I bought a 128GB TLC SSD(Sandisk X300) and a 128GB MLC SSD(Plextor M6S) at about the same time. I put them in ZFS mirror which is like RAID1.
Recently, I ran many stress tests on those SSDs for days without knowing that I was testing on SSDs instead of HDDs.
After 4 years and 4 months of lifetime, Plextor M6S started spewing many checksum errors while ZFS was scrubbing ZFS mirror. The errors were incorrigible.
I bought a new 128GB 3D TLC SSD, Transcend 230S SSD and replaced Plextor M6S with it.
It turns out that the warranty for Sandisk X300 was 5 years and the warranty for Plextor M6S was 3 years.
I suspect Sandisk X300 is going to start dying in one year since I ran stress tests on it, too.
I never ran trim because zfs on linux didn't support it. Recently, zfs on linux started supporting trim. I might be able to revive Plextor M6S with a manual trim operation, but it's safer to buy a new SSD. I'm going to have my computer run trim operations on SSDs every week with a cron job.
Recently, I ran many stress tests on those SSDs for days without knowing that I was testing on SSDs instead of HDDs.
After 4 years and 4 months of lifetime, Plextor M6S started spewing many checksum errors while ZFS was scrubbing ZFS mirror. The errors were incorrigible.
I bought a new 128GB 3D TLC SSD, Transcend 230S SSD and replaced Plextor M6S with it.
It turns out that the warranty for Sandisk X300 was 5 years and the warranty for Plextor M6S was 3 years.
I suspect Sandisk X300 is going to start dying in one year since I ran stress tests on it, too.
I never ran trim because zfs on linux didn't support it. Recently, zfs on linux started supporting trim. I might be able to revive Plextor M6S with a manual trim operation, but it's safer to buy a new SSD. I'm going to have my computer run trim operations on SSDs every week with a cron job.
Last edited:
Facebook
Twitter