SMART Disk Monitoring

Monitor disk health with AI-powered false positive reduction

Overview

SMART (Self-Monitoring, Analysis, and Reporting Technology) provides critical disk health information, but monitoring SMART reports can be challenging due to false positives. Telemetry.host’s AI-powered log analysis helps distinguish real problems from normal SMART metrics.

Basic SMART Monitoring

Simple Daily Check

#!/bin/bash
# smart-check.sh

# Run SMART self-test and get results
smartctl -a /dev/sda | curl -X POST \
  https://telemetry.host/ping/PROJECT_KEY/timeout/26h/disk-sda-health?create=1 \
  -H "Content-Type: text/plain" \
  --data-binary @-

Add to crontab:

# Daily SMART check at 3 AM
0 3 * * * /usr/local/bin/smart-check.sh

Multiple Disks

#!/bin/bash
# smart-check-all.sh

for disk in /dev/sd[a-z]; do
    [ -e "$disk" ] || continue
    
    DISK_NAME=$(basename "$disk")
    
    smartctl -a "$disk" | curl -X POST \
        "https://telemetry.host/ping/PROJECT_KEY/timeout/26h/disk-$DISK_NAME?create=1" \
        -H "Content-Type: text/plain" \
        --data-binary @-
done

Understanding SMART Metrics

SMART reports contain various attributes that look like errors but aren’t:

False Positive Examples

These are normal and don’t indicate problems:

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

General SMART Values:
Offline data collection status:  (0x82)	Offline data collection activity
					was completed without error.
					Auto Offline Data Collection: Enabled.
Self-test execution status:      (   0)	The previous self-test routine completed
					without error or no self-test has ever 
					been run.
Total time to complete Offline 
data collection: 		(  584) seconds.
Offline data collection
capabilities: 			 (0x7b) SMART execute Offline immediate.
					Auto Offline data collection on/off support.
					Suspend Offline collection upon new
					command.
					Offline surface scan supported.
					Self-test supported.
					Conveyance Self-test supported.
					Selective Self-test supported.
SMART capabilities:            (0x0003)	Saves SMART data before entering
					power-saving mode.
					Supports SMART auto save timer.
Error logging capability:        (0x01)	Error logging supported.
					General Purpose Logging supported.
Short self-test routine 
recommended polling time: 	 (   1) minutes.
Extended self-test routine
recommended polling time: 	 (1102) minutes.
Conveyance self-test routine
recommended polling time: 	 (   2) minutes.
SCT capabilities: 	       (0x50bd)	SCT Status supported.
					SCT Error Recovery Control supported.
					SCT Feature Control supported.
					SCT Data Table supported.

SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   081   064   044    Pre-fail  Always       -       134493480
  3 Spin_Up_Time            0x0003   091   091   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       34
  5 Reallocated_Sector_Ct   0x0033   100   100   010    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   072   060   045    Pre-fail  Always       -       14223679
  9 Power_On_Hours          0x0032   096   096   000    Old_age   Always       -       3623
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       7
184 End-to-End_Error        0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
188 Command_Timeout         0x0032   100   099   000    Old_age   Always       -       4295032833
190 Airflow_Temperature_Cel 0x0022   063   052   040    Old_age   Always       -       37 (Min/Max 21/45)
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       120
193 Load_Cycle_Count        0x0032   100   100   000    Old_age   Always       -       152
194 Temperature_Celsius     0x0022   037   048   000    Old_age   Always       -       37 (0 21 0 0 0)
195 Hardware_ECC_Recovered  0x001a   023   009   000    Old_age   Always       -       134493480
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x0023   100   100   001    Pre-fail  Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       2953 (7 183 0)
241 Total_LBAs_Written      0x0000   100   253   000    Old_age   Offline      -       10356940746
242 Total_LBAs_Read         0x0000   100   253   000    Old_age   Offline      -       1281966970

SMART Error Log Version: 1
No Errors Logged

SMART Self-test log structure revision number 1
No self-tests have been logged.  [To run self-tests, use: smartctl -t]

SMART Selective self-test log data structure revision number 1
 SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
    1        0        0  Not_testing
    2        0        0  Not_testing
    3        0        0  Not_testing
    4        0        0  Not_testing
    5        0        0  Not_testing
Selective self-test flags (0x0):
  After scanning selected spans, do NOT read-scan remainder of disk.
If Selective self-test is pending on power-up, resume after 0 minute delay.

The names contain “error”, “fail”, and “uncorrectable” but the RAW_VALUE of 0 means no issues.

AI-Powered False Positive Reduction

Telemetry.host’s AI understands SMART output context and won’t alert on these patterns. It analyzes:

• Attribute names vs actual values
• Threshold violations vs normal reporting
• Historical trends vs one-time spikes

What Triggers Real Alerts

The AI looks for genuine problems:

# Reallocated sectors increasing
  5 Reallocated_Sector_Ct   0x0033   095   095   010    Pre-fail  Always       -       42

# Current pending sectors (disk waiting to reallocate)
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       8

# SMART overall health failing
SMART overall-health self-assessment test result: FAILING_NOW

# Temperature too high
194 Temperature_Celsius     0x0022   045   055   000    Old_age   Always       -       55 (Max 70)

Advanced Monitoring Patterns

With Error Detection

Add basic error detection before AI analysis:

#!/bin/bash
# smart-check-errors.sh

DISK="/dev/sda"
OUTPUT=$(smartctl -a "$DISK")

# Check overall health
if echo "$OUTPUT" | grep -q "PASSED"; then
    STATUS="success"
    MESSAGE="Disk health check passed"
else
    STATUS="warning"
    MESSAGE="Disk health check did not pass - check details"
fi

# Send full output for AI analysis
echo "$OUTPUT" | curl -X POST \
    https://telemetry.host/ping/YOUR_MONITOR_ID \
    -H "Content-Type: application/json" \
    -d "{
        \"status\": \"$STATUS\",
        \"message\": \"$MESSAGE\"
    }" \
    --data-binary @-

NVMe Disks

NVMe disks have different SMART attributes:

#!/bin/bash
# nvme-smart-check.sh

DISK="/dev/nvme0n1"

# Get NVMe SMART data
nvme smart-log "$DISK" | curl -X POST \
    https://telemetry.host/ping/PROJECT_KEY/timeout/26h/nvme-health?create=1 \
    -H "Content-Type: text/plain" \
    --data-binary @-

RAID Arrays

Monitor RAID health alongside individual disks:

#!/bin/bash
# raid-smart-check.sh

# Check RAID status
RAID_STATUS=$(cat /proc/mdstat)

# Check individual disks in RAID
for disk in $(grep -o 'sd[a-z]' /proc/mdstat | sort -u); do
    smartctl -a "/dev/$disk" >> "/tmp/raid-smart.log"
done

# Send combined report
{
    echo "=== RAID Status ==="
    echo "$RAID_STATUS"
    echo ""
    echo "=== Individual Disk SMART ==="
    cat "/tmp/raid-smart.log"
} | curl -X POST \
    https://telemetry.host/ping/YOUR_MONITOR_ID \
    -H "Content-Type: text/plain" \
    --data-binary @-

rm "/tmp/raid-smart.log"

Monitoring Specific Attributes

Critical Attributes to Watch

#!/bin/bash
# smart-critical-attrs.sh

DISK="/dev/sda"

# Get specific critical attributes
REALLOCATED=$(smartctl -A "$DISK" | grep Reallocated_Sector_Ct | awk '{print $10}')
PENDING=$(smartctl -A "$DISK" | grep Current_Pending_Sector | awk '{print $10}')
UNCORRECTABLE=$(smartctl -A "$DISK" | grep Offline_Uncorrectable | awk '{print $10}')

# Check if any are non-zero
if [ "$REALLOCATED" -gt 0 ] || [ "$PENDING" -gt 0 ] || [ "$UNCORRECTABLE" -gt 0 ]; then
    STATUS="warning"
    MESSAGE="Disk showing signs of wear: Reallocated=$REALLOCATED, Pending=$PENDING, Uncorrectable=$UNCORRECTABLE"
else
    STATUS="success"
    MESSAGE="All critical SMART attributes normal"
fi

curl -X POST https://telemetry.host/ping/YOUR_MONITOR_ID \
    -H "Content-Type: application/json" \
    -d "{
        \"status\": \"$STATUS\",
        \"message\": \"$MESSAGE\",
        \"metadata\": {
            \"reallocated_sectors\": $REALLOCATED,
            \"pending_sectors\": $PENDING,
            \"uncorrectable_sectors\": $UNCORRECTABLE
        }
    }"

Temperature Monitoring

#!/bin/bash
# disk-temp-check.sh

DISK="/dev/sda"
TEMP=$(smartctl -A "$DISK" | grep Temperature_Celsius | awk '{print $10}')

if [ "$TEMP" -gt 50 ]; then
    STATUS="warning"
    MESSAGE="Disk temperature high: ${TEMP}°C"
elif [ "$TEMP" -gt 60 ]; then
    STATUS="error"
    MESSAGE="Disk temperature critical: ${TEMP}°C"
else
    STATUS="success"
    MESSAGE="Disk temperature normal: ${TEMP}°C"
fi

curl -X POST https://telemetry.host/ping/YOUR_MONITOR_ID \
    -d "{\"status\":\"$STATUS\",\"message\":\"$MESSAGE\"}"

Running SMART Tests

Short Test

# Start short test (takes 1-2 minutes)
smartctl -t short /dev/sda

# Wait for completion
sleep 120

# Get results
smartctl -l selftest /dev/sda | curl -X POST \
    https://telemetry.host/ping/YOUR_MONITOR_ID \
    -H "Content-Type: text/plain" \
    --data-binary @-

Long Test (Weekly)

#!/bin/bash
# weekly-smart-long-test.sh

DISK="/dev/sda"

# Start long test
smartctl -t long "$DISK"

# Long test takes hours - check status
echo "SMART long test started for $DISK"

# Schedule result check for later (e.g., 6 hours)
at now + 6 hours <<EOF
smartctl -l selftest "$DISK" | curl -X POST \
    https://telemetry.host/ping/YOUR_MONITOR_ID \
    -H "Content-Type: text/plain" \
    --data-binary @-
EOF

Add to crontab for weekly runs:

# Weekly long SMART test on Sunday at 2 AM
0 2 * * 0 /usr/local/bin/weekly-smart-long-test.sh

Cloud/VM Considerations

AWS EBS Volumes

AWS doesn’t expose SMART data for EBS volumes. Monitor using CloudWatch metrics instead or use instance store volumes.

Other Cloud Providers

Check if SMART data is available:

# Test if SMART is available
if smartctl -i /dev/sda > /dev/null 2>&1; then
    echo "SMART available"
else
    echo "SMART not available on this VM"
fi

Best Practices

1. Daily Monitoring

Run daily checks to catch degrading disks early:

0 3 * * * /usr/local/bin/smart-check.sh

2. Let AI Handle False Positives

Send full SMART output - the AI will filter out false positives:

# ✅ Good: Send full output for AI analysis
smartctl -a /dev/sda | curl -X POST {MONITOR_URL}

# ❌ Bad: Try to filter yourself
smartctl -a /dev/sda | grep -v "Error" | curl -X POST {MONITOR_URL}

3. Monitor All Disks

Don’t forget about secondary disks:

for disk in $(lsblk -d -n -o NAME | grep -E '^sd|^nvme'); do
    smartctl -a "/dev/$disk" | curl ...
done

4. Track Trends Over Time

The monitoring dashboard shows historical trends to spot gradual degradation.

5. Set Up Notifications

Configure Discord or email notifications for disk warnings - these are time-sensitive.

Troubleshooting

smartctl Command Not Found

Install smartmontools:

# Ubuntu/Debian
sudo apt-get install smartmontools

# RHEL/CentOS
sudo yum install smartmontools

# macOS
brew install smartmontools

Permission Denied

Run with sudo or add to sudoers:

# Run with sudo in cron
0 3 * * * sudo /usr/local/bin/smart-check.sh

# Or add to sudoers (be careful!)
echo "username ALL=(ALL) NOPASSWD: /usr/sbin/smartctl" | sudo tee -a /etc/sudoers.d/smartctl

SMART Not Supported

Some disks don’t support SMART:

smartctl -i /dev/sda | grep "SMART support"

If not supported, consider monitoring at a different level (RAID controller, etc.)

Example Real-World Alerts

✅ True Positive (Real Problem)

Critical Warning: Reallocated sector count increased from 5 to 47 in the last week.
Action: Replace disk immediately, backup data.

❌ False Positive (Not a Problem)

SMART Attribute: Offline_Uncorrectable
Current Value: 100
Threshold: 0
Raw Value: 0

AI Analysis: This is normal. Attribute name contains "uncorrectable" but raw value 
is 0, indicating no actual uncorrectable sectors.

Next Steps

• Learn about Notifications
• See backup monitoring examples
• Explore cron job monitoring