CODEWITHSUNDEEP
sql-server
nothrow
nothrow

Reputation: 16178

How can I cause consistency error in SQL Server?

I've got following error in dbcc checkdb output from our customer for one table (more of very similar lines):

Msg 8964, Level 16, State 1, Line 1
Table error: Object ID 212503619, index ID 1, partition ID 72057594046251008, alloc unit ID 72057594048675840 (type LOB data). The off-row data node at page (1:705), slot 0, text ID 328867287793664 is not referenced.

CHECKDB found 0 allocation errors and 49 consistency errors in table 'X' (object ID 2126630619).

This error was created when running upgrade of our software (if he restores DB from backup and run the upgrade again, the same issue reappears).

My question is - how can I possibly create this kind of error from my app? I always thought that this kind of error must be caused by some environmental (HDD) problem, but I've seen the same issue on the same table on another environment. I tried the same steps as him, but without success.

Thanks!

Upvotes: 2

Views: 1651

Answers (1)

usr
usr

Reputation: 171206

You're right, this is probably a severe bug in SQL Server. It is not possible to cause corruption using documented and supported T-SQL. To cause corruption you need

  1. hardware problems
  2. OS-level file system problems (filter drivers, ...)
  3. Undocumented commands like DBCC WRITEPAGE
  4. A severe bug

Can you single-step through the upgrade script? If not, try tracing it with SQL Profiler. Find the statement that first makes corruption appear.

Here is a simpler, less noisy command:

DBCC CHECKDB([AdventureWorks2012]) WITH NO_INFOMSGS, ALL_ERRORMSGS

Upvotes: 2

Related Questions