In Atlas, navigate to your cluster, open the 'Online Archive' tab, and click 'Configure Online Archive'.
Set the archival criteria: choose the date field (e.g., `createdAt`) and specify the age threshold in days after which documents are archived (e.g., documents older than 90 days).
Optionally specify a custom partition field to optimize query performance on archived data (e.g., `region`, `tenantId`).
Enable the archive and let Atlas begin the initial archival run; monitor progress in the Online Archive status panel.
Query archived and live data together via the federated endpoint — use the Atlas Data Federation connection string; queries transparently span both tiers.
Known gotchas
Archived documents are stored in Atlas-managed S3 (or your own S3 bucket if configured); queries on archived data are billed at Data Federation query rates, not standard Atlas rates.
Documents cannot be written back to the live collection from the archive automatically — restoration requires exporting and re-inserting.
The archival date field must be a BSON Date type; string-formatted dates are not supported as the archival criteria field.
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp