Some might tell you that you can use a performance monitoring tool such as Windows' PerfMon utility to determine that a defrag utility is doing its job. Sounds reasonable. And I would agree that you can get an idea of fragmentation's impact on a disk subsystem (a single disk or RAID/SAN). But that's not enough.
To really get a handle on whether a defrag tool is helping, it's important to look not just at the entire drive and overall statistics for the drive. It's very important to look at all the files on a drive. Because if you have, for example, a thousand files on your drive, and 950 are completely defragmented but the other 50 are a mess, your drive is not in the best possible shape it can be. So you need to look at all files on the drive, and you need to look at write times specifically too. Because if you are writing new files to your drive and your free space is not consolidated, you're going to run into performance problems. And to determine this, you need to use more than PerfMon. You should use a tool like hIOmon, which provides very detailed performance information not only at the device level, but also at the file level, giving you the true story of how much a defragmentation tool is helping.
For a detailed overview of determining the impact of file fragmentation and free space fragmentation on drive performance, check out Greg Hayes' (Raxco Software's 5-time Microsoft MVP) article on this subject.
This is why free space consolidation matters - without it, your drive not only fragments more quickly, but performance suffers when you are writing to the drive.
And running all the time, over and over, time and again, invisibly or in full view, without consolidating free space, won't change that fact. You might think you're going somewhere fast, but you're just spinning your wheels.
Comments