Question on defragging files

[trinaz]

Using Defraggler for my system defrags. My main C drive reports that the System Volumn Information file is fragmented...but when I try to defrag this file...I get a "no files defagged" message from Defraggler ?

Is this an issue I should be concerned about...and will things just keep humming away with or without defragmenting this file ?

Thanks...TR

 

[kado897]

There are certain files that Defraggler or any other defrag program will not touch. This is one of them.

 

[A Guy]

You might try Puran Defrag, it has a boot defrag option that may defrag your system files if they are not in use (during boot). A Guy

 

[reckey]

Or try the demo version of any of the full fledged commercial defraggers which defrag files that the built in cannot.

 

[cluberti]

If you run defraggler as system (psexec works well for this), it can defrag these files. The problem is they're owned by SYSTEM, and unless you have full access to the folder and it's contents, you can't touch it with a user-initiated defrag of an online (running) system. The default Windows 7 scheduled defrag task can potentially do this (it runs as SYSTEM), as will any defragmentation app that is launched as SYSTEM.

***CAVEAT***
There's a reason most defrag apps do NOT touch this location while the system is online. You MUST beware of odd behaviors when defragmenting this location as SYSTEM on a running Windows installation, and take a very important precaution before doing so - you need to completely disable the paging file (and reboot) before defragmenting this folder, or you risk bugchecking the next time you reboot after the defrag is run (Vista is more susceptible to this than Windows 7, but it pays to be safe just in case). This location contains just what the name sounds like - System Volume Information - and that includes information about the paging file. If the paging file exists when this location is defragmented, the information about that paging file can become corrupt in the System Volume Information folder, causing a bugcheck after a reboot. To avoid this, disable the paging file, reboot, defragment the drive as SYSTEM via psexec and defraggler, and then reboot again. You can re-enable the paging file at this point if you so choose safely.

 

[kado897]

Thanks, you learn something new every day. I suppose the real question is why would you want to, A few fragments here are not going to affect performance noticeably.

 

[cluberti]

Actually, this is one location that gets hit... A LOT. This, like the paging file, is actually quite a good thing to defragment for performance reasons on mechanical drives . I do this on average about once a month, if it becomes fragmented (I don't run with a paging file, but that's a personal preference). I've started moving to SSDs as well, where fragmentation doesn't really mean anything, but mechanical drives are still going to be around for a long time unfortunately.

 

[kado897]

I've not tried it but would not turning paging off and back on again have the same effect?

 

[seekermeister]

Actually, this is one location that gets hit... A LOT. This, like the paging file, is actually quite a good thing to defragment for performance reasons on mechanical drives . I do this on average about once a month, if it becomes fragmented (I don't run with a paging file, but that's a personal preference). I've started moving to SSDs as well, where fragmentation doesn't really mean anything, but mechanical drives are still going to be around for a long time unfortunately.

Hmm, on the last defrag, I believe that my System Volume Information was fragged into 25 pieces. Though I didn't care for the idea, I took it on face value that since the file was locked, that it should stay that way, but your statement regarding it's effect on performance got my attention. Would this be the reason that for many people, that the OS seems to bog down over time, and causes them to reinstall the OS as a means of restoring the original performance? I can imagine that there are other reasons also, but I'm wondering just how significant this is in the overall scenario?

EDIT: Correct that frag count to 114 pieces.

 

[seekermeister]

cluberti,

You inspired me to run my first offline defrag. I had only tried that once before, long ago, because at that time O&O ran it's own video driver for that process, and it caused problems that forced me to reinstall the OS. I learned afterward that there was another solution, but the experience left a bad taste in my mouth. It was surprising how much faster that the defrag was offline, that it is doing it from desktop. I haven't noticed any increase in performance, but it feels good that O&O now shows no fragmentation, including locked files. Thanks.

 

[cluberti]

- defragmenting your drive might not show much of a performance increase if the disk is fast and the cache is good, but it doesn't hurt either.

 

[kado897]

- defragmenting your drive might not show much of a performance increase if the disk is fast and the cache is good, but it doesn't hurt either.

It can hurt if you go at it too heavily. I have had it delete restore points. Even though I use Defraggler I only do quick defrags and individual files.

 

[cluberti]

If your disk is that fragmented, the problem isn't fragmentation - it's free space. Defraggler is actually quite good at defragmenting drives with low amounts of free space (15-20% free), but no defragmenter should be causing the deletion of restore points, although it can (and does) interact with VSS as would anything that hits the drive hard. Again, this is why defragmenting infrequently (perhaps once or twice a month) is more beneficial than doing it daily, or weekly, on a desktop-class system. On servers this can be different, but you aren't using restore points and such as a recovery solution in those instances, either (I hope!) .

 

[kado897]

If your disk is that fragmented, the problem isn't fragmentation - it's free space. Defraggler is actually quite good at defragmenting drives with low amounts of free space (15-20% free), but no defragmenter should be causing the deletion of restore points, although it can (and does) interact with VSS as would anything that hits the drive hard. Again, this is why defragmenting infrequently (perhaps once or twice a month) is more beneficial than doing it daily, or weekly, on a desktop-class system. On servers this can be different, but you aren't using restore points and such as a recovery solution in those instances, either (I hope!) .

No I don't rely on RPs but I have found them useful in the past.There is a thread on Defragler deleting RPs on their forum.
Restore points - Piriform Community Forums

 

[cluberti]

Again, it's not defraggler but the act of defragmenting the volume. The more data on a volume, the more likely defragmenting it will cause RPs to get deleted - VSS can only hold so much data, and the way VSS works means anything that moves (lots of) files around in a short time (thus consuming basically double the space on disk in a short time) can do this. Ultimately, you'll find that Diskeeper, Auslogics, JKDefrag, and even the inbox Windows defrag does this (and I am sure there are other agressive defragmenters that can be listed here as well). It's just a by-product of moving what is essentially the vast majority of both large and small files around on the disk at one time interval on a volume with the default cluster size (4K) rather than using 16K (or higher) cluster sizes. The drawback to that is that files less than 16K or that do not fall nicely on 16K boundaries will use more space than they would on a 4K cluster volume, so pick your poison .

 

[kado897]

Agreed. It's not just Defraggler any heavy defragging will cause loss of RPs which is why I do little and often. When I first used Defraggler to defrag my c: drive which was in a mess. I lost all my RPs.

 

[cluberti]

I've heard through the grapevine that doing a file defrag only (rather than full) doesn't cause the issue, but that would make sense (it only does a small amount of file moving versus the whole disk).

 

[Swindon42]

I have users with huge ost files that won't defrag. I've used the xp util and contig under users login / safe mode / admin, the works. Any suggestions as this really affects outlook performance. (cant kill VP execs, and pst/archive dissabled by GP) Any suggestion would be appreciated.

 

[cluberti]

Outlook performance is more affected by item number in folders rather than ost fragmentation. I would agree that very large ost files should be kept as contiguous as possible, but it is just as important to keep item counts in folders to 5000 items or less, that your real-time antivirus does not scan ost or pst files, and the user's mailbox is kept as tidy as possible. Having a multi-gb ost is more an indication that you need an archival solution rather than there is an outlook ost perf issue. Outlook prior to 2010 had a 20gb max *st file size, and 2010 raised that to 50gb. However, getting anywhere near that is stretching the performance envelope for jet-based database files like outlook uses.

However, I would say a quick and dirty way to defrag a file that otherwise won't defrag is to move it to another volume, defrag the source volume, and then move it back.