1

When i am extracting big rars it takes forever. However when i look in the task man cpu usage is only about 16%.

I am thinking that the hdd is the bottleneck, or maybe not? The hdd is an average 7200 rpm seagate.

Some know more about this?

2 Answers 2

5

For decompressing compressed data your disks will be the bottleneck. Decompressing is much easier on the CPU than compressing, so your CPU can most likely decompress data much faster than your disk can hand over the input and write out the decompressed output.

This is particularly true if you are extracting the archive to a location on the same disk, as the extra head movements (as they bounce back-and-forth between the location on the disk the archive is read from and the location the output is being written to) can reduce the read and write speed considerably.

1
  • A caveat: Most (all?) popular general data compression products available use what's called an asymmetric compression algorithm, with compression being much more complex (e.g. requiring more CPU time) than decompression. The basic idea being that you would only compress once, but you or many others might decompress several times. However, there are boutique and experimental compression programs with symmetric compression, and decompression can take just as long as compression. All that said, RAR is asymmetric, and the asker is very much I/O bound.
    – afrazier
    Apr 28, 2010 at 17:28
0

If your hd is fragmented it can certainly slow down the extraction, but Rar also extracts to a temp location before copying the final extract to the destination folder you selected.

Processor speed used to be the determining factor on archives way back, when loading a compressed file and then decompressing it took less time than loading the decompressed file straight from the hd. That's not really the case these days since people tend to work with much larger files.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .