One possible reason why your MySQL database is broken is because the server disk was full. This can cause data corruption and prevent the database from functioning properly. To repair a broken MySQL database, you need to first identify the type of error and the database engine used. There are different methods to troubleshoot and repair a MySQL database depending on these factors.

One method is to use the innodb_force_recovery option in the MySQL configuration file. This option allows you to start the MySQL server and perform some operations on the InnoDB tables even if they are corrupted. The innodb_force_recovery option can have values from 1 to 6, where higher values mean more drastic measures to recover the data. However, this option does not guarantee that the data will be fully recovered or that the corruption will be fixed. It is only a temporary solution to allow you to dump your data and restore it on a new instance.

To use the innodb_force_recovery option, you need to follow these steps:

1. Stop the MySQL server if it is running.
2. Edit the MySQL configuration file (usually located at /etc/my.cnf or /etc/mysql/my.cnf) and add the following line under the [mysqld] section:

innodb_force_recovery = N

where N is a value from 1 to 6 depending on the severity of the corruption. You can start with a low value and increase it gradually if the server does not start or if you cannot dump your data.

1. Save the configuration file and restart the MySQL server.
2. Connect to the MySQL server and try to dump your data using the mysqldump command or any other backup tool. For example, to dump a specific database, you can use:

mysqldump -u root -p --single-transaction --quick --lock-tables=false <database name> > backup.sql

After dumping your data, stop the MySQL server and remove or comment out the innodb_force_recovery line from the configuration file.

Start a new MySQL instance with a fresh data directory and restore your data from the backup file using the mysql command or any other restore tool. For example, to restore a specific database, you can use:

mysql -u root -p <database name> < backup.sql

Check your data for consistency and integrity.

If this method does not work, another method is to recover from .frm and .ibd files. These are files that store the table definition and data for InnoDB tables respectively. If you have these files, you can try to recreate the table structure and import the data into a new database. However, this method requires that you have a matching .frm and .ibd file for each table and that they are not corrupted.

To recover from .frm and .ibd files, you need to follow these steps:

1. Create a new database with the same name as the original one on a new MySQL instance.
2. Copy all the .frm files from the original data directory to the new data directory under the same database name.
3. Install mysqlfrm tool from website.
4. This tool will help you to recover the database structure.
5. Start the MySQL server and connect to it.
6. For each table, run the following query:

ALTER TABLE <table name> DISCARD TABLESPACE;

This will detach the table from its original .ibd file.

1. Stop the MySQL server and copy all the .ibd files from the original data directory to the new data directory under the same database name.
2. Start the MySQL server and connect to it.
3. For each table, run the following query:

ALTER TABLE <table name> IMPORT TABLESPACE;

This will attach the table to its new .ibd file.Check your data for consistency and integrity.

Here is a script that will do all the steps described (recovery_loop.sh):

#!/bin/bash
MUSER="root"
MPASS="pass"
MHOST="localhost"
MYSQL="/opt/lampp/bin/mysql"
DB="sample_db"

echo > /root/recovery_dump.sql
echo > /root/recovery_post.sql

TABLENAMES="$($MYSQL -u$MUSER -p$MPASS $DB -Bse 'show tables')"

for table in ${TABLENAMES[@]}
do
  echo $table;
  echo DROP TABLE IF EXISTS $table";" >> /root/recovery_dump.sql;
  mysqlfrm --server=root:pass@localhost:3306 /root/$DB/$table.frm --port=3310 --user=mysql --basedir /opt/lampp/ >> /root/recovery_dump.sql
  echo ";" >> /root/recovery_dump.sql;
  echo alter table $table discard tablespace";" >> /root/recovery_dump.sql;
  echo alter table $table import tablespace";" >> /root/recovery_post.sql;
done

and you have to run the following commands after the script is ready:

cp -R /opt/lampp/var/mysql/sample_db /root/

cd /opt/lampp/var/mysql/

/root/recovery_loop.sh

rm -rf /opt/lampp/var/mysql/sample_db/*

mysql -u root -p sample_db < /root/recovery_dump.sql

cp /root/sample_db/*.ibd /opt/lampp/var/mysql/sample_db/

chown -R mysql:mysql /opt/lampp/var/mysql/sample_db/

mysql -u root -p sample_db < /root/recovery_post.sql

mysqldump -u root -p sample_db > /root/sample_db_dump.sql

The last step is to check your data for consistency and integrity and restore sample_db_dump.sql to a fresh database.

These are some of the ways to repair a broken MySQL database. However, they are not foolproof and may not work in all cases. The best way to prevent data loss and corruption is to have regular backups of your database and monitor your disk space usage.