If your query is matching data within last 24 hours - your snapshot will start processing/delivering immediately.
If some of your matched data is older than 24 hours - it needs to be retrieved from S3 Glacier Deep Archive storage tier before delivery, which may take up to 72 hours.
Avoid queries that span the retention boundary (approximately 24 hours from now).Requests with max_age or time ranges that fall within ~24h ± 2h of the current time may include files that have already been migrated to archive storage tier. Attempting a dump for such queries can cause the dump to stall or remain incomplete because of files storage class transition.
Recommendations:
- We recommend using
max_age = 12h for initial testing to ensure fast delivery.
- For real-time data needs: use
max_age: "12h" or a narrower window to avoid the retention edge.
- For historical data (older than 24h): use explicit
min_date/max_date filters rather than max_age.
- If a dump appears stalled: we usually retry automatically, please open a ticket if it didn’t happen.
