ORA-15042 ASM Disk Is Missing - Disk Recovery and Replacement Guide

ORA-15042: ASM Disk “disk_name” Is Missing

Error Overview

Error Text: ORA-15042: ASM disk "disk_name" is missing

This error occurs when ASM (Automatic Storage Management) cannot find a disk that is supposed to be part of a diskgroup. The disk may have failed, been removed, or become inaccessible due to various reasons. This error indicates a potential risk to data redundancy and requires immediate attention to prevent data loss.

Understanding ASM Disk Management

ASM Disk States

ASM Disk Lifecycle States
├── CANDIDATE - Available for ASM use
├── MEMBER - Part of a diskgroup
├── FORMER - Was part of a diskgroup
├── PROVISIONED - Prepared but not used
├── CONFLICT - Header conflicts
└── FOREIGN - Contains non-ASM data

Disk Discovery Process

ASM Disk Discovery
├── Read ASM_DISKSTRING paths
├── Open and read disk headers
├── Validate ASM metadata
├── Check disk membership
└── Update v$asm_disk view

Common Causes

Hardware Issues

• Physical disk failure
• Controller failure
• Cable disconnection
• Power supply issues
• Storage array problems

Configuration Problems

# Path changes
# Multipathing issues
# Permission changes
# Disk partition changes
# Device naming changes

Operating System Issues

• Device driver problems
• Kernel module not loaded
• UDEV rules misconfigured
• Filesystem mounted over ASM disk
• SELinux/permissions blocking access

Storage Network Problems

• SAN connectivity issues
• iSCSI target problems
• Fiber channel path failures
• Network timeout issues

Diagnostic Steps

Identify Missing Disk

-- Connect to ASM instance
sqlplus / as sysasm

-- Find missing disk details
SELECT dg.name as diskgroup_name,
       d.name as disk_name,
       d.path,
       d.header_status,
       d.mount_status,
       d.state
FROM v$asm_disk d, v$asm_diskgroup dg
WHERE d.group_number = dg.group_number(+)
AND (d.mount_status = 'MISSING' OR d.header_status = 'FORMER')
ORDER BY dg.name, d.name;

-- Check specific diskgroup for missing disks
SELECT disk_number, name, path, mount_status, header_status, state
FROM v$asm_disk
WHERE group_number = (SELECT group_number FROM v$asm_diskgroup WHERE name = 'DATA')
AND mount_status = 'MISSING';

Verify Disk at OS Level

# List all available disks
ls -la /dev/oracleasm/disks/
lsblk
fdisk -l

# Check specific disk
ls -la /dev/oracleasm/disks/DISK1
file /dev/oracleasm/disks/DISK1

# Check disk accessibility
dd if=/dev/oracleasm/disks/DISK1 of=/dev/null bs=1M count=1

# Check multipath status
multipath -ll
multipath -v3 -ll | grep -A5 -B5 "DISK1"

Analyze System Logs

# Check system messages
dmesg | grep -i "error\|fail\|offline"
tail -100 /var/log/messages | grep -i "disk\|scsi\|error"

# Check ASM logs
tail -500 $ORACLE_BASE/diag/asm/+asm/+ASM/trace/alert_+ASM.log | grep -i "disk\|missing\|offline"

# Check disk events
journalctl -u oracleasm -n 100

Storage Validation

# For Oracle ASM disks
oracleasm status
oracleasm scandisks
oracleasm listdisks

# Query specific disk
oracleasm querydisk DISK1

# Check disk header
kfed read /dev/oracleasm/disks/DISK1 | head -20

Resolution Steps

1. Rediscover Missing Disk

Force Disk Discovery

-- Clear and reset discovery string
ALTER SYSTEM SET asm_diskstring = '' SCOPE=MEMORY;
ALTER SYSTEM SET asm_diskstring = '/dev/oracleasm/disks/*' SCOPE=MEMORY;

-- Check if disk reappears
SELECT path, header_status, mount_status, state
FROM v$asm_disk
WHERE path LIKE '%DISK1%';

-- Force specific path discovery
ALTER SYSTEM SET asm_diskstring = '/dev/oracleasm/disks/DISK1' SCOPE=MEMORY;
ALTER SYSTEM SET asm_diskstring = '/dev/oracleasm/disks/*' SCOPE=MEMORY;

Rescan at OS Level

# Rescan SCSI devices
for host in /sys/class/scsi_host/*; do
    echo "- - -" > $host/scan
done

# Rescan Oracle ASM disks
oracleasm scandisks

# Restart Oracle ASM
oracleasm stop
oracleasm start
oracleasm scandisks

2. Repair Disk Access

Fix Permissions

# Check and fix permissions
ls -la /dev/oracleasm/disks/
chown grid:asmadmin /dev/oracleasm/disks/DISK1
chmod 660 /dev/oracleasm/disks/DISK1

# For raw devices
chown grid:asmadmin /dev/sdb1
chmod 660 /dev/sdb1

# Update UDEV rules if needed
vi /etc/udev/rules.d/99-oracle-asmdevices.rules
udevadm control --reload-rules
udevadm trigger

Fix Multipath Issues

# Check multipath configuration
cat /etc/multipath.conf

# Reload multipath
multipath -r

# Check specific device
multipath -ll /dev/sdb

# Add device to multipath
multipath -a /dev/sdb

3. Bring Disk Online

Online Missing Disk

-- Attempt to online the disk
ALTER DISKGROUP data ONLINE DISK data_0001;

-- Force online
ALTER DISKGROUP data ONLINE DISK data_0001 FORCE;

-- Online by path
ALTER DISKGROUP data ONLINE DISK '/dev/oracleasm/disks/DISK1';

-- Online all missing disks
ALTER DISKGROUP data ONLINE DISKS IN FAILGROUP failure_group_1;

4. Replace Missing Disk

Add Replacement Disk

-- Drop missing disk and add replacement
ALTER DISKGROUP data
    DROP DISK data_0001 FORCE
    ADD DISK '/dev/oracleasm/disks/NEWDISK1' NAME data_0001_new
REBALANCE POWER 10;

-- Monitor rebalance
SELECT * FROM v$asm_operation;

-- Check disk status after rebalance
SELECT name, path, mount_status, header_status, state
FROM v$asm_disk
WHERE group_number = (SELECT group_number FROM v$asm_diskgroup WHERE name = 'DATA');

Advanced Recovery Procedures

Disk Header Recovery

# Backup disk header before repair attempts
dd if=/dev/oracleasm/disks/DISK1 of=/backup/disk1_header.bak bs=4M count=1

# Analyze disk header
kfed read /dev/oracleasm/disks/DISK1 | grep -E "grpname|dskname|dsknum|grptyp"

# Clear disk header if corrupted (WARNING: Destructive)
dd if=/dev/zero of=/dev/oracleasm/disks/DISK1 bs=4M count=1

# Re-add disk to ASM
oracleasm createdisk DISK1 /dev/sdb1

Force Mount with Missing Disks

-- Mount diskgroup despite missing disks
ALTER DISKGROUP data MOUNT FORCE;

-- Check redundancy status
SELECT file_number, redundancy_lowered, mirror_lowered, needs_resilver
FROM v$asm_file
WHERE group_number = (SELECT group_number FROM v$asm_diskgroup WHERE name = 'DATA')
AND (redundancy_lowered = 'Y' OR mirror_lowered = 'Y');

-- Drop missing disks after force mount
ALTER DISKGROUP data DROP DISK data_0001 FORCE;

Extract Data from Remaining Disks

# Use AMDU to extract files
amdu -diskstring '/dev/oracleasm/disks/*' -extract 'DATA.256' -output /recovery/

# Extract all files from diskgroup
amdu -diskstring '/dev/oracleasm/disks/*' -dump 'DATA' -output /recovery/dump/

Monitoring and Prevention

Missing Disk Detection Script

#!/bin/bash
# Monitor for missing ASM disks

ORACLE_HOME=/u01/app/19.0.0/grid
ORACLE_SID=+ASM
export ORACLE_HOME ORACLE_SID

check_missing_disks() {
    MISSING=$($ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF
    SET PAGESIZE 0 FEEDBACK OFF HEADING OFF
    SELECT COUNT(*)
    FROM v\$asm_disk
    WHERE mount_status = 'MISSING' OR header_status = 'FORMER';
EOF
    )

    if [ "$MISSING" -gt 0 ]; then
        echo "ALERT: $MISSING missing ASM disks detected"

        # Get details
        $ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF
        SET LINESIZE 200
        COL diskgroup_name FORMAT A15
        COL disk_name FORMAT A15
        COL path FORMAT A30
        SELECT dg.name as diskgroup_name,
               d.name as disk_name,
               d.path,
               d.mount_status
        FROM v\$asm_disk d, v\$asm_diskgroup dg
        WHERE d.group_number = dg.group_number(+)
        AND (d.mount_status = 'MISSING' OR d.header_status = 'FORMER');
EOF

        # Send alert
        echo "Missing ASM disks detected" | mail -s "ASM Disk Alert" [email protected]
    else
        echo "All ASM disks are present"
    fi
}

# Run check
check_missing_disks

Proactive Disk Health Monitoring

-- Create disk monitoring table
CREATE TABLE asm_disk_monitor (
    check_time TIMESTAMP,
    path VARCHAR2(256),
    diskgroup_name VARCHAR2(30),
    mount_status VARCHAR2(20),
    header_status VARCHAR2(20),
    state VARCHAR2(20),
    read_errors NUMBER,
    write_errors NUMBER
);

-- Monitoring procedure
CREATE OR REPLACE PROCEDURE monitor_asm_disk_health AS
BEGIN
    -- Log disk status
    INSERT INTO asm_disk_monitor
    SELECT
        SYSTIMESTAMP,
        d.path,
        dg.name,
        d.mount_status,
        d.header_status,
        d.state,
        d.read_errs,
        d.write_errs
    FROM v$asm_disk d, v$asm_diskgroup dg
    WHERE d.group_number = dg.group_number(+);

    -- Alert on issues
    FOR rec IN (
        SELECT path, mount_status, read_errs, write_errs
        FROM v$asm_disk
        WHERE mount_status != 'CACHED'
           OR read_errs > 0
           OR write_errs > 0
    ) LOOP
        DBMS_OUTPUT.PUT_LINE('Disk issue: ' || rec.path ||
                           ' Status: ' || rec.mount_status ||
                           ' Read Errors: ' || rec.read_errs ||
                           ' Write Errors: ' || rec.write_errs);
    END LOOP;

    COMMIT;
END;
/

-- Schedule regular monitoring
BEGIN
    DBMS_SCHEDULER.CREATE_JOB(
        job_name => 'ASM_DISK_HEALTH_CHECK',
        job_type => 'PLSQL_BLOCK',
        job_action => 'BEGIN monitor_asm_disk_health; END;',
        start_date => SYSTIMESTAMP,
        repeat_interval => 'FREQ=HOURLY',
        enabled => TRUE
    );
END;
/

Automated Recovery Procedure

#!/bin/bash
# Automated disk recovery script

ORACLE_HOME=/u01/app/19.0.0/grid
ORACLE_SID=+ASM
export ORACLE_HOME ORACLE_SID

recover_missing_disk() {
    local disk_path=$1
    local disk_name=$2
    local diskgroup=$3

    echo "Attempting to recover disk $disk_name at $disk_path"

    # Step 1: Check OS level
    if [ ! -e "$disk_path" ]; then
        echo "Disk path does not exist, scanning for new devices..."
        oracleasm scandisks
        sleep 5
    fi

    # Step 2: Check permissions
    if [ -e "$disk_path" ]; then
        chown grid:asmadmin "$disk_path"
        chmod 660 "$disk_path"
    fi

    # Step 3: Force rediscovery
    $ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF
    ALTER SYSTEM SET asm_diskstring = '$disk_path' SCOPE=MEMORY;
    ALTER SYSTEM SET asm_diskstring = '/dev/oracleasm/disks/*' SCOPE=MEMORY;
EOF

    # Step 4: Try to online disk
    $ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF
    ALTER DISKGROUP $diskgroup ONLINE DISK '$disk_name';
EOF

    # Check result
    RESULT=$($ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF
    SET PAGESIZE 0 FEEDBACK OFF HEADING OFF
    SELECT mount_status FROM v\$asm_disk WHERE name = '$disk_name';
EOF
    )

    if [ "$RESULT" = "CACHED" ]; then
        echo "Disk $disk_name successfully recovered"
        return 0
    else
        echo "Failed to recover disk $disk_name"
        return 1
    fi
}

# Main recovery loop
$ORACLE_HOME/bin/sqlplus -s / as sysasm << EOF > /tmp/missing_disks.txt
SET PAGESIZE 0 FEEDBACK OFF HEADING OFF
SELECT d.path || '|' || d.name || '|' || dg.name
FROM v\$asm_disk d, v\$asm_diskgroup dg
WHERE d.group_number = dg.group_number(+)
AND d.mount_status = 'MISSING';
EOF

while IFS='|' read -r path name diskgroup; do
    recover_missing_disk "$path" "$name" "$diskgroup"
done < /tmp/missing_disks.txt

Emergency Procedures

Critical Disk Recovery

-- Emergency script for critical diskgroups
DECLARE
    v_critical_dg VARCHAR2(30) := 'DATA';
    v_missing_count NUMBER;
BEGIN
    -- Count missing disks
    SELECT COUNT(*) INTO v_missing_count
    FROM v$asm_disk
    WHERE group_number = (SELECT group_number FROM v$asm_diskgroup WHERE name = v_critical_dg)
    AND mount_status = 'MISSING';

    IF v_missing_count > 0 THEN
        DBMS_OUTPUT.PUT_LINE('CRITICAL: ' || v_missing_count || ' disks missing from ' || v_critical_dg);

        -- Attempt force mount
        BEGIN
            EXECUTE IMMEDIATE 'ALTER DISKGROUP ' || v_critical_dg || ' MOUNT FORCE';
            DBMS_OUTPUT.PUT_LINE('Force mount successful');
        EXCEPTION
            WHEN OTHERS THEN
                DBMS_OUTPUT.PUT_LINE('Force mount failed: ' || SQLERRM);
        END;

        -- Drop missing disks
        FOR disk IN (
            SELECT name FROM v$asm_disk
            WHERE group_number = (SELECT group_number FROM v$asm_diskgroup WHERE name = v_critical_dg)
            AND mount_status = 'MISSING'
        ) LOOP
            BEGIN
                EXECUTE IMMEDIATE 'ALTER DISKGROUP ' || v_critical_dg ||
                                ' DROP DISK ' || disk.name || ' FORCE';
                DBMS_OUTPUT.PUT_LINE('Dropped missing disk: ' || disk.name);
            EXCEPTION
                WHEN OTHERS THEN
                    DBMS_OUTPUT.PUT_LINE('Failed to drop ' || disk.name || ': ' || SQLERRM);
            END;
        END LOOP;
    END IF;
END;
/

Best Practices

1. Monitor disk health continuously
2. Configure proper disk discovery strings
3. Maintain consistent device naming
4. Use ASM disk labels for easy identification
5. Implement redundancy appropriate to SLA
6. Regular disk path and permission audits
7. Automate disk recovery procedures

Related Errors

• ORA-15001: Diskgroup does not exist or is not mounted
• ORA-15017: Diskgroup cannot be mounted
• ORA-15040: Diskgroup is incomplete
• ORA-15063: ASM discovered an insufficient number of disks

Troubleshooting Checklist

• Identify missing disk name and path
• Check disk existence at OS level
• Verify disk permissions and ownership
• Scan for new/changed disk devices
• Check multipath configuration
• Force ASM disk rediscovery
• Attempt to online the disk
• Consider disk replacement if unrecoverable