I've got a pair of Intel DC S3520 480gb SSDs in a ZFS mirror. One of them has 36% wearout while the other has 34%. It seems that any intensive write operation at all locks up the host system and nightly backups to a remote host seem to make it just die. I've attached screenshots of both their SMART reports respectively, although I must admit I have no idea what I should be looking at here. I've got about only 5 VMs on them, 1 is a Windows VM and the others are Ubuntu respectively.
According to iotop one vm does write to those disks at around 5 ish Mb/s constantly though I'm not entirely sure this is the single handed culprit since there are period where it does write but io delay is <2%, but other periods when its >25%. I'm not sure if I should be replacing these SSDs at this point or not but it is an option, and I might be replacing them with Intel S4510 480gbs. Anyways, any pointers or places I should look for this problem?
According to iotop one vm does write to those disks at around 5 ish Mb/s constantly though I'm not entirely sure this is the single handed culprit since there are period where it does write but io delay is <2%, but other periods when its >25%. I'm not sure if I should be replacing these SSDs at this point or not but it is an option, and I might be replacing them with Intel S4510 480gbs. Anyways, any pointers or places I should look for this problem?
