Disabled defragmentation for SSD...but it seems like defrag is still running?

[996GT2]

I've set scheduled defragmentation to "off" and run defrag manually for the two hard disks in my system. In addition, Intel SSD toolbox also says that my system configuration settings are "optimal" for SSD.

But when I checked today, Windows defrag says that the last defrag run on my SSD (c drive) was on 5/6. This was strange, since scheduled defragmentation was disabled long before then. I also haven't run the TRIM utility since 5/2 (see screenshot), so why would Windows show a defrag run on the SSD on 5/6? Does anybody know what's going on here?

 

[996GT2]

Bump

 

[groberts101]

only thing I could possibly think of right off would be something like Norton system utility(or whatever they hell they call it these days) using Windows defrag as part of its maintenance routine.

 

[996GT2]

only thing I could possibly think of right off would be something like Norton system utility(or whatever they hell they call it these days) using Windows defrag as part of its maintenance routine.

You're right. Looks like it could be Norton's Idle Time Optimizer doing the defragmentation in the background. I've disabled it and hopefully it doesn't try to defrag the SSD again.

 

[capeconsultant]

I went to Norton Community for Beta 2010 Internet Security Suite and complained about this. Bad news. they had better fix this. Others were having same issues.

google "Norton Idle time optimizer" to find thread.

 

[996GT2]

I went to Norton Community for Beta 2010 Internet Security Suite and complained about this. Bad news. they had better fix this. Others were having same issues.

google "Norton Idle time optimizer" to find thread.

I've disabled Idle Time Optimizer, so it should be ok, right?

 

[groberts101]

been a while since I used Norton(it's such a PIG) but IIRC I had to do that on my sis-in-laws system when I installed an SSD in it. Seemed fine afterwards and I maintain it on occasion without seeing any sign of defrags being run after the fact.

 

[bobal]

I've always used diskeeper and it never defrags the ssd. Diskeeper also prevents most fragmentation from happening in the first place.

 

[evilspoons]

As arrogant as this sounds, my general advice is to take the Norton product and throw it out the window and just use Microsoft Security Essentials. It's been proven to be up there with the best of them for antivirus capabilities, and it stays the hell out of the way. It's only annoyed me when I've left my computer off for two months and its "oh noes you haven't run a scan in a long while" thing went off.

Oh yeah, and it doesn't cost $90 every 12 months.

 

[bobal]

I agree. Security Essentials is a very decent program.

 

[groberts101]

LOL... diskeeper IS a drive defrgamenter. It's just a much more SSD friendly version and limits the amount of data moved and helps to reduce the fragmentation to begin with. I use Perfect Disk 11 on my SSD every month or so to consolidate free space of the bitmap for greater efficiency. Yes.. even SSD has latency(low latency is not the same as no latency) and what most fail to realize is that fragmentation will reduce efficiency of logical volumes to the point that it can reduce bandwidth even on an SSD.

And back to the topic.. all I use is Security essentials myself due to the piggish ways that most AV programs are hogging resources. My dedicated virus protection is called secure erase and reimage back to fresh installed states. lol

 

[Emulex]

not a bad idea to defrag large fragment files using contig on ssd 6 million fragments could add up even with low latency

 

[d4a2n0k]

People still pay for anti-virus programs? I havent let a Norton product touch my computers since 1999. I agree with the above, get MS security essentials its free and doesnt take over your computer.

 

[Voo]

not a bad idea to defrag large fragment files using contig on ssd 6 million fragments could add up even with low latency

Yeah, because as we all know LBAs correlate to one specific physical address on SSDs Correct me if I'm wrong but since the whole premise of LBA_n being followed by LBA_n+1 doesn't hold for SSDs, defragging can't help.

 

[Emulex]

but there is overhead associated with fragments. times a million or 6.

 

[Voo]

but there is overhead associated with fragments. times a million or 6.

Yeah but that's solely in the hand of the controller.

The defragger just assumes that LBA_n will be followed by LBA_n+1 so it puts associated files at adjacent addresses - but since the controller will just shuffle the blocks around as it sees fit, that's a forlorn hope. The defragger doesn't even know the mapping of LBAs to physical addresses (I don't think anyone even publishes how to get to that data) so how should that work?

 

[Emulex]

i'm talking about operating system overhead. continuous blocks must take up less information than millions of fragments. Not controller fragment issues - operating system. In this case NTFS

 

[groberts101]

I'd be happy to start a new thread with those consolidation tests to prove the neighsayers wrong if you guys are game?

Emulex is dead on with his observation and logical losses do in fact get handed down the stack towards the SSD. If one understands the amount of overhead and hogging that Windows already imposes onto the SSD(why do you think an empty SSD always runs faster than one with an OS installed?) it's not very hard to imagine the efficiency losses from a bitmap that's less than optimised.

And NTFS is absolutley horrid for fragmentation and I welcome the day that Windows does away with it. In fact if Linux had better program compatibility I'd have moved away from the pig we call Windows long ago.

Furthermore.. AV programs often write to the drive sporadically and in small random files to the point that GC is often interupted(remember trim isn't always best at cleaning up and idle time GC is key for heavy users) and fragmentation continuiously increases as a result. They are obviously necessary evil's for some but they aren't worth the tradeoff for my system.

 

[Voo]

I'd be happy to start a new thread with those consolidation tests to prove the neighsayers wrong if you guys are game?

An interesting use for defragmentation, so absolutely.

 

[wsaenotsock]

Windows defrags your boot sector automatically (changing the scheduled tasks in disk defrag does nothing) see this thread

http://forums.anandtech.com/showthread.php?t=2158683&highlight=

 

[Makaveli]

I use Perfect Disk 11 on my SSD every month or so to consolidate free space of the bitmap for greater efficiency.

This caught my eye care to go into abit more detail?

 

[Emulex]

me too. about once a month. i like to keep things tidy

 

[Emulex]

btw remember like COMPACT , CONTIG defrags at the drive, or dir, or file level. it doesn't do any compacting like the diskeeper program.

one might contig your sql server files and logs while leaving the rest alone. seriously i did a huge zip file and it was taking so long that grew into millions of fragments since it ran into another few jobs.

and something old school in me thinks the less fragments - the less work a catastrophic or forensic recovery might be. you know if you had to piece together 10 pieces versus 10K - one might be easy to do.

 

[groberts101]

although I don't have time to ward off all the "SSD doesn't respond to logical data manipulation because access is already so low" or the old mandatory "logical and physical data relationships are irrelevant because the SSD is always fragmented" responses.. here's a quick teaser of what I've seen for more than a year now.







It should be noted that the gains I see are cumulative across a wider 6 drive array and aren't quite as pronuonced on single drives(unless fragmentation levels are higher or the bitmap is spread out badly) but as soon as I started heavily testing Tony-Trim over a year ago the potential gain became quite obvious.

With WELL over 500 specific tests for this particular optimization(not to mention the actual usage).. the consistency of gain is amazing(even at lower fragmentation levels shown above). Will start a nice jousting thread when the time permits.

 

[Phynaz]

I'm shocked that people on this forum think you can defrag an SSD.

People here should know better.