When using the scan key emitter, if redis-copy got disconnected or is killed abrubtply, it would be useful if it will return the cursor returned by the last SCAN operation and can be passed later as a parameter. This will be useful on subsequent attempts to avoid redoing DUMPs and RESTOREs from scratch (and displaying lots of FAIL messages).
Pull-request welcome. The REDIS_COPY_SCAN_CURSOR Environment variable would be a good fit (as would logging the cursor somewhere, or putting the last-started one in a temporary file somewhere)
When using the scan key emitter, if redis-copy got disconnected or is killed abrubtply, it would be useful if it will return the cursor returned by the last
SCAN
operation and can be passed later as a parameter. This will be useful on subsequent attempts to avoid redoingDUMP
s andRESTORE
s from scratch (and displaying lots of FAIL messages).