Monitor - Drive SMART Values

Import-Module $env:SyncroModule -DisableNameChecking
#Requires -Version 3.0

<# Version 1.1
.DESCRIPTION
Obtain and evaluate SMART data from drives and alert based on ID value thresholds.
Uses WMI so no outside tools are needed but ome hardware/configuration may not be compatible.

The threshold values used below are guesstimates as to what might be concerning. Sometimes
manufacturers use the same attribute number for different purposes, calculate things differently,
or use Worst vs RawValue. If you come across these exceptions, please report them or do a pull request
so the script can improve.  Don't take any one value as critical and adjust as you feel appropreiate for
your environment and risk level, or even remove attributes from monitoring entirely.

.NOTES
Created by Isaac Good with data processing function from https://github.com/Fantastitech/GetSmartWin/blob/master/getsmartwin.ps1
    v1.0 - 2018-11-05
    v1.1 - 2022-02-25
        Added - description, changlog and comments
        Added - if statements for SSD and spinner specific IDs
        Changed - split out ID 233 for Intel (and maybe others?) as they use Worst value instead of RawValue
#>

# If this is a virtual machine, we don't need to continue
$Computer = Get-CimInstance -ClassName 'Win32_ComputerSystem'
if ($Computer.Model -like 'Virtual*') {
    exit
}

# Clear our variable so we have good output when testing
$Warnings = @()

# Grab the instance names for disks that have WMI SMART data available
# This is where the "Not supported" error can be generated from, which we check for later
$disks = (Get-CimInstance -Namespace 'Root\WMI' -ClassName 'MSStorageDriver_FailurePredictStatus' | Select-Object 'InstanceName').InstanceName

foreach ($disk in $disks) {
    # Since MSStorageDriver_FailurePredictStatus and Get-PhysicalDisk don't share any values to
    # uniquely identify a disk so we can check if SSD, we have to go through an intermediate step
    # of getting the serial number to match against. Maybe there's a better way that I'm missing. #>

    # Strip instance name down to a value we can match against
    $product = [regex]::Matches($disk,"(?<=Prod_).*(?=\\)").Value

    # Grab the serial for the disk that matches the instance name
    $serial = (Get-CimInstance -ClassName Win32_DiskDrive | Where-Object { $_.PNPDeviceID -match "$product"}).SerialNumber

    # Use the serial to check if the disk is an SSD
    $ssd = Get-PhysicalDisk | Where-Object { $_.SerialNumber -eq $serial } | Where-Object -Property MediaType -eq SSD

    # Retrieve SMART data
    $SmartData = (Get-CimInstance -Namespace 'Root\WMI' -ClassName 'MSStorageDriver_ATAPISMartData' | Where-Object 'InstanceName' -eq $disk)
    [Byte[]]$RawSmartData = $SmartData | Select-Object -ExpandProperty 'VendorSpecific'

    # Starting at the third number (first two are irrelevant)
    # get the relevant data by iterating over every 12th number
    # and saving the values from an offset of the SMART attribute ID
    [PSCustomObject[]]$Output = for ($i = 2; $i -lt $RawSmartData.Count; $i++) {
        if (0 -eq ($i - 2) % 12 -and $RawSmartData[$i] -ne 0) {
            # Construct the raw attribute value by combining the two bytes that make it up
            [Decimal]$RawValue = ($RawSmartData[$i + 6] * [Math]::Pow(2, 8) + $RawSmartData[$i + 5])
            $InnerOutput = [PSCustomObject]@{
                DiskID   = $disk
                ID       = $RawSmartData[$i]
                #Flags    = $RawSmartData[$i + 1]
                #Value    = $RawSmartData[$i + 3]
                Worst    = $RawSmartData[$i + 4]
                RawValue = $RawValue
            }
            $InnerOutput
        }
    }

    # Display table of values
    $Output | Sort-Object ID | Out-String

    # Spinner-specific ID's
    if (-not $ssd) {
        # Spin Retry Count
        $Warnings += $Output | Where-Object ID -eq 10 | Where-Object RawValue -ne 0 | Format-Table

        # Recalibration Retries
        $Warnings += $Output | Where-Object ID -eq 11 | Where-Object RawValue -ne 0 | Format-Table

        # High Fly Writes
        $Warnings += $Output | Where-Object ID -eq 189 | Where-Object RawValue -ne 0 | Format-Table
    }

    # Reallocated Sectors Count / Retired NAND Blocks
    $Warnings += $Output | Where-Object ID -eq 5 | Where-Object RawValue -gt 1 | Format-Table

    # Spare Blocks Available
    $Warnings += $Output | Where-Object ID -eq 16 | Where-Object RawValue -lt 100 | Format-Table

    # Remaining Spare Blocks
    $Warnings += $Output | Where-Object ID -eq 17 | Where-Object RawValue -lt 100 | Format-Table

    # Unexpected Power Loss Count (Unsafe Shutdowns Count on PCIe SSDs)
    $Warnings += $Output | Where-Object ID -eq 174 | Where-Object RawValue -gt 300 | Format-Table

    # Used Reserved Block Count Total
    $Warnings += $Output | Where-Object ID -eq 178 | Where-Object RawValue -gt 1 | Format-Table

    # Used Reserved Block Count Total
    $Warnings += $Output | Where-Object ID -eq 179 | Where-Object RawValue -gt 1 | Format-Table

    # Unused Reserved Block Count or Percentage Remaining (Available Spare Blocks on PCIe SSDs)
    $Warnings += $Output | Where-Object ID -eq 180 | Where-Object RawValue -lt 30 | Format-Table

    # Erase Failure Count
    $Warnings += $Output | Where-Object ID -eq 182 | Where-Object RawValue -gt 1 | Format-Table

    # SATA Downshift Error Count or Runtime Bad Block (may indicate bad cable)
    $Warnings += $Output | Where-Object ID -eq 183 | Where-Object RawValue -gt 2 | Format-Table

    # End-to-End error / IOEDC
    $Warnings += $Output | Where-Object ID -eq 184 | Where-Object RawValue -ne 0 | Format-Table

    # Reported Uncorrectable Errors
    $Warnings += $Output | Where-Object ID -eq 187 | Where-Object RawValue -gt 1 | Format-Table

    # Command Timeout
    $Warnings += $Output | Where-Object ID -eq 188 | Where-Object RawValue -gt 1 | Format-Table

    # Temperature Celcius
    $Warnings += $Output | Where-Object ID -eq 194 | Where-Object RawValue -gt 50 | Format-Table

    # Reallocation Event Count
    $Warnings += $Output | Where-Object ID -eq 196 | Where-Object RawValue -gt 1 | Format-Table

    # Current Pending Sector Count
    $Warnings += $Output | Where-Object ID -eq 197 | Where-Object RawValue -ne 0 | Format-Table

    # Uncorrectable Sector Count
    $Warnings += $Output | Where-Object ID -eq 198 | Where-Object RawValue -ne 0 | Format-Table

    # UltraDMA CRC Error Count
    $Warnings += $Output | Where-Object ID -eq 199 | Where-Object RawValue -ne 0 | Format-Table

    # Soft Read Error Rate
    $Warnings += $Output | Where-Object ID -eq 201 | Where-Object Worst -lt 95 | Format-Table

    # SSD-specific ID's
    if ($ssd) {
        # SSD Percentage Lifetime Used % (Crucial/Micron)
        $Warnings += $Output | Where-Object ID -eq 202 | Where-Object Worst -gt 95 | Format-Table

        # SSD RAIN Successful Recovery Page Count (Crucial Brand)
        $Warnings += $Output | Where-Object ID -eq 210 | Where-Object Worst -gt 10 | Format-Table

        # SSD Life Left %
        $Warnings += $Output | Where-Object ID -eq 231 | Where-Object Worst -lt 20 | Format-Table

        # SSD Endurance Remaining or Available Reserved Space %
        $Warnings += $Output | Where-Object ID -eq 232 | Where-Object Worst -lt 20 | Format-Table

        # SSD Media Wear Out Indicator %
        if ($disk -match 'intel') {
            $Warnings += $Output | Where-Object ID -eq 233 | Where-Object Worst -lt 20 | Format-Table
        }
        else {
            $Warnings += $Output | Where-Object ID -eq 233 | Where-Object RawValue -lt 20 | Format-Table
        }
    }
} # End foreach

# Check FailurePredictStatus
$Warnings += Get-CimInstance -Namespace 'Root\WMI' -ClassName 'MSStorageDriver_FailurePredictStatus' |
    Select-Object InstanceName, PredictFailure, Reason |
    Where-Object {$_.PredictFailure -ne $False} | Format-Table

# Check general disk status and health
$Warnings += Get-PhysicalDisk |
    Select-Object FriendlyName, Size, MediaType, OperationalStatus, HealthStatus |
    Where-Object {$_.OperationalStatus -ne 'OK' -or $_.HealthStatus -ne 'Healthy'} | Format-Table

if ($Warnings) {
    $Warnings = $warnings | Out-String
    $Warnings
    Rmm-Alert -Category 'Monitor - Drive SMART Values' -Body "$Warnings"
    Exit 1
}

if ($Error) {
    if ($Error -match "Not supported") {
        ($notsup = "Not supported: You may need to switch from ACHI to RAID/RST mode, see the link for how to do this non-destructively: https://www.top-password.com/blog/switch-from-raid-to-ahci-without-reinstalling-windows/")
        Rmm-Alert -Category 'Monitor - Drive SMART Values' -Body "$notsup"
    }
    Rmm-Alert -Category 'Monitor - Drive SMART Values' -Body "$error"
    exit 1
}

Close-Rmm-Alert -Category "Monitor - Drive SMART Values"