Files
disk-health/old/harvester-v2.3.sh
2025-10-22 15:02:09 +08:00

531 lines
18 KiB
Bash

#!/bin/bash
# Disk Health Check Script for Harvester OS
# Checks SATA HDD, SATA SSD, SAS, NVMe, RAID controllers, and soft-raid
# Supports consumer and enterprise disk classification
SCRIPT_NAME=$(basename "$0")
VERSION="2.3"
# Color codes
RED=$(tput setaf 1)
GREEN=$(tput setaf 2)
YELLOW=$(tput setaf 3)
BLUE=$(tput setaf 4)
CYAN=$(tput setaf 6)
MAGENTA=$(tput setaf 5)
NC=$(tput sgr0)
# Function to print colored output
print_color() {
local color=$1
local message=$2
echo -e "${color}${message}${NC}"
}
# Check if smartctl is installed
command_exists() {
command -v "$1" >/dev/null 2>&1
}
if ! command_exists smartctl; then
print_color $RED "Error: smartctl is not installed. Please install smartmontools package."
exit 1
fi
# TBW endurance standards (using lowest numbers)
declare -A CONSUMER_TBW=(
["250"]=150
["500"]=300
["1000"]=600
["2000"]=1200
["4000"]=2400
)
declare -A ENTERPRISE_TBW=(
["250"]=450
["500"]=900
["1000"]=1800
["2000"]=3600
["4000"]=7200
)
# Function to get closest capacity tier
get_capacity_tier() {
local capacity_gb=$1
local tiers=("250" "500" "1000" "2000" "4000")
for tier in "${tiers[@]}"; do
if [[ $capacity_gb -le $tier ]]; then
echo $tier
return
fi
done
# For larger than 4TB, use 4TB tier with scaling
echo "4000"
}
# Function to get disk type and interface
get_disk_info() {
local disk=$1
local controller=$2
local smart_cmd="smartctl"
[[ -n "$controller" ]] && smart_cmd+=" -d $controller"
smart_cmd+=" -i $disk"
local info=$($smart_cmd 2>/dev/null)
local transport=""
local disk_type="UNKNOWN"
local is_enterprise=false
# Check if it's NVMe
if [[ "$disk" == /dev/nvme* ]] || echo "$info" | grep -qi "NVMe"; then
disk_type="NVMe"
transport="NVMe"
# Check for SAS
elif echo "$info" | grep -qi "SAS"; then
disk_type="SAS"
transport="SAS"
is_enterprise=true
# Check for SATA SSD
elif echo "$info" | grep -qi "Solid State Device"; then
disk_type="SSD"
transport="SATA"
# Check for SATA HDD
elif echo "$info" | grep -qi "Rotation Rate"; then
disk_type="HDD"
transport="SATA"
fi
# Check for enterprise features
if echo "$info" | grep -qi "ENTERPRISE\|EP\|SAS\|Xeon\|Xeons\|DualPort\|PowerLoss\|PLP"; then
is_enterprise=true
fi
# Check device type by model name
local model=$(echo "$info" | grep -i "Device Model:" | cut -d: -f2 | sed 's/^[ \t]*//')
if echo "$model" | grep -qi "PRO\|EP\|DC\|ENT\|ENTERPRISE\|MX500\|870QVO\|860QVO\|Crucial\|Samsung\|Intel"; then
if echo "$model" | grep -qi "PRO\|EP\|DC\|ENT\|ENTERPRISE"; then
is_enterprise=true
fi
fi
echo "$disk_type|$transport|$is_enterprise"
}
# Function to calculate TBW for SSD/NVMe
calculate_tbw() {
local disk_type=$1
local raw_value=$2
local sectors=$3
local tbw=0
if [[ -n "$sectors" && "$sectors" != "0" ]]; then
# Calculate from sectors (most common for SATA SSDs)
local bytes=$((sectors * 512))
tbw=$(echo "scale=2; $bytes / 1000 / 1000 / 1000 / 1000" | bc -l 2>/dev/null || echo "0")
elif [[ -n "$raw_value" && "$raw_value" != "0" ]]; then
if [[ "$disk_type" == "NVMe" ]]; then
# NVMe: raw value is in 32MB units
tbw=$(echo "scale=2; $raw_value * 32 / 1000 / 1000" | bc -l 2>/dev/null || echo "0")
else
# SATA SSD: various manufacturers
tbw=$(echo "scale=2; $raw_value * 32 / 1000 / 1000" | bc -l 2>/dev/null || echo "0")
fi
fi
echo "$tbw"
}
# Function to get estimated endurance
get_estimated_endurance() {
local capacity_gb=$1
local is_enterprise=$2
local disk_type=$3
# HDDs don't have TBW
if [[ "$disk_type" == "HDD" ]]; then
echo "N/A"
return
fi
local capacity_tier=$(get_capacity_tier "$capacity_gb")
if [[ "$is_enterprise" == "true" ]]; then
echo "${ENTERPRISE_TBW[$capacity_tier]}"
else
echo "${CONSUMER_TBW[$capacity_tier]}"
fi
}
# Function to estimate SSD lifespan with TBW remaining
estimate_ssd_lifespan() {
local power_on_hours=$1
local tbw_used=$2
local estimated_endurance=$3
local disk_type=$4
if [[ -z "$power_on_hours" || "$power_on_hours" -eq 0 ]]; then
echo "Unknown||Unknown|New"
return
fi
if [[ "$estimated_endurance" == "N/A" ]]; then
echo "N/A|N/A|N/A|HDD"
return
fi
local tbw_remaining=$(echo "scale=2; $estimated_endurance - $tbw_used" | bc -l 2>/dev/null || echo "0")
if [[ $(echo "$tbw_used > 0" | bc -l 2>/dev/null) -eq 1 ]]; then
local lifespan_used=$(echo "scale=1; $tbw_used * 100 / $estimated_endurance" | bc -l 2>/dev/null || echo "0")
local lifespan_remaining=$(echo "scale=1; 100 - $lifespan_used" | bc -l 2>/dev/null || echo "100")
if [[ $(echo "$lifespan_used >= 80" | bc -l) -eq 1 ]]; then
echo "${RED}${lifespan_remaining}%${NC}|${RED}${tbw_remaining} TB${NC}|High wear|$disk_type"
elif [[ $(echo "$lifespan_used >= 50" | bc -l) -eq 1 ]]; then
echo "${YELLOW}${lifespan_remaining}%${NC}|${YELLOW}${tbw_remaining} TB${NC}|Moderate wear|$disk_type"
else
echo "${GREEN}${lifespan_remaining}%${NC}|${GREEN}${tbw_remaining} TB${NC}|Healthy|$disk_type"
fi
else
echo "Unknown|${estimated_endurance} TB|New|$disk_type"
fi
}
# Function to estimate HDD lifespan
estimate_hdd_lifespan() {
local power_on_hours=$1
local reallocated_sectors=$2
local pending_sectors=$3
local start_stop_count=$4
local load_cycle_count=$5
local disk_type=$6
if [[ -z "$power_on_hours" ]]; then
echo "Unknown"
return
fi
local severity=0
# Critical issues
if [[ "$pending_sectors" -gt 0 ]]; then
echo "${RED}CRITICAL${NC} (Pending sectors: $pending_sectors)"
return
elif [[ "$reallocated_sectors" -gt 100 ]]; then
severity=$((severity + 3))
elif [[ "$reallocated_sectors" -gt 10 ]]; then
severity=$((severity + 2))
elif [[ "$reallocated_sectors" -gt 0 ]]; then
severity=$((severity + 1))
fi
# Usage-based assessment
if [[ "$power_on_hours" -gt 50000 ]]; then
severity=$((severity + 3))
elif [[ "$power_on_hours" -gt 30000 ]]; then
severity=$((severity + 2))
elif [[ "$power_on_hours" -gt 15000 ]]; then
severity=$((severity + 1))
fi
# Mechanical wear (for HDDs)
if [[ "$disk_type" == "HDD" ]]; then
if [[ "$start_stop_count" -gt 50000 ]]; then
severity=$((severity + 2))
elif [[ "$start_stop_count" -gt 20000 ]]; then
severity=$((severity + 1))
fi
if [[ "$load_cycle_count" -gt 500000 ]]; then
severity=$((severity + 2))
elif [[ "$load_cycle_count" -gt 200000 ]]; then
severity=$((severity + 1))
fi
fi
if [[ $severity -ge 5 ]]; then
echo "${RED}< 6 months${NC} (Multiple risk factors)"
elif [[ $severity -ge 3 ]]; then
echo "${YELLOW}6-18 months${NC} (Moderate wear)"
elif [[ $severity -ge 1 ]]; then
echo "${YELLOW}1-3 years${NC} (Light wear)"
else
echo "${GREEN}> 3 years${NC} (Healthy)"
fi
}
# Function to check soft-raid (MDRAID)
check_mdraid() {
local md_devices=()
if [[ -f /proc/mdstat ]]; then
while IFS= read -r line; do
if [[ $line =~ ^md[0-9]+ ]]; then
md_devices+=("/dev/${line%% *}")
fi
done < /proc/mdstat
fi
for md in "${md_devices[@]}"; do
if [[ -b "$md" ]]; then
print_color $MAGENTA "Found software RAID: $md"
local md_info=$(mdadm --detail "$md" 2>/dev/null)
if [[ -n "$md_info" ]]; then
echo "RAID Level: $(echo "$md_info" | grep "Raid Level" | cut -d: -f2 | sed 's/^[ \t]*//')"
echo "State: $(echo "$md_info" | grep "State" | head -1 | cut -d: -f2 | sed 's/^[ \t]*//')"
echo "Devices: $(echo "$md_info" | grep "Active Devices" | cut -d: -f2 | sed 's/^[ \t]*//')"
echo ""
fi
fi
done
}
# Function to check a single disk
check_disk() {
local disk=$1
local controller=$2
local smart_cmd="smartctl"
[[ -n "$controller" ]] && smart_cmd+=" -d $controller"
print_color $CYAN "Checking disk: $disk (Controller: ${controller:-direct})"
echo "=================================================="
# Get disk information
local disk_info=$(get_disk_info "$disk" "$controller")
local disk_type=$(echo "$disk_info" | cut -d'|' -f1)
local transport=$(echo "$disk_info" | cut -d'|' -f2)
local is_enterprise=$(echo "$disk_info" | cut -d'|' -f3)
# Get basic disk information
local info=$($smart_cmd -i "$disk" 2>/dev/null)
local health=$($smart_cmd -H "$disk" 2>/dev/null)
local attributes=$($smart_cmd -A "$disk" 2>/dev/null)
# Check if we can read the disk
if [[ $? -ne 0 ]] || [[ -z "$info" ]]; then
print_color $YELLOW "Cannot read disk information. It may be offline, unsupported, or need controller specification."
echo ""
return
fi
# Extract disk information
local model=$(echo "$info" | grep "Device Model:" | cut -d: -f2 | sed 's/^[ \t]*//')
[[ -z "$model" ]] && model=$(echo "$info" | grep "Model Number:" | cut -d: -f2 | sed 's/^[ \t]*//')
local serial=$(echo "$info" | grep "Serial Number:" | cut -d: -f2 | sed 's/^[ \t]*//')
local capacity=$(echo "$info" | grep "User Capacity:" | cut -d: -f2 | sed 's/^[ \t]*//' | cut -d'[' -f1)
local firmware=$(echo "$info" | grep "Firmware Version:" | cut -d: -f2 | sed 's/^[ \t]*//')
# Extract capacity in GB
local capacity_gb=0
if [[ $capacity =~ \[([0-9.]+)\s+GB\] ]]; then
capacity_gb=${BASH_REMATCH[1]%.*}
elif [[ $capacity =~ \[([0-9.]+)\s+TB\] ]]; then
capacity_gb=$(echo "${BASH_REMATCH[1]} * 1000" | bc -l 2>/dev/null | cut -d. -f1)
elif [[ $capacity =~ ([0-9,]+)\s+bytes ]]; then
local bytes=$(echo "${BASH_REMATCH[1]}" | tr -d ',')
capacity_gb=$(echo "scale=0; $bytes / 1000 / 1000 / 1000" | bc -l 2>/dev/null | cut -d. -f1)
fi
local health_status=$(echo "$health" | grep "result:" | cut -d: -f2 | sed 's/^[ \t]*//')
[[ -z "$health_status" ]] && health_status=$(echo "$health" | grep "SMART overall-health" | cut -d: -f2 | sed 's/^[ \t]*//')
# Extract SMART attributes
local power_on_hours=$(echo "$attributes" | grep "Power_On_Hours" | awk '{print $10}')
local reallocated_sectors=$(echo "$attributes" | grep "Reallocated_Sector_Ct" | awk '{print $10}')
local pending_sectors=$(echo "$attributes" | grep "Current_Pending_Sector" | awk '{print $10}')
local total_written=$(echo "$attributes" | grep -E "Total_LBAs_Written|Host_Writes_32MiB" | awk '{print $10}')
local host_writes_32mib=$(echo "$attributes" | grep "Host_Writes_32MiB" | awk '{print $10}')
local start_stop_count=$(echo "$attributes" | grep "Start_Stop_Count" | awk '{print $10}')
local load_cycle_count=$(echo "$attributes" | grep "Load_Cycle_Count" | awk '{print $10}')
# For NVMe disks using smartctl
if [[ "$disk_type" == "NVMe" ]]; then
local nvme_attributes=$($smart_cmd -x "$disk" 2>/dev/null)
power_on_hours=$(echo "$nvme_attributes" | grep "Power On Hours" | awk '{print $4}')
total_written=$(echo "$nvme_attributes" | grep "Data Units Written" | awk '{print $5}' | tr -d ',')
[[ -z "$total_written" ]] && total_written=$(echo "$nvme_attributes" | grep "Host_Writes_32MiB" | awk '{print $10}')
fi
# Display basic information
echo "Model: ${model:-Unknown}"
echo "Serial: ${serial:-Unknown}"
echo "Type: $disk_type"
echo "Interface: $transport"
echo "Class: $($is_enterprise && echo "Enterprise" || echo "Consumer")"
echo "Capacity: ${capacity:-Unknown}"
echo "Firmware: ${firmware:-Unknown}"
echo "Health: ${health_status:-Unknown}"
echo "Power On Hours: ${power_on_hours:-Unknown}"
# Disk type specific analysis
if [[ "$disk_type" == "HDD" ]]; then
echo "Realloc Sectors: ${reallocated_sectors:-0}"
echo "Pending Sectors: ${pending_sectors:-0}"
[[ -n "$start_stop_count" ]] && echo "Start/Stop Count: ${start_stop_count:-0}"
[[ -n "$load_cycle_count" ]] && echo "Load Cycle Count: ${load_cycle_count:-0}"
local lifespan=$(estimate_hdd_lifespan "$power_on_hours" "${reallocated_sectors:-0}" "${pending_sectors:-0}" "${start_stop_count:-0}" "${load_cycle_count:-0}" "$disk_type")
echo "Lifespan: $lifespan"
elif [[ "$disk_type" == "SSD" || "$disk_type" == "NVMe" || "$disk_type" == "SAS" ]]; then
local tbw_used=0
if [[ -n "$total_written" && "$total_written" != "0" ]]; then
tbw_used=$(calculate_tbw "$disk_type" "" "$total_written")
elif [[ -n "$host_writes_32mib" && "$host_writes_32mib" != "0" ]]; then
tbw_used=$(calculate_tbw "$disk_type" "$host_writes_32mib" "")
fi
local estimated_endurance=$(get_estimated_endurance "$capacity_gb" "$is_enterprise" "$disk_type")
echo "TBW Used: ${tbw_used} TB"
echo "TBW Endurance: ${estimated_endurance} TB"
local lifespan_info=$(estimate_ssd_lifespan "$power_on_hours" "$tbw_used" "$estimated_endurance" "$disk_type")
local lifespan_percent=$(echo "$lifespan_info" | cut -d'|' -f1)
local tbw_remaining=$(echo "$lifespan_info" | cut -d'|' -f2)
local wear_status=$(echo "$lifespan_info" | cut -d'|' -f3)
if [[ "$estimated_endurance" != "N/A" ]]; then
echo "TBW Remaining: $tbw_remaining"
echo "Lifespan: $lifespan_percent ($wear_status)"
fi
# Show mechanical attributes for SAS drives that might be SSDs
if [[ "$disk_type" == "SAS" ]]; then
echo "Realloc Sectors: ${reallocated_sectors:-0}"
echo "Pending Sectors: ${pending_sectors:-0}"
fi
else
print_color $YELLOW "Unknown disk type - limited information available"
fi
echo ""
}
# Function to detect RAID controllers and disks
detect_raid_disks() {
local controllers=("megaraid" "cciss" "areca" "3ware" "hpt" "aacraid" "auto")
local disks=()
# Check for direct disks first (SATA/SAS/NVMe)
for disk in /dev/sd[a-z] /dev/sd[a-z][a-z] /dev/nvme[0-9]n[0-9]; do
if [[ -b "$disk" ]]; then
disks+=("$disk:direct")
fi
done
# Check for RAID controllers
for controller in "${controllers[@]}"; do
for i in {0..31}; do
# Try different device patterns
for base_disk in "/dev/sda" "/dev/sg$i" "/dev/sr$i"; do
if smartctl -d "$controller,$i" -i "$base_disk" >/dev/null 2>&1; then
disks+=("$base_disk:$controller,$i")
break
fi
done
done
done
# Check for JBOD/passthrough disks on MegaRAID
if command_exists storcli; then
local jbod_disks=$(storcli /c0/eALL/sALL show all 2>/dev/null | grep -i "jbod\|unconfigured" | awk '{print $2}')
for disk in $jbod_disks; do
disks+=("$disk:megaraid,$disk")
done
fi
echo "${disks[@]}"
}
# Main function
main() {
print_color $BLUE "Disk Health Check Script v$VERSION for Harvester OS"
print_color $BLUE "===================================================="
echo ""
local disks=()
# Check for soft-raid first
check_mdraid
# If specific disk provided, check only that disk
if [[ $# -gt 0 ]]; then
for disk in "$@"; do
if [[ -b "$disk" ]]; then
disks+=("$disk:direct")
else
print_color $RED "Error: $disk is not a valid block device"
fi
done
else
# Auto-detect disks
print_color $CYAN "Auto-detecting disks..."
read -ra disks <<< "$(detect_raid_disks)"
fi
if [[ ${#disks[@]} -eq 0 ]]; then
print_color $YELLOW "No disks found via auto-detection"
print_color $CYAN "Trying direct disk access..."
# Try direct access to common disks
for disk in /dev/sda /dev/sdb /dev/sdc /dev/nvme0n1; do
if [[ -b "$disk" ]]; then
disks+=("$disk:direct")
fi
done
fi
if [[ ${#disks[@]} -eq 0 ]]; then
print_color $RED "No disks found or accessible"
echo "Try running as root or specifying disk paths manually"
exit 1
fi
print_color $GREEN "Found ${#disks[@]} disk(s) to check"
echo ""
# Check each disk
for disk_info in "${disks[@]}"; do
IFS=':' read -r disk controller <<< "$disk_info"
check_disk "$disk" "$controller"
done
print_color $BLUE "Check completed!"
}
# Usage information
usage() {
echo "Usage: $SCRIPT_NAME [DISK1 DISK2 ...]"
echo ""
echo "If no disks specified, auto-detects all available disks and RAID arrays"
echo ""
echo "Examples:"
echo " $SCRIPT_NAME # Check all auto-detected disks"
echo " $SCRIPT_NAME /dev/sda # Check specific disk"
echo " $SCRIPT_NAME /dev/nvme0n1 # Check specific NVMe disk"
echo " $SCRIPT_NAME /dev/sda /dev/nvme0n1 # Check multiple disks"
echo ""
echo "Supported: SATA HDD/SSD, SAS HDD/SSD, NVMe, Hardware RAID, Software RAID"
}
# Parse command line arguments
case "${1:-}" in
-h|--help)
usage
exit 0
;;
-v|--version)
echo "$SCRIPT_NAME version $VERSION"
exit 0
;;
*)
main "$@"
;;
esac