Re: Corruptions like crazy

[Andrew Grimo]

Glad to hear converting the tables to innodb worked well for you. The output of your mysqld.log file on the startup looks much better now.

Answer for Q1: Likely, you're phpMyAdmin backup is taken using a facility like mysqldump. Since phpMyAdmin can't necessarily handle large files, as your database grows, it may not be able to process the restores.

Backup script
# mysqldump -uroot -pMyRootPassword --databases musiciansjamnetwork | gzip -c > backup-$(date +"%Y-%m-%d_%H-%M").gz

Restore script (assuming the script above produced this date_time)
# gunzip backup-2015-04-29_03-03.gz
# mysql < backup-2015-04-29_03-03
or
# gunzip backup-2015-04-29_03-03.gz | mysql

You should also test your backup so that you know how it works, and that it works.

Answer for Q2: You don't really need to run mysqlcheck....so often. You could run it once a week maybe and add the --optimize flag which will defrag your table files and reorder your primary indexes, but is only beneficial if your tables are going through lots of changes such as inserts and deletes. It's your call.

Answer for Q3: The service named httpd is your apache web server...not the database. However, when it is stopped, it will no longer send traffic to your database. But that doesn't stop the database. The service named mysqld (or mysql depending on your rpm install) is the database itself. With Innodb tables, you shouldn't need to stop your database any more for maintenance activities. If you upgrade the database using the yum package manager, you should stop your database for that activity though.

Enjoy!
Andrew