Simple NodeJS utility that fetches Redash query results via API and upserts the data to desired Salesforce objects using the Salesforce REST Bulk API, leveraging the lightweight node-salesforce-connection package for session authentication and REST request abstraction.
Robust logging with winston.
$ git clone git@github.com:codygordon/redash-to-salesforce.git
$ cd redash-to-salesforce
$ npm i
First, copy .env.example
to a new .env
(make sure to leave .env.example
intact) and enter environmental variables (all are required for script to run).
$ cp .env.example .env
Env Variable | Value |
---|---|
SALESFORCE_HOSTNAME |
test.salesforce.com or login.salesforce.com or yourdomain.salesforce.com |
SALESFORCE_API_VERSION |
tested with 41 |
SALESFORCE_USERNAME |
example@example.com |
SALESFORCE_PASSWORD |
(make sure to update if changed) |
SALESFORCE_SECURITY_TOKEN |
Reset Salesforce Security Token |
REDASH_BASE_URL |
https://redash.yourdomain.com |
REDASH_API_KEY |
(find ReDash API key @ {redash_url}/users/{your_user_id}#apiKey) |
LOGS_PATH |
(log files in production will be saved in specified folder path) |
Second, edit lib\config.js
and enter objects into the config
constant (array) for each ReDash query / Salesforce object being processed, for example:
const config = [
{
redashQueryId: 55,
salesforceObject: 'contact',
salesforceExtId: 'your_user_id__c',
},
{
redashQueryId: 56,
salesforceObject: 'opportunity',
salesforceExtId: 'your_opportunity_id__c',
salesforceLookups: [{
relationship: 'contact__r',
extId: 'your_user_id__c',
}],
},
]
Each upsert process will be handled synchronously in the config
array's indexed order allowing for lookups to be handled properly (i.e., a Contact must exist before a related Opportunity can reference it), so make sure the objects are in the order you want them to run.
You should set up a dedicated API User to use with this script.
For each Salesforce object with data being upserted into there must be a separate ReDash query with appropriate columns and data formatting.
The script will fail if the following conditions aren't met:
- All columns in query must exist as fields in Salesforce object
- ReDash query column must match existing Salesforce field names exactly
- Exception: Luckily, Salesforce field names are case insensitive when using the API, as the ReDash query API only exports field names in lowercase
- Custom fields must end with
__c
(this will not display in the #table query results, but it will still export correctly when using the API)
- Date columns in query must be in the format of
YYYY-MM-DD
- For lookups the script will convert the query column into the required nested JSON format
- The column name and the config's
salesforceLookups.relationship
value must both exactly match the relationship name (i.e.,Contact__r
) - The column values must contain the related object's external ID that matches the config's
salesforceLookups.extId
- The column name and the config's
Webpack is used to build into single production dist file:
$ npm run build
You may start the script in development
mode (which uses nodemon and will log to the console) by running:
$ npm start
You can either have an external service such as crontab
run the built script at redash-to-salesforce\dist\index.js
, or run:
$ npm run prod
In production
mode a log file will be created in the specified directory each time the script is run and no logs will be sent to the console.
- Write test suites in Mocha