Operations are callable units of work. Operations can be written in Python, SQL, and JavaScript and they may be private or called by other operations within the application. Here's a quick look at basic usage examples for Python and SQL.

Python Operations#

Python in Transposit is helpful when you need to encode business logic, make API requests, run queries, or quickly transform the data resulting from one of your queries.

A basic example of a Python operation:

def run(params):
    api.log("Hello World")
    return {
        "mission": "complete"
    }

To run another operation from inside a Python operation, use the run method. The first argument is the operation reference; the second argument is a map of operation parameters to run with:

api.run("this.operation_name", {"parameter_name": parameter_value})

To run a SQL statement inside a Python operation, use the query method:

api.query("select * from connection.operation")

To print to the console, use api.log:

api.log("hello world")

Learn more in the Python operations reference.

SQL Operations#

SQL is designed to easily join, filter, and organize your data in the way you need it. It abstracts away the details of data representation and allows you, the developer, to express your intent, while the relational engine translates that intent, optimizes it, and executes it. In this way, SQL offers a simpler way of stating this intent than you would ever get writing your own custom code.

A primary use of SQL operations is querying data. Here's an example of a basic SELECT statement:

SELECT * FROM database.get_users WHERE lastName='smith' LIMIT 50

Data can be joined, as you'd expect in SQL:

SELECT * FROM database.get_users
AS A JOIN database.get_user AS B ON A.id = B.id
WHERE A.lastName='smith' LIMIT 50

Creating SQL operations from a template (choose SQL from template in the menu) starts you off with scaffolded code based on the syntax and parameters of a given data connection operation.

Learn more in the SQL operations reference.

Usage limits#

Transposit enforces usage limits. This helps to protect our infrastructure as well as your API rate limits.

A request may consist of many operations. A request starts when a developer runs an operation in the console, a scheduled task is started or an endpoint is accessed.

Console limits#

Currently Transposit does not support multiple users modifying an application in the console simultaneously. Doing so will cause unexpected behavior and lost code.

If you need to have more than one user modify an application in the console, coordinate to make sure that each person accesses the console in a sequential fashion.

Request limits#

Memory#

• Each request is given a quota of 1 GB of memory allocated.
• This is not configurable by the developer.

CPU time#

• Each request is given a quota of 1 minute of CPU time.
• This is not configurable by the developer.

Execution timeout#

• Each request is given a quota of 1 minute of execution time.
• This is configurable by the developer, under Code > Operations > Properties. If you configure this timeout, it only applies when this operation is the first operation to run.
• The maximum value is 5 minutes.
• If operations are running in parallel (for example, if you use runBulk), you'll use CPU time more quickly than execution time; if you use timing events like setTimeout, you'll likely use CPU time less quickly than execution time.

Operations#

• Each request is given a quota of 1000 operations calls.
• Each call to an operation counts against this limit.
• This is not configurable by the developer, but inlining operations is a workaround.

API call limits#

To make sure you don't accidentally run through your API rate limits for your data connections, Transposit limits the number of API calls.

• By default, the limit is 500 per data connection per request.
• This can be changed for a particular data connector by the data connector's author.
• This can be further overridden in your application in the settings for the data connection. Navigate to Data connection > Configure, and edit the value under Max API calls.