When the instance won't start.
Pull intact rows from a crashed, corrupted or powered-off database — and prove their integrity. Post-mortem extraction with built-in integrity evidence.
//What is verified here
- Skip damaged blocks (
--skip-corrupt/--max-corrupt) — recovered, not guessed GA - Recover deleted rows (
--include-deleted,_DELETEDcolumn) GA - Read straight from ASM raw devices and offline volumes — disk scanning BETA · file extraction EXP
- Per-file / per-LOB SHA-256 sealing +
manifest.json, with compliance mapping GA - Consistency certificate: Stop-SCN + fuzzy-bit per datafile, tiered verdicts (T3-CLEAN / T2-COMMITTED), chain-of-custody GA
The SHA-256 sealing works at file level (audit report) and at LOB side-file level (manifest.json) — not per individual table row. For chain-of-custody that means: every delivered file and every extracted document is hash-sealed and reproducible. It is precisely per-file / per-LOB SHA-256 sealing + manifest, not "per-record".
//Step by step
Read straight from ASM raw devices / offline volumes
reclaimdb asm-scan /dev/oracleasm/disks/* reclaimdb asm-extract /dev/oracleasm/disks/* -o /tmp/extracted/ reclaimdb restore -m meta.json --csv ./out/ /tmp/extracted/*.dbf
Skip damaged blocks instead of aborting
A normal import would refuse; reclaimDB logs broken blocks and pulls out the intact rows.
reclaimdb --skip-corrupt restore -m meta.json --csv ./out/ /data/*.dbf # With an abort threshold for audits: reclaimdb --max-corrupt 50 restore -m meta.json --csv ./out/ /data/*.dbf
Carry deleted rows forensically
reclaimdb --include-deleted restore -m meta.json --csv ./out/ /data/*.dbf
Recoverable deleted rows appear with an extra _DELETED column.
Produce integrity evidence (audit report)
The audit command produces a sealed report with SHA-256 per file (plus SHA-256 per LOB side-file in manifest.json) and compliance mapping (SOX/PCI/ISO/BSI/NIST) as JSON + print-ready HTML. For point-in-time consistency evidence, --consistency-cert adds Stop-SCN + fuzzy-bit per datafile with tiered verdicts and a chain-of-custody record (scan window, size/mtime, optional SHA-256).
reclaimdb audit -m meta.json --report-dir /var/audit/ /backup/*.dbf
//Relevant exit codes
Test this case on your data
Send us the kind of source you actually have — a .dmp, a .dbf, an RMAN backup piece or a storage snapshot. You get a signed evaluation binary, a sample report, and an honest assessment of which paths apply to your case.