This issue is resolved. All affected tables have been repartitioned with the correct configuration, and data pulls are working normally again.
What you need to do: Re-run your ingestion jobs to catch up on data since October 15.
We apologize that this took longer to fix than we first estimated, and for the disruption to your data pipelines.
If you still see errors on any of these tables, please contact Support and include the table name and the error message.
Posted Oct 16, 2026 - 14:06 PDT
Update
Repartitioning the affected tables is taking longer than expected. The work is still in progress with our engineers. We’ll post another update by then, or sooner if anything changes.
Until the fix is finished, data reads on the affected tables will keep failing. Please hold off on re-running ingestion for these tables until we confirm the fix is complete. We’re sorry for the extended disruption.
Posted Oct 16, 2026 - 10:36 PDT
Identified
We’ve found the root cause of todays incident. Our partition scheme was updated for a few tables which prevented successful connection when attempting to refresh your data. That’s why the service returned invalid responses when those tables were read.
We’re now repartitioning the affected tables with the correct configuration. We expect the fix to be finished in about 2 hours.
You don’t need to do anything for now. Once the fix is in place, please re-run your process. We’ll post another update when the work is done.
Posted Oct 16, 2026 - 08:05 PDT
Investigating
We’re looking into an issue where queries against 37 Data Connect tables fail with this error: "DS_INVALID_RESPONSE_FROM_DS_SERVER: The http response returned invalid json"
Impact: -Data reads on the 37 tables listed below fail. Schema and metadata queries (for example, information_schema) still work. -All other tables in the share are working normally. -These tables were last ingested successfully on October 15, 2026. Data from that load and earlier is unaffected, but no new data can be pulled from these tables until this is fixed.
Our engineering team is investigating at this time and we will follow up once this issue is resolved. We’re sorry for the disruption this may cause to your downstream processes.
Posted Oct 16, 2026 - 05:43 PDT
This incident affected: Data and Partner Connect (Data Connect Operational Status).