Current State

Customers cannot rewind the Jira connector cursor themselves for a specific project.

When incremental ingestion misses data, we need to contact Fivetran Support to rewind the cursor for the affected Jira project.

Using a full table re-sync is not an equivalent workaround, as it can reprocess significantly more data than necessary, increase MAR consumption and cost, and take hours or days for large tables.

Desired State

Provide a self-service option to rewind the Jira cursor for a specific project to a chosen date or timestamp, through:

• Fivetran UI
• Fivetran API
• Ideally both

This would allow customers to re-ingest only the required time range instead of performing a full historical re-sync.

Nice to have: Display the current Jira cursor timestamp for each project in the Fivetran UI, making it easier to identify and troubleshoot ingestion gaps.

Why This Is Useful for Everyone

For customers:

• Faster recovery from data-integrity issues
• Lower MAR consumption compared with full table re-syncs
• Less dependency on Fivetran Support

For Fivetran:

• Fewer support requests for manual cursor rewinds
• Less operational work for Support
• A safer and more targeted recovery mechanism for Jira connectors