Hello,
I've been running a Samsung 1TB 850 PRO SSD in a USB enclosure with Mac OS 10.12.6 on a mid 2012 rMBP 15. Recently I've noticed that DriveDx has been giving me errors with Flash Read Fail count at 644 (started at 13 about two weeks back). Running the Full Self Test has initially shown no errors before this thing started and subsequently I've seen that it has been giving errors at LBA 546,559,560. Yesterday when I re-ran the extended test it passed without errors.
When I do a full scan of the disk with Techtool Pro 14 I get 32 bad blocks starting at the above address. Using CCC to backup this device there is one file which is not readable. Manually copying it off is also not possible.
Using TTP's benchmarking I've noticed the write speed of the drive has decreased to 200MB/sec with read speed at 400MB/sec. I thought this may be a cable or enclosure problem although why at the same address, but using a different cable and enclosure yields the same.
The drive has a 10 year warranty and I'm only into year 4 with only 1.7 out of 300 TB written to it. Anyhow is this a lack of trim problem with some blocks wearing out because of that or is my drive failing and I should return it? Would deleting the corrupted file allow the drive to mark its blocks/pages as bad and leave it at that? I've read that it's not uncommon for some SSDs to have bad "blocks". And DriveDx and TTP's SMART warning are both panicking because of the same bad blocks/pages being re-read and giving errors?
I don't have a Thunderbolt enclosure to test this drive in, unfortunately.
Any advice? Thanks.
I've been running a Samsung 1TB 850 PRO SSD in a USB enclosure with Mac OS 10.12.6 on a mid 2012 rMBP 15. Recently I've noticed that DriveDx has been giving me errors with Flash Read Fail count at 644 (started at 13 about two weeks back). Running the Full Self Test has initially shown no errors before this thing started and subsequently I've seen that it has been giving errors at LBA 546,559,560. Yesterday when I re-ran the extended test it passed without errors.
When I do a full scan of the disk with Techtool Pro 14 I get 32 bad blocks starting at the above address. Using CCC to backup this device there is one file which is not readable. Manually copying it off is also not possible.
Using TTP's benchmarking I've noticed the write speed of the drive has decreased to 200MB/sec with read speed at 400MB/sec. I thought this may be a cable or enclosure problem although why at the same address, but using a different cable and enclosure yields the same.
The drive has a 10 year warranty and I'm only into year 4 with only 1.7 out of 300 TB written to it. Anyhow is this a lack of trim problem with some blocks wearing out because of that or is my drive failing and I should return it? Would deleting the corrupted file allow the drive to mark its blocks/pages as bad and leave it at that? I've read that it's not uncommon for some SSDs to have bad "blocks". And DriveDx and TTP's SMART warning are both panicking because of the same bad blocks/pages being re-read and giving errors?
I don't have a Thunderbolt enclosure to test this drive in, unfortunately.
Any advice? Thanks.
