Salesforce Connection FAQ
What happens if source system columns are changed? There is a new extraction option to ignore metadata changes. Selecting it, new columns will be added to the extraction and deleted columns will be filled with null values. The extraction will not fail.
I receive an "URI Too Long" error message. What does it mean? That means that the table contains too many columns (usually 100+) for the request to be completed. To resolve that, unselect irrelevant columns in the table configuration.
How does Celonis ensure that extractions do not have a negative impact on the productive Salesforce system performance? Salesforce ensures that by enforcing API limits that Celonis extractions are bound to.
Why does the connection using stored credentials fail, even though I'm using the correct username/password? Common reason: for stored credentials authentication, a security token has to be pasted right after the password (without spaces). For example, enter 'yourpasswordyourtoken' in the connection configuration's password field. (New security token can be generated in the user's Salesforce settings >> Reset My Security Token.)
How is Salesforce data secured during extraction? Firstly, Celonis performs read-only operations on your Salesforce data (updates, deletions, etc. are not performed). Secondly, data transfer from Salesforce to Celonis is secured through HTTPS, ensuring an encrypted information exchange. Additionally, numerous certifications and white papers in the Celonis Security and Trust Center confirm/describe how Celonis ensures highest data security.
How does Celonis protect stored Salesforce passwords? Stored passwords are encrypted in the App DB.
In summer 2022 Salesforce will deprecate and no longer support Salesforce Platform API versions < v30.0: How does this impact my existing Salesforce Extractions? More information on the deprecated versions can be found here. The specific API version that you are using depends on your current Salesforce instance. Our Extractor always uses the latest version from the available ones in the instance. We are working on extending the extractor logs to make it visible which API version is being used.