← All articles
integritymaintenancehealth-check

CHECKDB skipped until corruption shows up

Why integrity checks get postponed, what that costs, and how production teams make CHECKDB practical again.

Corruption does not schedule a convenient appointment. It appears during month-end, during a failover, or during a restore you desperately need.

Yet many estates only discover that DBCC CHECKDB has not completed successfully in months — or ever — during a formal review.

Why teams skip it

  • “It takes too long”
  • “It blocks users” (often a misconception about how it is run)
  • “Storage is reliable now”
  • “We rely on backups” (which may already contain corruption)

None of those remove the need for a known-good integrity signal.

What a health check looks for

  • Last successful CHECKDB per database (and whether results were reviewed)
  • Jobs that exist but fail, time out, or only cover some databases
  • Physical-only shortcuts used without understanding trade-offs
  • Alerting that ignores integrity failures
  • Environments where CHECKDB cannot finish inside the maintenance window

Making CHECKDB sustainable

  • Right-size the window; do not pretend a 2TB database needs zero resources
  • Prioritise critical databases if you must phase rollout
  • Monitor duration trends — sudden growth is a warning
  • Treat failure as a severity incident, not a ticket for “next month”
  • Pair integrity work with tested restores

The consulting angle

Skipping integrity checks is rarely malice. It is usually capacity, fear of impact, or inherited neglect. A structured health review surfaces the gap, sizes the real effort, and puts CHECKDB back into a plan the business can fund.

Need hands-on SQL Server help?

Allay Data Solutions offers maintenance, tuning, and assessments from Port Elizabeth, South Africa.

Get in touch