In LOGS, datasets and attachments are not permanently deleted immediately. Instead, a controlled, multi-stage deletion and cleanup mechanism is used to protect against accidental data loss. 

Dataset Deletion Workflow 

Before a dataset can be permanently removed, it must first be moved to the Trash.

Move to Trash

• Any user with write permissions to a dataset can move it to the trash.
• Trashed datasets are removed from searches and suggestions.
• Files associated with the dataset are not deleted immediately.

Permanent Deletion from Trash

• Only global administrators can permanently delete datasets from the trash.
• The automatic cleanup process applies only to datasets that have been permanently deleted from the trash.

Automatic Cleanup

• After a dataset is permanently deleted, its files are moved to a separate storage location.
• This relocation process runs once every 24 hours.
• Moved files remain in this isolated location for an additional 30 days.
• During this period the data files can be manually restored by a admin with access to the actual data storage. The original metada of the dataset however is lost.
• After the 30‑day retention period, the files are permanently deleted from the server.

Attachment Cleanup

Attachments undergo dedicated service scans for unreferenced attachments. Such attachments are first marked as deleted.
They then follow the same cleanup process as dataset files:

• 24-hour protection period
• File relocation
• Final deletion