Operations¶
This page covers the day-to-day commands you use after sync is enabled.
Status¶
Use this to confirm that the replica is enabled, the replica ID is set, and the journal is growing.
Doctor¶
sync doctor combines:
- journal integrity
- retention safety
- peer lag
- unresolved conflicts
- recent session summaries
- guidance text
Use it before prune or before upgrading a schema.
Peer Lag¶
The sync doctor and sys_sync_peer_lag surfaces show inbound and outbound lag per peer. If lag grows, inspect the latest sessions and the peer binding.
Retention¶
Use dry run first:
If the safe watermark is lower than the requested prune point, the command fails unless --allow-data-loss is set.
--allow-data-loss is a deliberate override. It should be used only when you have another copy of the data or you are intentionally discarding history.
Maintenance Routine¶
Recommended order for a healthy replica:
sync statussync pendingsync conflictssync doctorsync prune --dry-runsync pruneonly when the dry run is safe
Output Shapes¶
sync statusis a compact key-value table or JSON object.sync doctoris a report object in JSON or a human table with extra sections.sync prunereturnsrequested_through,effective_through,pruned,dry_run,allow_data_loss, andblocked_by_json.