API Rest developed using Spring Boot, responsible for providing the data managed by the system. In this project, there is also a Worker service, which processes asynchronous messages regarding executions of scientific workflows, tasks, jobs and other commands that may be setup by the platform. For authentication OAuth 2.0 and JWT was used.

To test the project locally, simply clone the repository and import into Eclipse (or any other IDE of your choice) as a Gradle project. After that, you will need to run the gradle build and run the SQLs from the Docs / SQLs / 1.0.0 folder in your local PostgreSQL installation.

The platform allows the user to setup executors for the following types of computational models:

• WORKFLOW

• EXECUTABLE

• COMMAND

• WEB SERVICE

The platform allows the user to setup execution environments of the following types:

• CLOUD (AWS)

• CLUSTER (LNCC)

• SSH

There is also a possibility to setup VPN conection for these environemnts:

• VPN

• VPNC (Cisco VPN)

The API accepts filters, sorting, aggregation functions, grouping and field projection. Another important aspect is that the API will only answer success if the correct credentials are present in the header of the request. Every request (except for the login endpoint and some other specific endpoints) must contain the Authorization header with the value Bearer {AUTENTICATION_TOKEN}. An authentication token can be obtained by making a request to /login endpoint twith the correct credentials. Here is an example of a possible filter operation:

localhost:9500/v1/computational_models?count=[name,currentVersion]&sort=[creationDate=asc]&groupBy=[creationDate]&filter=[currentVersion>1.0]

The available options of filters to be applied:

• Equals: "=eq=" or "=" (may be used to compare if value is equal to null)

• Less than or equal: "=le=" or "<="

• Greater than or equal: "=ge=" or ">="

• Greater than: "=gt=" or ">"

• Less than: "=lt=" or "<"

• Not equal: "=ne=" or "!=" (may be used to compare if the value is other than null)

• In: "=in="

• Out: "=out="

• Like: "=like="

Logical operators in the url:

• AND: "_and_" or just ";"
• OR: "_or_" or just ","

The Projection follows the following syntax, and the json response will only have with the specified fields:

projection = [field1, field2, field3...]

The Sorting follows the following syntax (where sortOrder may be asc or desc):

sort = [field1 = sortOrder, field2 = sortOrder...]

GroupBy follows the following syntax (groupBy does not accept projections parameters and is expected to be used along with an aggregation function):

groupBy = [field1, field2, field3...]

It performs Sum function in the specified fields, and follows the following syntax:

sum = [field1, field2, field3...]

It performs function of Avg in the specified fields, and follows the following syntax:

avg = [field1, field2, field3...]

It performs Count function in the specified fields, and follows the following syntax:

count = [field1, field2, field3...]

It performs Count Distinct function in the specified fields, and follows the following syntax:

count_distinct = [field1, field2, field3...]

• offset (DEFAULT_OFFSET = 0)
• limit (DEFAULT_LIMIT = 20 and MAX_LIMIT = 100)

Service responsible for processing message broker RabbitMq messages. These messages dictate whether a computational model will initiate an execution or abort an execution of a model executor/data extractor. The tasks may be manageble throught the platform (the executor, extractors and the environment in which the task/workflow/command will execute). There is also a cron task that verifies the status of models executing in Clusters.

Front-end that consumes exposed API resources. The front-end design was written using angular 1. To install the project, simply run the commands in the front-end project directory:

• npm install
• bower install

Finally, just run the gulp server command and access the url:

localhost: 9502

Gulp is a task automator that performs some tasks on the front-end project.

• gulp server: runs the server locally, performing the build for the dist folder;
• gulp build: performs the build of the static files for the dist folder, handling the sass files and all the plugins and dependencies;
• gulp build-production: performs the same procedure as above, but running javascripts uglify;