MySQL Server Start Failed: Troubleshooting Startup Errors

# MySQL Server Start Failed: Troubleshooting Startup Errors

You attempt to start MySQL and it fails immediately:

systemctl start mysqld
# Job for mysqld.service failed because the control process exited with error code.
# See "systemctl status mysqld" and "journalctl -xe" for details.

Or when running manually:

mysqld --user=mysql
# mysqld: Can't change dir to '/var/lib/mysql/' (Errcode: 13 - Permission denied)

The server won't start, and you need to diagnose the root cause quickly.

Step 1: Check the Error Log

The MySQL error log is your primary diagnostic tool:

```bash # Find the error log location mysqld --help --verbose 2>/dev/null | grep -A 1 "Default options"

# Common locations cat /var/log/mysql/error.log cat /var/lib/mysql/$(hostname).err cat /usr/local/mysql/data/$(hostname).err ```

Look for specific error patterns:

[ERROR] Can't start server: Bind on TCP/IP port: Address already in use
[ERROR] Do you already have another mysqld server running on port: 3306?
[ERROR] Aborting

Step 2: Check for Port Conflicts

Another MySQL instance or process might be using port 3306:

```bash # Check what's using port 3306 netstat -tlnp | grep 3306 ss -tlnp | grep 3306 lsof -i :3306

# Kill any existing MySQL processes pkill -9 mysqld pkill -9 mysql ```

If another service is using the port, either stop that service or change MySQL's port:

# /etc/my.cnf or /etc/mysql/my.cnf
[mysqld]
port = 3307

Step 3: Check Data Directory Permissions

Permission issues are a common cause of startup failures:

```bash # Check MySQL data directory ownership ls -la /var/lib/mysql/

# The directory should be owned by mysql user chown -R mysql:mysql /var/lib/mysql/ chmod 750 /var/lib/mysql/ ```

Check the MySQL user's home directory and tmp directory:

```bash # Check mysql user exists id mysql

# Create user if missing useradd -r -s /sbin/nologin mysql

# Check tmp directory permissions ls -la /tmp/ chmod 1777 /tmp/ ```

Step 4: Check Configuration File Syntax

Syntax errors in my.cnf can prevent startup:

```bash # Validate configuration syntax mysqld --validate-config

# Or use my_print_defaults my_print_defaults mysqld ```

Common configuration errors:

```ini # Wrong: Missing quotes for paths with spaces datadir = /var/lib/mysql data

# Correct: datadir = /var/lib/mysql-data

# Wrong: Invalid option name inno_db_buffer_pool_size = 1G

# Correct: innodb_buffer_pool_size = 1G ```

Check for conflicting options:

```bash # Check all config files that are loaded mysqld --print-defaults

# Common config locations cat /etc/my.cnf cat /etc/mysql/my.cnf cat /etc/mysql/conf.d/*.cnf cat ~/.my.cnf ```

Step 5: Check InnoDB Corruption

InnoDB corruption can prevent startup:

[ERROR] InnoDB: Attempted to open a previously opened tablespace. Previous tablespace database/table uses space ID: 54 at filepath: ./database/table.ibd
[ERROR] InnoDB: Cannot open tablespace database/table which uses space ID 54

Start MySQL in recovery mode:

# Add to my.cnf under [mysqld]
[mysqld]
innodb_force_recovery = 1

Recovery levels: - Level 1-2: Minimal, allows SELECT and some operations - Level 3: More aggressive, may lose data - Level 4-6: Severe recovery, may have significant data loss

```bash # Start with recovery mode systemctl start mysqld

# If successful, dump data immediately mysqldump --all-databases --routines --triggers --events > backup.sql

# Stop and remove corrupted data systemctl stop mysqld rm -rf /var/lib/mysql/ibdata1 rm -rf /var/lib/mysql/ib_logfile*

# Reinitialize (WARNING: destroys all data) mysqld --initialize --user=mysql

# Restore from backup mysql < backup.sql ```

Step 6: Check Disk Space

Full disks prevent MySQL from writing logs and temp files:

```bash # Check disk usage df -h

# Check inodes (for small files issue) df -i

# Check MySQL data directory size du -sh /var/lib/mysql/

# Clean up old logs if needed rm -f /var/lib/mysql/mysql-bin.*.index # Then purge old binary logs from MySQL ```

Step 7: Check SELinux or AppArmor

Security modules can block MySQL access:

```bash # Check SELinux status getenforce

# Temporarily disable for testing setenforce 0

# Check SELinux context ls -Z /var/lib/mysql/

# Reset MySQL context restorecon -Rv /var/lib/mysql/

# For AppArmor (Ubuntu/Debian) aa-status apparmor_parser -R /etc/apparmor.d/usr.sbin.mysqld ```

Step 8: Check Missing Required Directories

MySQL needs certain directories to exist:

```bash # Check and create required directories mkdir -p /var/run/mysqld chown mysql:mysql /var/run/mysqld chmod 755 /var/run/mysqld

mkdir -p /var/log/mysql chown mysql:mysql /var/log/mysql chmod 750 /var/log/mysql

# Check tmpdir exists mkdir -p /tmp chmod 1777 /tmp ```

Step 9: Check Binary Log Issues

Corrupted binary logs can prevent startup:

```bash # Check for binary log errors in error log grep "binary log" /var/log/mysql/error.log

# Temporarily disable binary logging for recovery # Add to my.cnf [mysqld] skip-log-bin

# Or rename corrupted log files cd /var/lib/mysql mv mysql-bin.000001 mysql-bin.000001.corrupted ```

Step 10: Check for Table Corruption

MyISAM table corruption:

```bash # Check and repair MyISAM tables myisamchk -r /var/lib/mysql/database/*.MYI

# For severe corruption myisamchk -o /var/lib/mysql/database/*.MYI ```

For InnoDB, use recovery mode as described earlier.

Verification

After fixing the issue:

```bash # Start MySQL systemctl start mysqld

# Check status systemctl status mysqld

# Verify MySQL is listening netstat -tlnp | grep 3306

# Connect and verify mysql -u root -p -e "SELECT 1;"

# Check error log for clean startup tail -f /var/log/mysql/error.log ```

Look for these good signs in the error log:

[Note] mysqld: ready for connections.
Version: '8.0.33'  socket: '/var/lib/mysql/mysql.sock'  port: 3306

Common Startup Error Quick Reference

Prevention Best Practices

1. 1.Always backup before making configuration changes
2. 2.Test config changes: mysqld --validate-config
3. 3.Monitor disk space and set alerts
4. 4.Keep error log accessible and check regularly
5. 5.Use LVM or ZFS snapshots for quick recovery

```bash # Test startup without actually starting mysqld --user=mysql --verbose --help | head -20

# Always check logs after config changes systemctl restart mysqld && journalctl -u mysqld -n 50 ```