release: bump version to 0.68.1

docs/technical/en/backend-collectors.md

@@ -77,6 +77,8 @@ Manual trigger, data clearing, and cache clearing now enter the PostgreSQL data
 
 Data deletion runs in batches so AIS-scale tables are not locked by one huge statement. A `clear_data` job clears `collected_data`, then source-specific AIS derived tables, and broadcasts `records_processed` as it goes; the console queue renders only user-facing text such as `Deleting data` and `Delete complete`, while internal table names remain in logs and raw task details. After AIS cleanup, the backend runs `ANALYZE ais_raw_observations` so datasource-list estimates converge quickly. The datasource directory uses PostgreSQL statistics for AIS record counts by default to avoid a cold-start `count(*)`; opening a single datasource detail row requests the exact count for that source.
 
+The CelesTrak TLE collector prefers the complete `active` catalog. If CelesTrak returns the "GP data has not updated" HTTP 403, the backend first reuses the active raw download cache under `$PLANET_CACHE_DIR/downloads/celestrak`; if that cache is missing, it enters fallback group mode. Fallback group mode uses only currently valid public CelesTrak groups, including `starlink`, `gps-ops`, `galileo`, `glo-ops`, `beidou`, `geo`, `iridium-next`, `stations`, `visual`, `weather`, `science`, `cubesat`, `amateur`, and `last-30-days`. Disaster-recovery fallback prefers valid local group caches before touching the network, so small-group update windows or unreliable HEAD metadata do not incorrectly fail recovery. Console `Clear Data` and `Clear Cache` jobs only touch database rows, Earth layer cache, and dashboard cache; they do not remove this raw download cache.
+
 ## III. Collector List
 
 | Collector | Data type | Content | Frequency |