Application programming interfaces (APIs) are like fuel for digital transformation, enabling companies to develop applications and microservices quickly. DataTrucker.IO is a new, free and open source (Apache 2.0) no-code/less-code API backend that decreases the time required to build APIs.
DataTrucker reads simple JSON and YAML configs and produces the code to convert it into an API. In addition to creating the code, it also hosts the codebase on a Node.js server and makes it immediately available for public consumption.
DataTrucker handles most of the everyday activities a developer needs to do on a new project, including:
- Creating an API endpoint with specified business logic (using simple plugins)
- Applying standard role-based access control (RBAC)
- Applying authorization logic
- Applying hardening on endpoints
- Handling log management
- Connecting to a variety of systems
- Modularizing business logic
- Doing it all with little to no code
Installing DataTrucker
This tutorial shows how to install DataTrucker on OpenShift and build an API for a Postgres database. The process is similar in the Kubernetes environment.
[ Download The API owner's manual to learn 7 best practices of effective API programs. ]
Step 1: Create a namespace called trucker
oc new-project trucker
Step 2: Download and install the application
DataTrucker.IO is available in OpenShift's OperatorHub and can be added to your cluster as an Operator.
Step 3: Navigate into the Operators
Click Operators in the left navigation pane and open the DataTrucker.IO Operator.
Step 4: Create a DataTrucker config by running the YAML object
Create a Persistent Volume Claim (PVC) for a database backend. (Note: The Postgres database provided below uses Crunchy Data containers to get started. I recommend using a hardened geo-redundant database for a production workload.)
- Create a PVC called
samplepvc. - Create an instance of a DatatruckerConfig object.
- Before you click Create, ensure
TempDB.enabledis True in the DatatruckerConfig object. This is required for prototyping the demo below.
You can find an example in DataTrucker's GitLab repository.
oc apply -f DatatruckerConfig.yaml
The config object Kind: DatatruckerConfig creates the following:
-
Postgres database: You can create a temporary non-hardened Crunchy Data database for following this tutorial by enabling the following in the DataTrucker config (recall that for a production workload, you should use a hardened geo-redundant database):
TempDB: enabled: true pvc: samplepvc -
Database configuration: In production systems, use a geo-redundant Postgres database:
user: testuser password: password databasename: userdb hostname: db type: pg port: 5432 -
Crypto configuration: Create the crypto configuration:
API: cryptokeys: |- ....See detailed information in the Production hardening section of the DataTrucker.IO documentation.
-
API server backend configuration: Create the API server backend configuration:
API: name: API loginServer: |- .... managementServer: |- .... jobsServer: |- ....
Step 5: Create a login and management endpoints
Create an endpoint for obtaining a login token:
apiVersion: datatrucker.datatrucker.io/v1
kind: DatatruckerFlow
metadata:
name: login-endpoint
spec:
Type: Login
DatatruckerConfig: < the name of the config object created in step 4 >
Create an endpoint for RBAC management and credentials:
apiVersion: datatrucker.datatrucker.io/v1
kind: DatatruckerFlow
metadata:
name: management-endpoint
spec:
Type: Management
DatatruckerConfig: < the name of the config object created in step 4 >
The items above create the deployments and service endpoints for the user interface (UI) and the management API.
Step 6: Expose the management endpoint
Expose the routes:
$ oc get svc | grep endpoint
login-endpoint ClusterIP 10.217.5.89 <none> 80/TCP 3m43s
management-endpoint ClusterIP 10.217.5.220 <none> 80/TCP 3m29s
management-endpoint-ui ClusterIP 10.217.4.42 <none> 80/TCP 3m28s
$ oc expose svc management-endpoint-ui
route.route.openshift.io/management-endpoint-ui exposed
$ oc expose svc login-endpoint
route.route.openshift.io/login-endpoint exposed
$ oc get routes
NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
login-endpoint login-endpoint-trucker.apps-crc.testing login-endpoint 8080 None
management-endpoint-ui management-endpoint-ui-trucker.apps-crc.testing management-endpoint-ui 9080 None
Step 7: Log into the UI with a web browser
- Create an admin user:
Image
(Gaurav Shankar, CC BY-SA 4.0) - Log in as that admin user:
Image
(Gaurav Shankar, CC BY-SA 4.0)
Step 8: Create a Postgres credential for the API
Now that DataTrucker is installed and configured, you are ready to build an API.
Create Postgres credentials for the database of your choice:
- Expand the left navigation bar.
- Select Credentials.
- Open the Postgres Credentials pane.
- Click Create Credentials.
- Enter your database details.
Step 9: Create a Postgres API
Create a Flow object with the following job spec.
---
apiVersion: datatrucker.datatrucker.io/v1
kind: DatatruckerFlow
metadata:
name: my-first-api
spec:
DatatruckerConfig: datatruckerconfig-sample
JobDefinitions:
- credentialname: db < cred name from step 8 >
job_timeout: 600
name: postgres1
restmethod: POST
script: 'select ''[[userinput]]'' as userinput; ' < query you want to execute>
tenant: Admin
type: DB-Postgres
validations:
properties:
userinput:
maxLength: 18
pattern: '^[a-z0-9]*$'
type: string
type: object
- credentialname: db < cred name from step 8 >
job_timeout: 600
name: postgres2
restmethod: GET
script: select * from information_schema.tables < query you want to execute>
tenant: Admin
type: DB-Postgres
Type: Job
This spec creates a new microservice to host the API, and the microservice has two APIs on its route:
- postgres1:
- Gets current date and user sent parameter into the SQL
- Is a POST request
- Sanitizes input for the user-input variable
userinput
- postgres2:
- Gets a list of tables available
- Is a GET request
[ Free download: Advanced Linux commands cheat sheet. ]
Now, search for the service:
$. oc get svc | grep my-first-api
my-first-api ClusterIP 10.217.5.116 <none> 80/TCP 45s
$. oc expose svc my-first-api
route.route.openshift.io/my-first-api exposed
$. oc get routes | grep my-first-api
my-first-api my-first-api-trucker.apps-crc.testing my-first-api 8080 None
Now that you have a URL, go test it out.
The URL is in the following form:
http://<your api route>/api/v1/jobs/<name of the JobDefinitions defined in the yaml>
In the above example, you created two JobDefinitions:
- postgres1 of type POST
- postgres2 of type GET
Step 10: Test out your APIs
Get a login token from the login endpoint:
curl --location --request POST 'http://login-endpoint-trucker.<wilcard.domain>/api/v1/login' \
--header 'Content-Type: application/json' \
--data-raw '{
"username": "xxx",
"password": "xxxxxxxx",
"tenant": "Admin"
}'
Response:
{
"status": true,
"username": "xxx",
"token": "xxxxxxxxxxxx"
}
Now use the login token against your APIs:
-
The first API:
curl --location --request POST 'http://my-first-api-trucker.<wilcard.domain>/api/v1/jobs/postgres1' \ --header 'Authorization: Bearer xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx \ --header 'Content-Type: application/json' \ --data-raw '{ "userinput": "myfirstresponse" }' Response: { "reqCompleted": true, "date": "2021-09-05T22:05:58.064Z", "reqID": "req-3w", "data": { "command": "SELECT", "rowCount": 1, "oid": null, "rows": [ ............. -
The second API:
curl --location --request GET 'http://my-first-api-trucker.<wilcard.domain>/api/v1/jobs/postgres2' \ --header 'Authorization: Bearer xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx' Response: { "reqCompleted": true, "date": "2021-09-05T22:03:58.389Z", "reqID": "req-35", "data": { "command": "SELECT", "rowCount": 185, "oid": null, "rows": [ { " .......
For more information, watch DataTrucker.IO's overview video.
This article is adapted from the DataTrucker.IO blog and is republished with permission.
