Announcement

Collapse
No announcement yet.

Linux 3.3 Kernel: Btrfs vs. EXT4

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #21
    Originally posted by energyman View Post
    with an SF controller who does compression you do not gain any space - but you are wearing the SSD down if you compress the data. Because the controller will not be able to compress it further, defeating one of the wear leveling mechanism.
    So if you write 50GB to the SF SSD without compression, you'll see less than 50GB used on the filesystem?

    Comment


    • #22
      no, you see 50gb used. But internally it is only say 40 gb, saving the flash chips from some dangerous write cycles.

      Comment


      • #23
        Originally posted by energyman View Post
        no, you see 50gb used. But internally it is only say 40 gb, saving the flash chips from some dangerous write cycles.
        Exactly. Personally I'd use compression anyway to save space, unless it hurt write performance by >50%. The lifetime of the SSDs is fine anways regardless, so this is the route I'd go

        Comment


        • #24
          emm, no - the write cycles are not 'ok'. The smaller the structure the less write cycles you get. Modern flash chips write cycles are pathetic.

          Comment


          • #25
            Originally posted by energyman View Post
            http://marc.info/?l=linux-ext4&m=133052231227201&w=2

            oh look, btrfs a lot faster than ext4 in a real world example.
            Well that's borked thread. Follow https://lkml.org/lkml/2012/2/29/210

            As one can notice, ext4 does not keep layout close thus requires more seeks to get the data read. Despite that it makes certain operations slower it also will eat your HDD(s) faster.

            -df

            Comment


            • #26
              Check this out - https://lkml.org/lkml/2012/2/29/210

              Comment


              • #27
                Originally posted by energyman View Post
                emm, no - the write cycles are not 'ok'. The smaller the structure the less write cycles you get. Modern flash chips write cycles are pathetic.
                Sure. Do you have a source that shows a rough estimated lifetime of a Sandforce SSD? (normal use)

                If it's 5 years than even halfing that to 2.5 years is definitely OK. As always, it depends on your usage scenario.

                I think Anandtech has something but I can't find it at the moment.

                Comment


                • #28
                  http://www.anandtech.com/show/5508/i...to-sandforce/7 - don't fill a complete Sandforce SSD with a compressed file system, or it's performance (and presumably lifetime) will decrease substantially.

                  Comment


                  • #29
                    Originally posted by AnonymousCoward View Post
                    http://www.anandtech.com/show/5508/i...to-sandforce/7 - don't fill a complete Sandforce SSD with a compressed file system, or it's performance (and presumably lifetime) will decrease substantially.
                    Thanks. I can't wait for SSDs to become good and usable.

                    Comment


                    • #30
                      Originally posted by AnonymousCoward View Post
                      http://www.anandtech.com/show/5508/i...to-sandforce/7 - don't fill a complete Sandforce SSD with a compressed file system, or it's performance (and presumably lifetime) will decrease substantially.
                      Hm, very interesting drawback of filesystem compression...

                      Comment

                      Working...
                      X