Where Are MySQL Logs Stored? Default Paths (Ubuntu, CentOS, Docker)

MySQL keeps several distinct logs, and only the error log is on by default. On Debian and Ubuntu it lives at /var/log/mysql/error.log. On RHEL, CentOS, and Fedora it is /var/log/mysqld.log for Oracle MySQL, or /var/log/mariadb/mariadb.log for MariaDB. The path is controlled by the log_error system variable; check it with SHOW VARIABLES LIKE 'log_error';. The general query log and slow query log are off by default and must be enabled explicitly.

Quick Reference: MySQL Log File Paths

The four MySQL log types: error log (startup, shutdown, errors), general query log (every statement), slow query log (queries over long_query_time), and binary log (replication/recovery, at /var/lib/mysql/binlog.* by default).

How to View and Tail MySQL Logs

On Debian/Ubuntu:

sudo tail -f /var/log/mysql/error.log

On RHEL/CentOS (MySQL):

sudo tail -f /var/log/mysqld.log

Via the systemd journal when log_error is empty:

sudo journalctl -u mysql -f
sudo journalctl -u mariadb -f

Grep for errors:

sudo grep -iE 'error|\[ERROR\]' /var/log/mysql/error.log

In Docker:

docker logs -f <container>

Finding the Exact Path When It Varies

Ask the running server — this is the only reliable method:

SHOW VARIABLES LIKE 'log_error';          -- error log path
SHOW VARIABLES LIKE 'datadir';            -- base for relative paths
SHOW VARIABLES LIKE 'general_log%';       -- general_log (ON/OFF) + general_log_file
SHOW VARIABLES LIKE 'slow_query_log%';    -- slow_query_log + slow_query_log_file
SHOW VARIABLES LIKE 'log_output';         -- FILE, TABLE, or NONE

If log_output is TABLE, the general and slow logs are written to mysql.general_log and mysql.slow_log tables instead of files — query them with SQL. A relative log_error resolves against datadir.

Ubuntu / Debian Detail

Config is split across /etc/mysql/mysql.conf.d/mysqld.cnf (MySQL) or /etc/mysql/mariadb.conf.d/50-server.cnf (MariaDB). The error log defaults to /var/log/mysql/error.log. Edit, then restart:

sudo systemctl restart mysql      # or: mariadb

RHEL / CentOS / Fedora Detail

The main config is /etc/my.cnf with drop-ins under /etc/my.cnf.d/. MySQL's log_error defaults to /var/log/mysqld.log; MariaDB uses /var/log/mariadb/mariadb.log. SELinux labels these with mysqld_log_t, so relocating logs requires semanage fcontext + restorecon.

Docker Detail

The official mysql and mariadb images log to stderr, so docker logs <container> shows the error log and there is no file inside the container by default. To persist files, mount a volume and point the config at it:

docker run -d --name db \
  -e MYSQL_ROOT_PASSWORD=secret \
  -v /host/mysql-log:/var/log/mysql \
  mysql --log-error=/var/log/mysql/error.log

Enabling and Changing Logs

In a config file under [mysqld]:

[mysqld]
log_error = /var/log/mysql/error.log
log_output = FILE
slow_query_log = 1
slow_query_log_file = /var/log/mysql/slow.log
long_query_time = 1
general_log = 0
general_log_file = /var/log/mysql/general.log

general_log, slow_query_log, and their file paths can be toggled at runtime without a restart:

SET GLOBAL slow_query_log = 'ON';
SET GLOBAL long_query_time = 1;
SET GLOBAL general_log = 'OFF';   -- keep off in production; it is very verbose

log_error requires a restart to change.

Log Rotation

Debian/Ubuntu ship /etc/logrotate.d/mysql-server (or mariadb); RHEL ships /etc/logrotate.d/mysqld. A typical config:

/var/log/mysql/*.log {
    daily
    rotate 7
    missingok
    create 640 mysql adm
    compress
    sharedscripts
    postrotate
        test -x /usr/bin/mysqladmin && /usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf flush-logs
    endscript
}

The postrotate runs mysqladmin flush-logs so MySQL reopens its files after the rename. Force a test run:

sudo logrotate -f /etc/logrotate.d/mysql-server

Common Errors You'll Find in MySQL Logs

• [ERROR] Access denied for user 'x'@'host' (using password: YES) — wrong credentials or no grant for that host; fix with GRANT/CREATE USER.
• [ERROR] Can't start server: Bind on TCP/IP port: Address already in use — port 3306 is taken by another mysqld; stop the duplicate instance.
• [ERROR] InnoDB: Cannot allocate memory for the buffer pool — innodb_buffer_pool_size is larger than available RAM; lower it.
• [ERROR] Table './db/table' is marked as crashed — a MyISAM table needs REPAIR TABLE or myisamchk.
• [Warning] Aborted connection ... (Got timeout reading communication packets) — client timed out or max_allowed_packet is too small for the payload.
• Disk is full writing '/var/lib/mysql/...' — the data volume is out of space; free space before writes stall.

Troubleshooting: Logs Missing or Empty

• Empty log_error — with no path set, errors go to stderr/journal; check journalctl -u mysql or docker logs.
• Logging to tables — if log_output = TABLE, query mysql.general_log and mysql.slow_log instead of looking for files.
• General/slow log off — both are disabled by default; enable them with the variables above before expecting output.
• Wrong path — a relative log_error resolves against datadir; always confirm with SHOW VARIABLES LIKE 'log_error';.
• Permissions — files are owned by the mysql user (mode 640); use sudo.
• Won't start — startup crashes log to the journal even when a file is configured, because mysqld fails before opening the file.
• Docker — no error.log inside the image unless you mounted a volume and set --log-error; otherwise use docker logs.