SQL Server Data Recovery
Direct answer
Microsoft SQL Server fails from corrupted MDF/LDF/NDF files, SUSPECT or EMERGENCY databases, growing page errors (824, 823, 829), truncated transaction log and bad updates. HD Doctor recovers 91% of SQL Server cases via technical page repair, table extraction with MDF parser and log reconstruction. With 24+ years and 380+ SQL Server cases solved.
Critical: do NOT run DBCC CHECKDB with REPAIR_ALLOW_DATA_LOSS without backup, do NOT DETACH/ATTACH a SUSPECT database, do NOT truncate log with database in alert.
How SQL Server organizes data
SQL Server uses MDF (Master Data File, primary), NDF (secondary, optional for partitioning) and LDF (Log Data File, transaction log). 8 KB pages with GAM/SGAM/PFS/IAM headers. Transaction log uses Full, Bulk-logged or Simple model. Common failures: page tear, bad block in data file, corrupt log after abrupt shutdown and instances in SUSPECT after I/O error.
Common SQL Server symptoms
- Database in SUSPECT, EMERGENCY or RECOVERY_PENDING state
- Error 824/823/829 (page consistency errors)
- DBCC CHECKDB reports multiple errors
- Error 5172 (header file is incorrect) on attach
- MDF doesn't open with "Cannot open database"
- Transaction log grew until disk full and stuck
- BACKUP database fails with "page not found"
- Application reports growing deadlocks and slow queries
Most frequent SQL Server causes
| Cause | % | Recoverable? |
|---|---|---|
| Page corruption (824, 823 errors) | 30% | ✅ Yes, page repair |
| Database SUSPECT after I/O error | 22% | ✅ Yes, EMERGENCY mode + repair |
| Corrupted transaction log | 18% | ✅ Yes, log reconstruction |
| Storage failure (RAID/SAN under MDF) | 12% | ✅ Yes, storage recovery first |
| Header file incorrect (5172) | 10% | ✅ Yes, MDF header repair |
| Bad SQL Server update | 5% | ✅ Yes, data in separate files |
| Other (accidental delete, drop) | 3% | ✅ Yes, file carving |
Source: HD Doctor internal stats on 380 SQL Server cases between 2022 and 2025.
What NOT to do with a failing SQL Server
- 1.DO NOT run DBCC CHECKDB with REPAIR_ALLOW_DATA_LOSS without backup. This mode wipes corrupted pages even if data is still extractable.
- 2.Do not DETACH SUSPECT database. SUSPECT database may not reattach, causing complete loss.
- 3.Do not truncate transaction log with database in alert. Truncate before checkpoint can make recovery impossible.
- 4.Do not run RESTORE with TRUNCATE_ONLY. Deprecated command that destroys the log.
- 5.Do not replace MDF with old backup without extracting new data. Restore overwrite loses post-last-backup transactions.
- 6.Do not run shrink on corrupted database. Shrink reorganizes pages and can amplify corruption.
How HD Doctor recovers SQL Server
We work on copies of MDF/LDF/NDF, never the original. We repair pages via technical parser or extract tables individually.
- 1
MDF/LDF/NDF file intake
You send database files or server drives. For TDE encryption, specific instructions.
- 2
Diagnosis within 24h
Page analysis, headers, SQL Server version identification and corruption type.
- 3
Free written quote with scope
Technical analysis before approval, listing viable tables and estimating loss.
- 4
Technical page repair
For page corruption, we use proprietary parser that rebuilds GAM/SGAM/IAM/PFS.
- 5
Extraction via MDF parser
When database can't be attached, we extract tables individually analyzing native MDF format.
- 6
Transaction log reconstruction
For corrupted LDF, we create new log and mark last stable checkpoint.
- 7
EMERGENCY mode + DBCC
When viable, we set EMERGENCY, backup and run DBCC CHECKDB with controlled scope.
- 8
Extracted data validation
We compare row counts, referential integrity and checksums with test instance.
- 9
Delivery + final report
Database restored or tables in BAK/SQL/CSV format, signed engineer report.
Turnaround and SLA
| Scenario | Turnaround |
|---|---|
| SUSPECT database (1 MDF) | 5–10 business days |
| Page corruption without physical damage | 7–12 business days |
| Storage failure + MDF recovery | 12–22 business days |
| Encrypted database (TDE) | +5–10 business days for analysis |
- 24h emergency SLA available for production SQL Server.
- No Data, No Charge policy: if we can't recover the critical tables you flagged, you don't pay for the service. Diagnosis is free in 92% of cases.
Versions and environments supported
We service SQL Server 2005, 2008, 2008 R2, 2012, 2014, 2016, 2017, 2019, 2022, 2025. Editions: Express, Web, Standard, Enterprise, Developer. Azure SQL Database (snapshots), Azure SQL Managed Instance. Types: row-store, columnstore, in-memory OLTP. We support TDE, Always Encrypted, AlwaysOn Availability Groups, Failover Cluster.
Why HD Doctor for SQL Server
- 🏛️24+ years focused exclusively on data recovery
- 🔬Class 100 cleanroom + in-house SQL Server infrastructure
- 🧠Native MDF parser + page repair + log reconstruction
- ⚡24h emergency SLA for production SQL Server
- 🤝Only Western Digital Platinum Partner with a regional lab
- ⚖️Signed engineer report valid for forensics and insurance
SQL Server FAQ
Database in SUSPECT, still recoverable?
Yes, in 90% of cases. SUSPECT is protection state when SQL Server detects corruption. We set EMERGENCY, extract tables via parser and rebuild.
Error 823/824 in production. What to do?
Page corruption errors. Stop the application, backup MDF/LDF and send for analysis. Do NOT run DBCC CHECKDB with REPAIR_ALLOW_DATA_LOSS before diagnosis.
Lost LDF. Recoverable with only MDF?
Yes, in 85% of cases. Without LDF, database has no transaction log but we extract data. There may be loss of uncommitted transactions.
Database encrypted with TDE. Can you?
Yes, with master key and TDE certificate. Without these keys, the MDF is mathematically unrecoverable.
Server storage caught fire. Any chance?
Yes. We recover the physical storage drives first (RAID/SAN), then extract MDF/LDF files and process.
How does the quote work?
Diagnosis is free. After technical assessment within 24h we send a detailed quote.