I’ve got a problem with my server, which is probably connected to the backup script. The problem looks like this: about one minute after the bash script with the mysqldump is executed my server stops delivering the dynamic pages (because of no answer from the MariaDB, static files are being delivered without any problem at all), no logs are being created and I can’t even log in to the ssh. Everything gets back to normal after the reset. The database itself isn’t very big, it has almost 50GB and on my hard drive, there is 100GB of free space. After the last “crash” I’ve checked the size of the file that was created by mysqldump and it’s almost 5GB, so if the problem really is with space, then something is eating 95GB and that file (or files) are inside of /tmp/ directory. Am I right?
How can I check why it is happening? Is there something that I could do, to prevent the hard disk from getting to 100% usage while I’m creating the database’s backup?
And maybe one more question, not related to the mysqldump directly, is there something that I can set, so ssh won’t get messed up when the disk is full?
As for the server itself, it’s Debian Stretch with MariaDB v.10.1.26. The dump line contains the settings: all databases, 3x ignore table and default character set.
Edit: One more thing that could be important. There is a mix of engines. Some of the tables are innodb, and some are myisam.