Today’s tip…

 

Starting with Windows 8 and Windows Server 2012, there has been a fundamental change to NTFS in how we handle file system inconsistencies.  Previously, there were a number of reasons that Windows would mark the file system dirty and urge the user to run a Chkdsk /f.  Depending on the size of the data set, this could mean hours or even days of downtime. 

 

The new approach is to catalog the inconsistencies, verify them, and fix them while the volume is still online when possible.  If it is not possible to repair the file system while online, instead of prompting the user to run a full Chkdsk /f, they will be prompted to run a Chkdsk /spotfix.  This will take the volume offline and fix the cataloged issues, rather than needlessly checking the entire file system.  This reduces the time of running Chkdsk to minutes or even seconds.

 

While it is still possible to run a traditional Chkdsk /f to check the consistency of all metadata, it is not likely that it will be needed.

 

This has been improved in Windows 8.1 and Windows Server 2012 R2 in that a larger range of inconsistencies will now be fixed while the volume is still online.  The result being even less downtime.

 

Also, I want to take a moment to thank Deborah Pisano of Microsoft News Center for her blurb about Tip of the Day in The Fire Hose blog today.