Product and documentation updates about the Salesforce integration from the team at Stitch.
Stay up-to-date by subscribing to our RSS feed.
Salesforce integration: New version (v2) now available!
A new version (v) of our integration is now available!
We’ve updated the LightningUriEvent stream to use the EventIdentifier field as the Primary Key.
Salesforce (v1) update: Increased request timeout
We’ve updated our (v) integration to increase the request timeout from 30 seconds to five minutes.
With this change, the integration can support objects that take longer than 30 seconds to return.
New extraction error reference for Salesforce
We’ve added a new resource for troubleshooting Extraction errors in integrations. If Stitch receives an error during Extraction, you can use this reference to locate the error, pinpoint the cause, and identify next steps for resolution.
New feature: Integration changelogs
Introducing: Integration changelogs!
Check out the history of our integrations and stay in the loop on updates with dedicated changelogs for these integrations:
We’ll add changelogs for other popular integrations and destinations in the weeks to come. If there’s a specific integration you’d like to see us work on, let us know by creating an issue in the Stitch Docs GitHub repo.
Salesforce (v1) update: Removed FieldHistoryArchive object
We’ve determined that the FieldHistoryArchive object is incompatible with the integration’s current method of querying, and as a result have removed it from the integration.
Refer to the tap repository for more info.
Salesforce (v1) improvement: Improved OPERATION_TOO_LARGE error messaging
We’ve updated the integration to log OPERATION_TOO_LARGE errors more clearly. The following error will now display in the Extraction Logs when this issue is encountered:
OPERATION_TOO_LARGE: exceeded 100000 distinct who/what's. Consider asking your Salesforce System Administrator to provide you with the `View All Data` profile permission. (Stream: [STREAM_NAME])
Salesforce (v1) improvement: Error message formatting
There are now actionable error messages when QUERY_TOO_COMPLICATED errors are encountered for the integration.
Salesforce (v1) integrations: Remove NetworkUserHistoryRecent object
We’ve determined that the NetworkUserHistoryRecent object is incompatible with the integration’s current method of querying, and as a result have removed it from the integration.
Refer to the tap repository for more info.
Salesforce (v1) integrations: Remove DataType object
We’ve determined that the DataType object is incompatible with the integration’s current method of querying, and as a result have removed it from the integration.
Refer to the tap repository for more info.
Salesforce (v1) integrations: LoginEvent is now Full Table
We’ve determined that the LoginEvent object doesn’t support ordering by CreatedDate, meaning that Stitch can’t reliably replicate this data incrementally. This table now has a forced Replication Method of Full Table.
Salesforce (v1) integrations: Improved PK chunking error messages
Error messages resulting from PK (Primary Key) chunking in integrations have been extended. Previously, only timeout issues were covered by the integration. This improvement adds logic that covers additional situations, such as failure to write query results.
Salesforce (v1) integrations: Location data types unsupported for Bulk API
We’ve determined that the locations fields aren’t supported by the Bulk API. If selected, the following erorr will surface during Extraction:
[FIELD_NAME] cannot query compound address fields or geolocations with bulk API
Refer to the tap repository for more info.
Salesforce (v1) integrations: Remove SiteDetail object
We’ve determined that the SiteDetail object is incompatible with the integration’s current method of querying, and as a result have removed it from the integration.
Refer to the tap repository for more info.
Salesforce (v1) integrations: Convert integer zero values to prevent schema violations
’s API occasionally returns integer fields with a value of 0.0, resulting in schema violation errors during Extraction. To prevent this error, the integration now converts 0.0 to 0 if the field can be an integer.
Salesforce (v1) integrations: Remove Announcement object
We’ve determined that the Announcement object is incompatible with the integration’s current method of querying, and as a result have removed it from the integration.
Refer to the tap repository for more info.
Salesforce (v1) integrations: Remove *ChangeEvent tables
We’ve determined that *ChangeEvent tables aren’t queryable via the REST or Bulk APIs, and have removed them from the integration. Refer to the tap repository for more info.
Salesforce (v1) integrations: Correct date chunking for Bulk API replication
For integrations using the Bulk API, when a query to the API times out, Stitch would cut the requested date range in half and re-try the request. However, after replicating the records, the integration wouldn’t attempt to replicate the second half of the original window, resulting in missed records.
Because the integration only updates Replication Keys when records are extracted, Replication Keys wouldn’t advance if records weren’t received in the first half of the date range. This meant that the integration wouldn’t ever try to replicate records in the second half of the date range.
This fix adds a check to see if the date range was chunked, and if so, queries the second half of the chunked range and extracts records, if any exist.
Salesforce (v1) integrations: BackgroundOperationResult is now Full Table
We’ve determined that the BackgroundOperationResult object doesn’t support ordering by CreatedDate, meaning that Stitch can’t reliably replicate this data incrementally. This table now has a forced Replication Method of Full Table.
Salesforce (v1) integrations: Fix for OpportunityFieldHistory parent
Previously, OpportunityFieldHistory exports failed due to a function in the integration incorrectly returning OpportunityField as the parent. This fix adds a check for the FieldHistory suffix, ensuring that the parent is correctly returned as Opportunity.
Salesforce (v1) integrations: Add suffix to parent tables ending in __
Previously, integrations attempted to perform PK (Primary Key) chunking for custom tables using an incorrect parent table name. This fix now appends c to parent tables ending in __, ensuring chunking is performed correctly.
Salesforce (v1): ContentFolderItem is no longer supported
We’ve determined that the ContentFolderItem object is incompatible with Stitch’s querying strategy, and as a result is no longer available for replication via the REST API option.
Salesforce (v1): Additional incompatible objects
We’ve found that the following objects are incompatible with Stitch’s querying strategy, and as a result are no longer available for replication:
LookedUpFromActivityAttachedContentNoteQuoteTemplateRichTextData
Salesforce (v1): Primary Key (PK) chunking
We’ve implemented ‘s Primary Key (PK) chunking feature into our integration.
If the integration is using the Bulk API and encounters a timeout error, Stitch will now re-try using the Sforce-Enable-PKChunking header in the API request. This approach ‘chunks’ data into small batches, allowing the integration to more efficiently extract large volumes of data.
Salesforce (v1): Improved record type handling
We’ve added handling for several record types:
bytecalculatedDataCategoryGroupReferencemasterRecord
Extraction logs now available!
Detailed Extraction Logs are now available for select integrations in Stitch. This new feature presents detailed information about the extraction process in these integrations, and lets you:
- Inspect, copy, and download extraction log files up to 50MB in size
- View historical logs over the past seven days
- Visualize how often extraction runs, how long it takes, and when it errors
- View logs for extraction jobs currently in progress
This feature is available for the following integrations:
We’ll be working over the coming weeks to bring these logs to the rest of our integrations. Check out the docs for more info on this new feature.
Salesforce integration: New version (v1)
A new version (v) of our integration is now available! Major changes include:
- Choose between Bulk or REST API
- Configure how much of your API quota Stitch can use
- Connect to Sandbox accounts
- Choose whether Stitch automatically replicates new fields
- Reset the Replication Key for individual tables in Stitch
- Toggle between Incremental or ull Table Replication for tables
- Access the latest objects and fields from the newest version of the API
- Removed
sf_prefix for tables for all newly created integrations
Learn more about the integration and these features in our integration documentation.