached
--out-settings-file <out-settings-file> Dump current settings to file
--priv-key-file <priv-key-file>
Path to server private key file, must be PKCS8 with single PEM-encoded, unencrypted key [env: TLS_PK_FILE=]Some of the flags default to the environment variables explained above is not set. Some of them have default values as fallback in case nor the env variable nor the flag is set.
JSON file configuration
Additionally if the you can load the whole configuration from a json file providing the path to the file within the --in-settings-file.
An example of the contents of the file would be like this:
{
"address" : "0.0.0.0:3030",
"tls" : {
"cert_file" : "./foo/bar.pem",
"priv_key_file" : "./bar/foo.pem"
},
"cors" : {
"allowed_origins" : ["https://foo.test", "https://test.foo"],
"max_age_secs" : 60
},
"db_url": "./database.sqlite3",
"block0_path": "./test/bin.test"
}There is an option to dump a configuration into a JSON file with the --out-settings-file providing the path to the out file.
This option will dump the configuration with the defaults, already set environment variables or provided flags into the file.
Playground
The VIT servicing station offers a GraphQL playground endpoint (api/v0/graphql/playground) designed to offer a user friendly approach to the API.
Some considerations when using the playground:
- There is no need of an
API-Tokenfor accessing it. But, the token is needed for any request performed by the playground. In order to use your API-Token just add it inside the json configuration for headers in the bottom left part of the webpage. For example:{ "API-Token" : "your api token here" } - Since the API-Token is needed for any request. Both the docs and the schema will not be available till the header is filled.
- Depending on the deployment and/or your server configuration you may need to configure properly the CORS settings in order to make the playground work as expected.
CLI
The vit-servicing-station-cli is an accompanying tool to interact with some of the ecosystem.
Right now it offers the following commands:
api-token
generate
It is possible to generate api tokens (URL safe base64 encoded) with a simple command. For example:
❯ ./vit-servicing-station-cli api-token generate
F-4QxU3FrbH7qgIt can be combined with two (optional) arguments:
--nnumber of tokens to generate--sizelength (in bytes) of the tokens
add
We can add a token to some db using the tool too:
./vit-servicing-station-cli api-token add --db-url ../../db/vit_station_new.db --tokens 1CNDAo43fo4ktQ 0wNbdTDMJCFcnwWe need to provide the url to the database where we want it to be inserted (with --db-url) and the tokens we want too
insert (with --tokens followed by the tokens).
Notice that the token is in the same URL safe base64 encoding as we generated in the previous command.
If not specifying the --tokens argument the cli will read the input from the standard input the tokens we want to insert.
This enables the command to be piped from another command, for example:
./vit-servicing-station-cli api-token generate --size 10 --n 10 | ./vit-servicing-station-cli api-token add --db-url ../../db/vit_station_new.db
Formed in 2009, the Archive Team (not to be confused with the archive.org Archive-It Team) is a rogue archivist collective dedicated to saving copies of rapidly dying or deleted websites for the sake of history and digital heritage. The group is 100% composed of volunteers and interested parties, and has expanded into a large amount of related projects for saving online and digital history.

