General fragmentation question - afterhours

I'm not sure I understand your question. Techtool told you that your defrag would result in just 60Gb of data on a 500 Gb hard drive. The 460 is the actual usable storage space on a 500 Gb drive after formatting.

The formatting process takes 6 to 8% of the drive space. You aren't getting ripped off (much as slimy legal eagles try to class-action themselves into independent wealth on the coattails of technologically ignorant juries). You get 500 Gb of storage, and once you decide how to format that storage, you will have to use up some of it for the electronic fences the OS and drive uses to keep file A from bleeding into file B. This is no different than buying an acre of land, and building a hundred stockades for your cattle. Add up each stockade, and you get less than a full acre because the fencing or walls take up space.

does that help?

Reply to This

Wednesday, September 24 2008 @ 09:29 AM PDT

General fragmentation question - Crazy Dave

Just wondering how exactly you cloned the HD. If you used block-level cloning software, this would have created an identical hard disk image,and would not have defragmented your files at all. A true block level clone does not modify any data or file mappings. You could do a hundred block-level clones of a volume and still have file fragments, because the process completely ignores the files themselves, concentrating on the underlying volume itself.

There are advantages and disadvantages to the block-level method:
If your hard disk is only 25% full like you mentioned, the software may actually clone the ENTIRE volume (including empty blocks) all 100%, to the other hard disk. This will obviously take much longer than only copying the bits that are occupied with data. Some software does a block-level clone of only the used portion of the HD, but some only clones the entire volume. Some software allows you to change this setting depending on requirements.
If your hard disk is almost full and has many fragmented files, block-level copies can be faster than file-level, because the cloning application will completely ignore a volume structure and simply copy the hard drive in bulk. Also, the hard disk reads all data one block after another. The slowest part about a hard disk is the time taken moving the read/write head to a different place on the hard disk platter, which is not necessary during a block-level copy. This is where the two values 'Random R/W' and 'Sustained R/W' are useful. If you're copying fragmented files, then the hard disk will thrash around searching for every single piece of the file before it can complete the copy. Block-level copies everything contiguously, wasting no time on file seeking.

If you performed a file-based backup then restored the files to another volume, that's a different story altogether. This proceess reads every file fragment from the source volume and sticks them back together into one neat unfragmented file when it writes them to the target volume. This is what you want if your goal is to defragment the hard disk.

I see many people complaining about slow defragmentation speeds, but degfragmenting is almost completely dependent on hard disk speed. If you have to defragment 60 GB of files, it will take exactly the same amount to degrafment it as it will to physically copy 60GB of files from one place to another. Get a stop-watch and time how long it takes to duplicate a 1GB file. If it takes 1 minute to duplicate 1GB, then defragmenting 60GB of files will take rougly 60 minutes. Defragmenting a 750GB hard disk will take over 12 hours. If your hard disk and computer are slower, it make take 3 minutes to copy 1 GB, therefore defragmenting 60GB will take roughly 3 hours. The physical access times on hard disks have remained virtually unchanged for the last 5 years or so, so a PowerMac G4 will take about the same amount of time to defragment a disk as a quad-core MacPro. There have been advances when comparing IDE to SATA, but many of those are due to better caching techniques which don't affect sustained transfers which are needed in the defragmentation/cloning processes.

Reply to This

Wednesday, September 24 2008 @ 12:27 PM PDT