steve626 wrote:
HWTech wrote:
SMART failure notices are not always catastrophic for an SSD.
Maybe you have more experience with this than I do, but I beg to differ here. I have never seen SMART failures that were not catastrophic or close (in time) to catastrophic, on HDs as well as SSDs.
The macOS "SMART Status" indicator only ever triggers once a SMART health attributes exceeds the manufacturer's expectations for a "Pre-Fail" attribute. There are also "Old Age" or "Lifespan" attributes which usually are not configured to trigger a Failed SMART status, however, an app like DriveDx will list those attributes with a "Warning", but macOS itself will not be aware of these "Old Age" attributes reaching these limits.
There is a huge difference here between Hard Drives and SSDs. With HDs I have only seen the macOS/Disk Utility show a "SMART Status Failed" condition one time.....maybe two twice over 20 years or so since SMART has been around. HDs usually become too slow long before the SMART status failure can ever be triggered. Generally if DriveDx reports a HD having any "Warning" or "Failing" notices, then it indicates the HD should be replaced (at least in my own personal experience).
Unfortunately with SSDs things are a bit more complicated. Usually an SSD's health needs to be manually interpreted. Apps like DriveDx can alert a user to a change in the SSD's health so that the user can then manually interpret the new health details to determine if the SSD has a serious problem, or whether to wait until the next alert to reassess. Unfortunately most people don't really know how to interpret an SSD's health report. Even I have trouble assessing an SSD's health especially with the newer SSDs where very little health information is presented.
In my own personal experience, the majority of SSD failures are due to the SSD's controllers no longer being able to communicate with the computer. The other time where I've seen SSDs fail is after writing PBs of data to the SSD. I can only recall one maybe two times where an SSD completely failed for another reason. Of course with some of these "bad" SSD's I am resetting the SSD's hardware to factory defaults which is why I've been able to continue using "bad" SSDs without any further problems and have seen few other actual failures.
The thermal SMART warning are a warning of hardware damage that has already occurred.
That is probably accurate if the temperatures went beyond the "Critical" temperature range for too long.
Read and write errors growing beyond SMART limits are also serious, the drive has basically worn out. If the drive is trying to reallocate sectors in growing number, this is also a failing drive. SMART also warns when the drive has reached its limit on read/write cycles on the flash.
You are correct if nothing if nothing can be done to stop the errors. However, in my personal experience performing a hardware reset of an SSD can "fix" many SSD issues.
I recently had an Apple NVMe integrated SSD in an Apple Intel USB-C laptop show write errors (Actually "Media & Data Integrity" errors). I used a Linux utility to reset the SSD to factory defaults by using that SSD's built-in hardware secure erase feature. I was able to keep using this SSD and have not seen any more write errors since.
I have seen SSDs have trouble reallocating bad blocks which can cause some of the SSD's SMART attributes to degrade quickly & perhaps even trigger a failing status for the SSD, but that doesn't mean the SSD is actually bad. Again, I have been able to reset those SSDs which stopped the underlying problem to cause some SMART attributes to appear to degrade quickly. The SSD's firmware assumes that bad blocks get reallocated quickly, but when they don't reallocate it can cause a SMART attribute to appear worse than it actually is.
This same hardware reset can also improve the performance of an SSD, but at the loss of any existing data on the SSD.
In my experience, SSDs with SMART errors are no longer reliable and can experience very slow performance. By not reliable, I mean that data can be lost or corrupted at any time.
Yes, if an SSD is left untreated, then that will certainly happen.
However, I have found that utilizing an SSD's built-in hardware secure erase feature will reset an SSD's internal hardware to factory defaults thereby eliminating many of the SSD problems....a Crucial tech support agent informed me of this years ago and it has worked for non-Crucial SSDs as well including Apple's own SSDs. It has been extremely rare in my personal experience for an SSD to continue having issues after such a reset. The one time where this was insufficient was due to a firmware bug, but once that SSD's firmware had been updated & another hardware reset was performed....the SSD was perfectly fine (actually we had several of this exact model SSD with the problem).