DropScan: decoding does not prove attribution
Physical scans can find pages and schema candidates, but tablespace IDs, index IDs, table names and decodable DDL may recur across instances or object generations.
DropScan records page sources, object generations, conflicts and possible schema matches before recovering confirmed objects. Missing pages and external-value files are recorded separately. Parsing some records does not mean that all data affected by the incident has been recovered.
Continue with PDU
See PDU's features, operating documentation and real recovery cases.