5 Replies Latest reply on Jul 8, 2013 9:58 AM by joe_intel

    Media_Wearout_Indicator going up?

    AndersKvist

      I have been monitoring Media Wearout Indicator for a while, but I saw something odd in our Munin graphs, so I dumped the output of smartctl every minute and here are the result:

       

      ...

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

           [120 lines with value:092 removed]

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   092   092   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      233 Media_Wearout_Indicator 0x0032   091   091   000    Old_age   Always       -       0

      ...

       

      I tought the Media_Wearout_Indicator were going down from 100 as the drive were wearing out and that I could expect failure when it hits 0 - how is it going up again?! The value have been on 091 all the time since I've started monitoring it, but like the above, the value is 092 for a couple of hours before going down to 091 again...

       

      === START OF INFORMATION SECTION ===

      Device Model:     INTEL SSDSA2M120G2GC

      Serial Number:    ******************

      Firmware Version: 2CV102HD

      User Capacity:    120,034,123,776 bytes

      Device is:        Not in smartctl database [for details use: -P showall]

      ATA Version is:   7

      ATA Standard is:  ATA/ATAPI-7 T13 1532D revision 1

      Local Time is:    Tue Jul  2 14:39:14 2013 CEST

      SMART support is: Available - device has SMART capability.

      SMART support is: Enabled

       

      /Anders

        • 1. Re: Media_Wearout_Indicator going up?
          joe_intel

          I would recommend monitoring this value with the Intel® SSD Toolbox just to make sure the reading is correct. Additionally, you may run some diagnostic tests on the drive with the same software utility.

           

          Intel® SSD Toolbox

          https://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=18455

          • 2. Re: Media_Wearout_Indicator going up?
            AndersKvist

            It looks like Intel SSD Toolbox is Windows only?

             

            The disks are in a server running Linux, that's why I read the values with smartctl...

             

            Reading in the Intel SSD Toolbox user guide it looks like it uses S.M.A.R.T. the same way as smartctl to read the attributes, so the tool shouldn't give any other values. I'm using it for some Samsung SSDs as well that increases quite fast compared to Intel SSDs...

             

            E9

            Media Wearout Indicator

            Reports the number of cycles the NAND media has undergone. The normalized value declines linearly from 100 to 1 as the average erase cycle count increases from 0 to the maximum rated cycles. Once the

            normalized value reaches 1, the number will not decrease, although it is likely that significant additional wear can be put on the device.

             

            /Anders

            • 3. Re: Media_Wearout_Indicator going up?
              joe_intel

              Yes, the Intel® SSD Toolbox is for Windows* only.

               

              I did not fully understand your statement regarding the Samsung* SSDs. Is the same happening with those?

               

              We may assume that we will get the same readings from both software utilities; however, it is better to know for sure and discard the possibility of a software issue.

              • 4. Re: Media_Wearout_Indicator going up?
                AndersKvist

                No, we just see the Samsung disks wearing very fast compared to our Intel disks - values only going down...

                 

                Unfortunately, it's not an option to boot Windows - it's a Linux server with quite high load and everything depends on it

                 

                I know this isn't a disk made for server use, but now that Intel also makes SSDs for serveruse (S3700 and S3500), are there any plans to make own utilities that can run on Linux?

                 

                The above mentioned server is scheduled for retirement within a half year and will be replaced with a equal server with S3500 disks, so I will be able to test the disks, but it's going to take a while...

                • 5. Re: Media_Wearout_Indicator going up?
                  joe_intel

                  At this time there are no plans to release software for other operating systems.

                  You may let us know about the results in case you have the possibility to schedule a downtime for testing.