DBCC Checks and Terabyte-Scale Databases

This document covers some of the options available to reduce the impact of running DBCC checks on VLDBs. ... Source : Stuart Ozer - Technical Note - SQLCAT.com

Running DBCC on a VLDB

In this Blog post, Paul Randal discusses the options for running DBCC CHECKDB on VLDB's  ... Source : Paul Randal - SQL Server Storage Engine - Blog Post

How long will CHECKDB take to run?

There are many factors that determine how long a DBCC check is likely to take on a particular database. In this blog post, Paul covers the top 10 ... Source : Paul Randal - SQLskills.com - Blog Post

Sparse Files; Under the Covers

Sparse files enable features such as online DBCC checks and Database Snapshots fast and efficient, but there is a fair degree of mystery surrounding how they actually work. This post addresses that very topic ... Source : Sarah Henwood & Bob Dorr - CSS SQL Server Engineers - Blog Post

Running DBCC on a user defined snapshot

DBCC uses a snapshot to reduce the performance impact on users whilst running. The problem with this is that the snapshot is created on the same disk(s) as the database, and may lead to failure if all the disk space is consumed by the snapshot during the dbcc check. In this post, Kalen talks about running the DBCC check on a user defined snapshot, which allows you to define the disk location for the snapshot, and therefore have more control of disk space usage during the DBCC check ... Source : Kalen Delaney - SQLblog.com - Blog Post 

Database Checksums - Is CHECKDB still necessary?

The database checksum feature introduced in SQL Server 2005 goes a long way to detecting physical corruption as early as possible, however, DBCC CHECKDB (and related commands) are still very much needed, as Bob Dorr points out in this post ... Source : Bob Dorr - CSS SQL Server Engineers - Blog Post

Error 825, a.k.a Impending Doom

A common (and recommended) practice is to setup SQL Server alerts for all high severity errors, typically level 16 . Amongst other things, this captures errors 823 and 824, both of which relate to I/O errors. Error 825 is raised when an I/O error occurs, but subsequent reads succeed. The problem here is that this error is raised as a level 10, and therefore not captured by a lot of monitoring systems. As Paul points out, creating an alert for this particular error is crucial in receiving an early warning of possible future I/O problems, or as he puts in, Impending Doom ... Source : Paul Randal - SQLSkills.com - Blog Post

DBCC CHECKDB: Last Success and Completion Status

Do you use SQL Agent Jobs to run DBCC CHECKDB? If the job completes without error, do you assume that the check was successful? In this blog post, Paul highlights the danger of this assumption, and tells us how we can find out the last successful CHECKDB operation for a given database ... Source : Paul S. Randal - sqlSkills.com - Blog Post