SQL databases typically contain two types of data files, i.e. primary and secondary data file. The primary data (or .mdf) file is the beginning of your database and has links to other files in the database. All data files in the database other than the primary data file are known as the secondary data files. They have ‘.ndf’ filename extension. However, corruption may occur in any of these data files, causing much downtime and in some cases, complete database inaccessibility.
SQL database corruption may result due to an instant power failure in the middle of a transaction or due to a hardware malfunction in the disk subsystem. Predominantly, hardware malfunction is the major cause of corruption. Other reasons that can incur corruption in the SQL database are database duplication, creating triggers on system tables, renaming the primary file group, etc. This corruption may remain obscure till the time you do not interact with the corrupt data. If corruption affects the backup files, you can never restore the database to its previous consistent state.
Some error messages received frequently in the event of SQL database corruption are listed below:
- Table error: Object ID 0, index ID 0, and page ID (1:105). The PageId in the page header = (0:0)
- Internal error. buffer provided to read column value is too small
- Conflict occurred in database’db_name’, table ‘table_name’, column ‘column_name’
- Memory or buffer error, space provided to read column is too small
To get away with these corruption problems, you should always have a good backup strategy in the first place. Using the most recent backup, you can perform MDF recovery and retrieve all your invaluable data stored in the database. However, it is usually experienced that maintaining the most updated backup is not always possible.
In such circumstances, you should use the DBCC CHECKDB command-line tool to repair data corruption. To check the physical consistency of the database, run the following command:
DBCC CHECKDB('neo') WITH PHYSICAL_ONLY; GO
The two major repair options used with DBCC CHECKDB are:
This option will usually try to repair all reported errors and might also cause some data loss. Therefore, you should use this option only as the last resort
This option does not incur data loss. For example, if a non-clustered index is damaged, you can be easily repaired by rebuilding the index.
If the methods discussed above fail to resolve the problem, use Stellar Repair for MS SQL. The software is embedded with several robust algorithms to extensively scan the corrupt database and perform a safe and reliable
MDF repair. The unique ‘Preview’ feature provides sneak-peek into the actual recovery results. Moreover, you can save the recovered queries, views, stored procedures, etc. in a separate text file that can be stored at any user-specified location in the system.
You need to follow the series of steps given below to perform precise recovery of your corrupt SQL database using SQL Recovery software:
- Main Interface of Stellar Repair for MS SQL is shown below along with the ‘Browse’ dialog box. In the dialog box, click ‘Browse’ to select an SQL database or click ‘Search’ to find and locate all SQL databases in the system.
- When the scan process is completed, all recoverable objects are displayed in a tree-view in the left pane. You can click an object to see its preview in the right-pane.
- Now, save the repaired database. You will get four options to save the database; MSSQL(MDF), CSV, HTML and XLS. When you click on the MDF option then, you need to insert the ‘SERVER\INSTANCENAME’ information. After specifying the instance and all related information click Save button.