How to Check SSD Health: The Four Values That Matter
SSD health is measured differently from hard drive health, and the number most people look at is not the one that matters. This covers what to check, what each value means, and when to replace.
How do I check SSD health?
Every SSD reports its own health data through the SMART interface, and the operating system or controller can read it.
Three routes.
The RAID controller or management interface. On a server, iLO, iDRAC, IMM or the controller utility reports drive health per bay. This is the route to use on a production server, since it does not require taking the drive offline.
Operating system tools. Standard SMART utilities read the drive directly, which gives more detail than most controller summaries.
Manufacturer utilities. Vendor-specific tools present the same underlying data with clearer labels, which helps because SMART attribute numbering varies between manufacturers.
Our drive health guide covers reading health data generally; this covers what differs on flash.
What to check on an SSD: the four values that matter
Percentage used, or remaining life. The headline number. It reports how much of the drive’s rated write endurance has been consumed. An SSD at 90% used has 10% of its rated writes left, regardless of how fast it still performs today.
Total bytes written (TBW) or data units written. The raw figure behind the percentage. Compare it against the drive’s rated endurance from the datasheet. Our endurance guide covers converting between DWPD and TBW.
Media and data integrity errors. Unrecoverable read errors. Should be zero. Any non-zero value on an enterprise SSD is worth investigating.
Available spare. SSDs reserve blocks to replace failed ones. As that reserve depletes, the drive is approaching end of life. A falling spare percentage is an early warning.
Why power-on hours mean little on an SSD
Because an SSD wears by writes, not by time.
A hard drive has moving parts that wear as they spin, so hours are a rough proxy for wear. An SSD has none. A drive that has been powered for five years doing almost no writing is in better condition than one powered for one year under heavy write load.
The practical consequence when buying used: ask for percentage used, not hours. A seller quoting only hours on an SSD has not understood what they are selling. Our hours guide covers the mechanical side, where hours do matter.
When should I replace an SSD?
Three triggers.
Percentage used approaching the rated limit. Plan the replacement before it arrives rather than after. Most enterprise SSDs continue working past 100% of rated endurance, but the manufacturer no longer stands behind them and failure risk rises.
Available spare falling. The drive is consuming its reserve blocks, which means media is failing.
Any media errors. On an enterprise SSD these should be zero.
And a fourth that is not a health value: the drive has gone read-only. Some SSDs enter read-only mode when endurance is exhausted, to preserve the data. That is the drive telling you it is finished.
Monitoring rather than checking
A one-off check tells you the state today; monitoring tells you the trend.
Set alerts on percentage used and available spare, and route them somewhere a person reads. On a server, the management controller can do this. Our monitoring guide covers why the alert path matters more than the sensor.
The value of monitoring on SSDs is that endurance consumption is predictable. If a drive has used 30% in a year, you know roughly when to plan its replacement, which is not true of mechanical drive failure.
Frequently asked questions
How do I check SSD health?
Through SMART data, read either from the RAID controller or management interface on a server, from operating system utilities, or from manufacturer tools. The controller route works without taking the drive offline.
What is the most important SSD health value?
Percentage used, or remaining life. It reports how much of the drive’s rated write endurance has been consumed, which is what actually limits an SSD’s life.
Do power-on hours matter for an SSD?
Very little. SSDs wear by writes, not time. A drive powered five years with light writing is in better condition than one powered one year under heavy write load. Ask for percentage used instead.
What is available spare on an SSD?
Reserve blocks the drive keeps to replace failed ones. A falling spare percentage means media is failing and is an early warning of end of life.
When should I replace an SSD?
When percentage used approaches the rated limit, when available spare is falling, or when any media errors appear. Plan the replacement before the limit rather than after.
Why has my SSD gone read-only?
Some SSDs enter read-only mode when write endurance is exhausted, to preserve the data already on them. Copy the data off and replace the drive.
Buying a used SSD from us? Ask for the percentage used and media error count. We check both before listing.
