Announcement

Collapse
No announcement yet.

Samsung 980 NVMe SSD Linux Performance

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

  • #21
    Originally posted by stormcrow View Post
    I should point out that Bad Things Happen when Linux encounters full storage devices. That's why it still isn't a good idea to put all your data on a single partition, regardless the defaults on the major distributions. Bugs and misconfiguration happen. This has become even more the case with SSDs which require enough free cells to (hopefully) move data from failing cells to unused ones. Some MFGs are good at over provisioning. Some are not. I don't think it's wise with the obvious new race to the bottom to test out with your important data just who is going to win that race. It's also one reason spinning rust isn't going away any time soon. Static long lived data is better off on a traditional hard drive as they're generally better with refreshing the surface than SSDs which may or may not bother to check if the charge in the storage cells need a refresh. Testing is showing SSD cell charge may not be as stable as MFGs claim.
    Have you tried to recover data loss on a hdd ?
    Its better on tape, bluerays, and encrypted cloud zip and hdd. Just make many copies instead of rely on hdd which will be 1000 dollars to recover.

    Comment


    • #22
      Originally posted by torsionbar28 View Post
      With ZoL, the regular scrub of the pool does precisely this, since every block is checksum'd on ZFS.
      So, Could you share more about your backup routine?
      I have files I want to be sure are well backup (Family Photos / Videos).
      I'm concerned which mechanism could validate them over time (That the HD itself didn't miss a bit flip etc...).

      Comment


      • #23
        Originally posted by stormcrow View Post
        I should point out that Bad Things Happen when Linux encounters full storage devices. That's why it still isn't a good idea to put all your data on a single partition, regardless the defaults on the major distributions. Bugs and misconfiguration happen. This has become even more the case with SSDs which require enough free cells to (hopefully) move data from failing cells to unused ones. Some MFGs are good at over provisioning. Some are not. I don't think it's wise with the obvious new race to the bottom to test out with your important data just who is going to win that race. It's also one reason spinning rust isn't going away any time soon. Static long lived data is better off on a traditional hard drive as they're generally better with refreshing the surface than SSDs which may or may not bother to check if the charge in the storage cells need a refresh. Testing is showing SSD cell charge may not be as stable as MFGs claim.
        Why couldn't SSDs move used cells around when storage is full, instead of writing to the very few free cells?

        Comment


        • #24
          I just replaced two 980 PRO with SN850ies. I noticed the performance is a lot better when re-formatting with 4KiB sector size using the NVME CLI.

          Comment


          • #25
            Originally posted by ThoreauHD View Post
            This is a great advertisement for the WD SN850. I didn't realize Samsung made such sad trash. I mean, they make their own memory, controllers, and flash. How do you fuck that up?
            I don't know where you live, but in my country, WD only sells a handful of new trash and mostly 3-10 year old trash (yes, it really is that old!). Seagate ftw

            Comment


            • #26
              Originally posted by tildearrow View Post

              Agreed. And I am still scared of buying an SSD, feeling that it will fail all of a sudden (unlike HDD which fails gradually).
              i use ssd's for a long time and had some failures.
              the good thing about ssd's is that you know when they will fail (there are exception though). just look at smart.

              so i do prefer ssd's over hdd's.
              (personally i by micron server ssd's)

              Comment


              • #27
                With their hard-won reputation it's taken so long to build, it's almost incomprehensible that Samsung would even consider releasing something like this.

                Comment


                • #28
                  Originally posted by stormcrow View Post
                  I should point out that Bad Things Happen when Linux encounters full storage devices.
                  Care to elaborate?

                  Comment


                  • #29
                    Maybe you didn't have it on hand but I would have liked to see how it compares to the 980 pro.

                    Comment


                    • #30
                      Originally posted by Royi View Post
                      So, Could you share more about your backup routine?
                      I have files I want to be sure are well backup (Family Photos / Videos).
                      I'm concerned which mechanism could validate them over time (That the HD itself didn't miss a bit flip etc...).
                      Take a look at the ZoL project, here. This is really the big selling point of ZFS, is that it has the ability to actively identify and correct silent corruption aka "bit flips". Personally I tend to use older hardware purchased second hand on e bay and other places, and like you, have data such as family photos that I intend to keep for many years.

                      Regular filesystems cannot detect or correct bit flips. Hence the name "silent" corruption. RAID arrays can detect a change, as the data no longer matches the parity, but they cannot correct it, as they don't know which one is wrong. ZFS can not only determine if a bit flipped, it can also correct it, and is therefore effectively immune to silent corruption. It also has a clever snapshot system and a send/receive feature that makes exporting the snapshots to external media really easy, for backup purposes.

                      It does have some drawbacks however. It generally performs slower than regular filesystems due to the overhead of all these safety features. And it's strongly recommended to use a system with ECC memory (i.e. Xeon, EPYC, Opteron) to avoid bit flips that occur in RAM while the data is cached or in-flight. Fortunately, this kind of enterprise grade hardware can be had inexpensively on the second hand market. It doesn't need a ton of horsepower, so a DDR3 era system works perfectly fine. I recently bought a Supermicro motherboard with Opteron CPU, cooler, and 64 GB of Registered ECC memory for $225 on e bay, a pretty good value for this purpose.

                      Comment

                      Working...
                      X