{"id":100165,"date":"2022-06-07T11:04:22","date_gmt":"2022-06-07T11:04:22","guid":{"rendered":"https:\/\/www.stellarinfo.com\/blog\/?p=100165"},"modified":"2026-05-22T11:03:21","modified_gmt":"2026-05-22T11:03:21","slug":"how-to-fix-page-corruption-in-sql-server","status":"publish","type":"post","link":"https:\/\/www.stellarinfo.com\/blog\/how-to-fix-page-corruption-in-sql-server\/","title":{"rendered":"How to Fix Page Corruption in SQL Server: Complete Guide [2026]"},"content":{"rendered":"

Page is the primary storage unit in SQL Server. All the information in SQL Server database (MDF\/NDF) files is stored in pages of 8 KB size. Each page starts with a 96-byte header that stores information, like page number, page type, free space etc. If a page gets corrupted, then it can impact the entire database. In this case, you may receive different errors when opening the database file or even not able to access the database. Therefore, fixing page-level corruption immediately is important.<\/p>

What is Page-Level Corruption in SQL Server? <\/h2>

It is a specific type of corruption where only specific pages in the database are affected instead of complete database. When this happens, you may face errors like incorrect PFS free space information error<\/a>, error 8966<\/a>, and more.<\/p>

Pages may contain table data (data pages), index structures (index pages), large objects (text\/image pages), or allocation metadata such as GAM, SGAM, and PFS pages. In SQL Server, an extent is a group of eight physically connected pages that help in efficiently managing these pages. Allocation maps like Global Allocation Map (GAM), Shared Global Allocation Map (SGAM), and Index Allocation Map (IAM) track page usage.<\/p>

Here is a table showing the types of SQL Server pages used to store and manage data:<\/p>

Page Types<\/strong><\/td>Description<\/strong><\/td><\/tr>
Data Page<\/td>Stores actual table data (rows)<\/td><\/tr>
Index Page<\/td>Stores index structures for faster data retrieval<\/td><\/tr>
Text\/Image Page<\/td>Stores large objects such as TEXT, NTEXT, and IMAGE<\/td><\/tr>
GAM (Global Allocation Map)<\/td>Tracks allocated extents<\/td><\/tr>
SGAM (Shared Global Allocation Map)<\/td>Tracks mixed extents with available space<\/td><\/tr>
IAM (Index Allocation Map)<\/td>Maps extents used by a table or index within a 4GB GAM interval<\/td><\/tr>
PFS (Page Free Space)<\/td>Tracks page status and available free space<\/td><\/tr><\/tbody><\/table><\/figure>

What Causes Page Corruption in SQL Server?<\/h2>

The following table outlines the most common causes of page-level corruption in SQL Server database and their descriptions:<\/p>

Cause of Corruption in SQL Database Page<\/strong><\/td>Description<\/strong><\/td><\/tr>
Hardware\/disk issues<\/td>Bad sectors, issues in drive, I\/O errors, etc. Can cause errors such as error 823<\/td><\/tr>
Torn Page writes<\/td>Incomplete page write to disk — triggers error 824<\/td><\/tr>
Checksum mismatch<\/td>Page verification fails on read – results in Error 824<\/td><\/tr>
Sudden server shutdown<\/td>Crash during write operation – causes page corruption errors<\/td><\/tr>
Ransomware\/malware attack<\/td>Encryption or modification of pages in MDF\/NDF files – causes corruption errors<\/td><\/tr>
Software bugs<\/td>OS-level or SQL Server bugs affecting read\/write operations – leads to corruption scenarios<\/td><\/tr><\/tbody><\/table><\/figure>

Common SQL Server Page Corruption Errors<\/h2>

Here are some page-level corruption errors in SQL server.<\/p>

SQL Server Error 823<\/h3>

It is one of the common page errors which occurs when SQL Server fails to read or write a database page. It occurs due to hardware failure, bad disk sectors, or I\/O controller issues. Another variant of this error is cyclic redundancy check (CRC) error.<\/a><\/p>

SQL Server Error 824<\/h3>

This is one of the errors reported by DBCC CHECKDB. It occurs when this command fails internal validation checks on the database. The major reasons are page checksum failures, incorrect page header values, and others.<\/p>

SQL Server Error 825<\/h3>

Basically, it is a read-retry warning. SQL Server successfully read the page but only after retrying the operation. This is an early warning that the storage subsystem is degrading. The page was read successfully this time — but if the problem progresses, Error 823 failures will follow. It occurs due to disk or controller issues.<\/p>

How to Detect Page Corruption in SQL Server Database?<\/h2>

Before proceeding to the solutions to fix corruption, you need to check the SQL database for corruption. There are various ways to check page-level corruption in SQL Server database. Some of them are explained below:<\/p>

1. Run the SELECT Statement to Query the suspect_pages Table<\/h3>

SQL Server keeps a log of all the suspect pages in the suspect_pages table in msdb (system) database, containing all the corrupt pages that are reported by SQL queries. This table gets automatically updated when the server detects checksum failure, read\/write, or torn pages error. You can check the suspect_pages table to see the corrupted pages in the database. For this, you can run the SELECT statement as given below:<\/p>

SELECT * FROM msdb..suspect_pages<\/p>

\"result<\/figure>

The above command will return a grid with a row of each corrupted page. In the above image, you can see that the page, with id 368, is corrupted.<\/p>

Key Terms to Understand suspect_pages Results<\/strong><\/p>

To interpret the output of the msdb..suspect_pages table, focus on these core columns:<\/p>

database_id:<\/strong> Identifies the affected database.<\/p>

file_id:<\/strong> Specifies the database file (MDF\/NDF) where corruption occurred.<\/p>

page_id:<\/strong> Identifies the exact corrupted page within the file.<\/p>

event_type<\/strong>: Indicates the type of corruption error detected on the page:<\/p>