DBRecover for Oracle 2605 GUI User Guide: C6 Validation Walkthrough

1. Scope and Operating Model

C6X260516 is a reference validation schema rather than a small training dataset. It is designed to exercise the same GUI recovery paths customers use in real Oracle incidents. A useful walkthrough verifies that DBRecover unloads real data, not just object names.

The C6 package contains two layers:

Current validated state

Official-guide coverage target

This C6 guide follows the same customer-facing scope as the public DBRecover for Oracle user guide: installation and startup, license registration, recovery wizard loading, dictionary and no-dictionary workflows, CSV export, Data Bridge, DDL export, diagnostics, Dump Reader, ASM file handling, and real-world recovery scenarios for corruption, missing SYSTEM, ransomware or damaged files, DELETE, TRUNCATE, DROP TABLE, and DROP TABLESPACE.

2. C6 Sample Inventory

In this walkthrough, the C6 sample directory is shown as:

D:\dbrecover-samples\c6x260516_oo19c_v1

The sample was generated on Oracle 19c OO19C with schema C6X260516 and tablespace C6X260516_TS.

Required directory layout

Clean DBF files

Dump files

Manifest and proof files

3. Install and Start DBRecover

DBRecover for Oracle is portable software. Customer packages include launch scripts for the target platform. Development helper scripts are not part of the customer workflow.

Windows startup

1. Unzip the DBRecover package to a local directory, for example D:\DBRecover\dbrecover-for-oracle-2605.
2. Open the unzipped directory.
3. Double-click dbrecover.bat.
4. If Windows blocks the package, unblock the ZIP before extracting or allow the signed package according to your endpoint policy.
5. For large C6 testing, add the DBRecover program directory and the C6 sample directory to antivirus exclusions if your security policy allows it.

Linux startup

1. Unzip the Linux package on a workstation or recovery server with a graphical desktop.
2. Run chmod +x dbrecover.sh.
3. Run ./dbrecover.sh.
4. If you connect remotely, verify X11, VNC, or another graphical desktop before starting the GUI.

UNIX startup

For AIX, HP-UX, Solaris, and other UNIX-like environments, install a compatible Java runtime and start the JAR from a graphical desktop session:

java -jar dbrecover-for-oracle.jar

Main menu orientation

License registration

1. Open Options and choose Register.
2. Enter or confirm the Oracle DB_NAME that the license was issued for.
3. Paste the license key and click Register.
4. Open Options and choose About to confirm registration details.

4. Load the Clean C6 DBFs in Dictionary Mode

Use dictionary mode when SYSTEM01.DBF, SYSAUX01.DBF, and the target user datafiles are available. This is the primary C6 workflow.

Step 1: open Recovery Wizard

1. Start DBRecover.
2. Open Tools.
3. Select Recovery Wizard.
4. Read the welcome page and click Next.

Step 2: choose dictionary mode

1. Select the option for recovery with SYSTEM tablespace.
2. Use ASM options only for ASM-sourced files.
3. Use no-SYSTEM mode only for the dedicated dbf_system_missing scenario.
4. Click Next.

Step 3: confirm database options

1. Keep Endian as Little Endian for this Oracle 19c Linux x86-64 sample.
2. Use automatic dictionary character-set detection. C6 was generated as AL32UTF8 and DBRecover maps it to UTF-8 internally.
3. Keep Block Size as 8192.
4. Keep Offset as 0 for normal DBF files.
5. Keep database version on automatic detection unless support gives a specific override.
6. Click Next.

Step 4: add the clean C6 files

1. Click Choose Files.
2. Select files from D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean.
3. Include SYSTEM01.DBF, SYSAUX01.DBF, UNDOTBS01.DBF, C6X260516_TS01.DBF, C6X260516_TS02.DBF, and C6X260516_DROP_TS01.DBF.
4. Keep block size 8192 and offset 0 for each DBF unless a header-specific investigation requires manual values.
5. Click Load.

External file list option

For repeatable validation, prepare a text file and use the external file-list loader. Keep one file per line:

"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\SYSTEM01.DBF" 8192 0
"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\SYSAUX01.DBF" 8192 0
"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\UNDOTBS01.DBF" 8192 0
"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\C6X260516_TS01.DBF" 8192 0
"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\C6X260516_TS02.DBF" 8192 0
"D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_clean\C6X260516_DROP_TS01.DBF" 8192 0

Load completion checks

5. Browse the Loaded C6 Schema

After the wizard finishes, expand the database tree, then expand Users, then C6X260516. The left tree should expose the C6 schema and its object folders.

Expected object surface

Tables that must be visible

At a minimum, confirm these named tables or expected baseline tables in the tree or table-search dialog:

Schema and table context menus

Right-click the schema node for schema-level actions. Right-click a table for table-level actions.

6. Export Recovered Data to CSV

CSV export is a primary customer workflow and a C6 validation gate. Do not validate CSV export only by checking that a file was created. Validate row counts, quoting, line breaks, dates, numbers, NULLs, and special table structures.

Single-table CSV export

1. Load the clean C6 database in dictionary mode.
2. Expand C6X260516 and right-click C6_CSV_QUOTING_EDGE.
3. Select the table-level CSV export action.
4. Choose an output directory outside the DBF sample tree, for example D:\dbrecover-output\c6-csv.
5. Start the export and wait for the progress dialog to finish.
6. Open the generated CSV in a text editor that can display CRLF and quotes.
7. Confirm row count 64.

CSV quoting rules to verify

Partitioned table export

1. Right-click C6_PARTITION_RANGE_CASE.
2. Export it to CSV.
3. Compare the output to COUNT(*) = 12000 and SUM(amount) = 226098840.00.
4. Confirm rows from all partitions are present. A pass that only exports one partition is invalid.

IOT export

1. Right-click C6_IOT_EDGE.
2. Export the table to CSV.
3. Compare the output to COUNT(*) = 4096 and SUM(balance) = 93220188.16.
4. Confirm the row order and key columns are stable enough for customer review.

Large payload export

C6 contains two large payload tables to prove that the 32 GB files contain real recoverable rows:

CSV pass and fail rules

7. BasicFile and SecureFile LOB Validation

LOB testing is one of the main reasons C6 exists. A LOB result is valid only when DBRecover materializes the actual bytes or text. A locator string, a preview-only value, or a zero-byte output file is a failure.

LOB objects

GUI steps

1. Load the clean C6 DBF set.
2. Right-click C6_LOB_BASICFILE_MEDIA and export the table.
3. Repeat for C6_LOB_SECUREFILE_MEDIA.
4. Inspect the output directory for materialized BLOB files or the configured LOB output format.
5. Record actual BLOB byte lengths for each row.
6. Confirm the largest BasicFile and SecureFile BLOB outputs are at least 67108864 bytes.
7. Open the CLOB and NCLOB output and confirm UTF-8 text is not garbled.

Invalid LOB results

Damaged LOB block scenario

The controlled damaged LOB file is:

D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_damaged\c6x260516_ts02_lob_block_flip.dbf

The damage generation log records a modified SecureFile BLOB extent from C6_LOB_SECUREFILE_MEDIA.BLOB_PAYLOAD. In the damaged-file workflow, DBRecover must report the affected LOB accurately. It must not silently report all LOBs as recovered when one known LOB block was modified.

8. Deleted, Truncated, Dropped, and Dropped-Tablespace Recovery

C6 includes separate expected baseline tables so that recovery tests compare actual output to known-good data. This prevents false passes caused by an empty dictionary state.

Recovery case matrix

DELETE workflow

1. Load the clean C6 files.
2. Locate C6_DELETE_CASE and C6_DELETE_EXPECTED.
3. Use the table-level deleted-row recovery option for C6_DELETE_CASE.
4. Export the recovered deleted rows.
5. Compare row count and numeric sums to C6_DELETE_EXPECTED.

TRUNCATE workflow

1. Locate C6_TRUNCATE_CASE in the loaded schema. It should be visible but empty.
2. Use the truncate recovery workflow or raw table recovery path provided by the GUI.
3. Export recovered rows.
4. Compare to C6_TRUNCATE_EXPECTED, which has 1500 rows.

DROP TABLE and DROP PURGE workflows

1. Use dropped-object recovery features to search for dropped table metadata and extents.
2. For normal drop, use C6_DROP_TABLE_EXPECTED as the comparison source.
3. For purge, use C6_DROP_PURGE_EXPECTED as the comparison source.
4. Record recovered row count, exported file size, and any unsupported metadata warning.

DROP TABLESPACE workflow

1. Load the clean C6 files and include C6X260516_DROP_TS01.DBF.
2. Use the raw datafile recovery workflow to inspect the retained dropped-tablespace file.
3. Recover data for the dropped tablespace scenario.
4. Compare recovered data to C6_DROP_TABLESPACE_EXPECTED, which has 1500 rows.

9. No-SYSTEM Manual Recovery

Use this scenario when SYSTEM01.DBF is missing, unreadable, or not provided by the customer. C6 provides a dedicated no-SYSTEM file set:

D:\dbrecover-samples\c6x260516_oo19c_v1\dbf_system_missing

GUI steps

1. Start a new recovery session.
2. Select the mode for recovery without SYSTEM tablespace.
3. Add files from dbf_system_missing. Do not add SYSTEM01.DBF.
4. Keep block size 8192 and offset 0.
5. Use manual column definition or the GUI's no-dictionary table-definition workflow.
6. Define C6_NODICT_MIXED_TYPES with its scalar, RAW, BLOB, and CLOB columns.
7. Unload the table and compare row count to 128.

What makes the test valid

• The input directory must not contain SYSTEM01.DBF.
• The GUI must not reuse dictionary metadata from a previous clean C6 session.
• The output must contain real values for scalar, RAW, BLOB, and CLOB fields.
• The final report must record C6_NODICT_MIXED_TYPES with status=PASS.

10. Damaged and Ransomware File Workflows

C6 includes controlled damage artifacts. Use only copied damaged DBFs; never damage the clean set.

Health Checker steps

1. Open the Health Checker or datafile diagnostic utility.
2. Add one damaged DBF at a time.
3. Run the check and save the diagnostic output.
4. Record whether the tool reports file header, block, checksum, or LOB-level problems.
5. Do not mix damaged files with clean files unless the scenario explicitly requires it.

11. ASM File Workflow

The public user guide includes ASM source-file handling. C6 covers that workflow through logs\c6_asm_manifest.csv, which maps ASM-style aliases to copied DBF files.

GUI steps

1. Open Tools.
2. Select ASM File Clone or the ASM copy workflow available in the build.
3. Use the ASM alias names from logs\c6_asm_manifest.csv as the source reference.
4. Copy or map the ASM files to local DBF files under the C6 sample directory.
5. Load the resulting local DBFs with Recovery Wizard.
6. Confirm the loaded schema matches the clean C6 dictionary-mode result.

Validation

The manifest gate is c6_asm_manifest_count = 2. A valid ASM test must prove that the copied DBFs can be opened and loaded, not only that alias strings exist.

12. Oracle Dump Reader for EXPDP and Classic EXP

C6 includes both export formats. Use Dump Reader to verify that DBRecover handles Data Pump and classic exp artifacts.

Files

GUI steps

1. Open the Dump Reader utility from the GUI.
2. Open D:\dbrecover-samples\c6x260516_oo19c_v1\exports\c6x260516_expdp.dmp.
3. Confirm that C6 objects are listed.
4. Check that C6_LOB_BASICFILE_MEDIA, C6_LOB_SECUREFILE_MEDIA, C6_IOT_EDGE, C6_PARTITION_RANGE_CASE, C6_BRIDGE_CHILD, and C6_CSV_QUOTING_EDGE are discoverable.
5. Repeat with c6x260516_classic_exp.dmp.
6. Record any classic exp unsupported-type warnings from c6x260516_classic_exp.log.

13. DDL Export and Object Browsing

C6 verifies that DBRecover exports schema definition details, not just table rows. The DDL scenario is anchored by C6_DDL_TABLE and related objects.

Required DDL objects

GUI steps

1. Load the clean C6 DBF set.
2. Right-click the C6X260516 schema node.
3. Choose Export DDL.
4. Select tables, indexes, constraints, views, stored procedures or functions if present, sequences, triggers, synonyms, and DB links.
5. Export DDL to an output directory.
6. Open the DDL file and search for C6_DDL_TABLE, C6_LOOPBACK, trigger definitions, materialized view definitions, and synonyms.

14. Data Bridge to a Target Oracle Database

Data Bridge validates that DBRecover can move recovered data into a target database and preserve relational data shape. C6 uses parent, child, and LOB bridge tables.

Prepare the target database

1. Use a clean target schema that is not C6X260516, for example C6_BRIDGE_TARGET.
2. Confirm target tablespace capacity. C6 LOB and large-payload tables are intentionally large.
3. Use a smaller subset first if you only need to verify connection setup.
4. For full validation, run the required C6 bridge tables and compare counts and sums.

Create a Data Bridge connection

1. Right-click the C6X260516 schema node or a specific C6 table.
2. Select Data Bridge.
3. Create or choose an Oracle connection.
4. Enter host, port, service name or SID, user, and password.
5. Test the connection and save it.

Run C6 bridge validation

1. Bridge C6_BRIDGE_PARENT, C6_BRIDGE_CHILD, and C6_BRIDGE_LOB.
2. After the load finishes, run row-count and sum checks on the target database.
3. Confirm C6_BRIDGE_CHILD has COUNT(*) = 5000 and SUM(child_amount) = 21879375.00.
4. Confirm LOB byte lengths in C6_BRIDGE_LOB match the source.
5. Record the target schema name, DBRecover build number, and validation SQL output.

SELECT COUNT(*), SUM(child_amount)
FROM C6_BRIDGE_CHILD;

15. Local Query and Diagnostic Utilities

Use local query tools to spot-check recovered content before running long exports.

Local Query

1. Open the Local Query utility from the loaded C6 session.
2. Run simple checks against visible C6 tables.
3. Compare results to logs\c6_validity_manifest.csv.

Useful C6 checks

SELECT COUNT(*), SUM(amount) FROM C6_PARTITION_RANGE_CASE;
SELECT COUNT(*), SUM(balance) FROM C6_IOT_EDGE;
SELECT COUNT(*), MAX(DBMS_LOB.GETLENGTH(blob_payload)) FROM C6_LOB_BASICFILE_MEDIA;
SELECT COUNT(*), MAX(DBMS_LOB.GETLENGTH(blob_payload)) FROM C6_LOB_SECUREFILE_MEDIA;
SELECT COUNT(*), SUM(child_amount) FROM C6_BRIDGE_CHILD;

Oracle Query utility

Use Oracle Query when a live Oracle source or target database is available. It is useful for comparing source SQL results to DBRecover output and Data Bridge output.

16. Final C6 Release Validation

C6 has two levels of automated validation. Quick mode verifies layout, manifests, and artifact presence. Full mode additionally requires actual DBRecover unload accuracy results.

Quick gate

Use quick mode after copying or refreshing the sample:

.\scripts\validation-tests\run-c6x260516-validation-test.ps1 -Mode quick -SkipHashes -SkipDumpParse -SkipDataAccuracy

The reference package passed quick validation with report:

target\c6x260516-validation-report\quick-run

Full gate

Use full mode before validation sign-off:

.\scripts\validation-tests\run-c6x260516-validation-test.ps1 -Mode full

The reference package passed full validation with report:

target\c6x260516-validation-report\full-run

Result summary:

Tests run: 9, Failures: 0, Errors: 0, Skipped: 0
BUILD SUCCESS
hashEnabled=True
dumpParseEnabled=True
dataAccuracyEnabled=True
validityManifestRequired=True

The full gate must consume DBRecover-generated output and write:

D:\dbrecover-samples\c6x260516_oo19c_v1\logs\c6_unload_accuracy_report.csv

Required unload accuracy rows

Manual sign-off checklist

• Clean C6 DBFs load successfully and show C6X260516.
• logs\c6_local_sha256_verify.csv has OK for 13 key files.
• Both C6 32 GB DBFs are present and have exact size 34,358,697,984 bytes.
• Both dump files open in Dump Reader.
• BasicFile and SecureFile 64 MB BLOB outputs are materialized bytes.
• IOT and partitioned table outputs match count and sum.
• DELETE, TRUNCATE, DROP, DROP PURGE, and DROP TABLESPACE cases compare to expected baselines.
• No-SYSTEM mode uses the dbf_system_missing directory and does not add SYSTEM01.DBF.
• Damaged-file checks use only dbf_damaged copies and do not alter dbf_clean.
• Full mode exits successfully and every required row in c6_unload_accuracy_report.csv has PASS.

17. Complete C6 Scenario Coverage Matrix

This table is generated from the current C6 manifest and serves as the checklist for this guide. A complete validation run should account for every row.

18. Troubleshooting

DBRecover starts slowly or Windows CPU is high

C6 contains multiple 28 GB to 32 GB files and the validation process reads them completely. Endpoint protection, OneDrive, Dropbox, and file indexers can rescan those files. If policy allows, exclude the C6 sample path and DBRecover output path from real-time scanning.

The C6 schema tree is empty

• Confirm SYSTEM01.DBF was included for dictionary mode.
• Confirm C6X260516_TS01.DBF and C6X260516_TS02.DBF are both present.
• Confirm block size is 8192 and offset is 0.
• Check the DBRecover log for AL32UTF8, dictionary, or file-header warnings.

A table exports zero rows

• Check whether the target is intentionally empty, such as C6_TRUNCATE_CASE.
• If it is not intentionally empty, compare against c6_validity_manifest.csv.
• For IOT or partitioned tables, verify that the selected export path supports the specific structure.

LOB columns are empty or look like locators

• Check the selected LOB export option.
• Confirm output files are materialized bytes, not locator strings.
• Compare BasicFile and SecureFile output separately.
• For full validation, record BLOB byte lengths in c6_unload_accuracy_report.csv.

DMP opens but some objects are missing

• Check whether the test uses expdp or classic exp. Classic exp may warn about unsupported Oracle types.
• Review c6x260516_expdp.log and c6x260516_classic_exp.log.
• Use c6_export_manifest.csv as the artifact checklist.

Full validation fails after quick validation passed

This is expected when unload output has not yet been generated. Quick mode proves that the sample and manifests are present. Full mode requires DBRecover output and c6_unload_accuracy_report.csv.

19. Customer or Release Handoff Package

A complete C6 handoff package for validation contains:

• The full C6 sample directory.
• This user guide.
• The C6 generation runbook and validation documentation.
• The quick or full validation report directory under target\c6x260516-validation-report\.
• DBRecover output directories used for CSV, LOB, DDL, DMP, Data Bridge, and recovery workflow checks.
• For validation sign-off, logs\c6_unload_accuracy_report.csv with PASS for every required table.