Download of backup of database inserts failed html

When a database backup is started for download in the browser the following code is inserted into the SQL file:

mysqldump failed :

mysqldump: Couldn't execute 'show events': Cannot proceed, because event scheduler is disabled (1577)





Howdy -- thanks for your report!

Which MySQL version is it that you're using there?

If it's different from the version that came with your distro, which repository did it come from?

pixel_paul's picture
Submitted by pixel_paul on Tue, 09/24/2019 - 10:37 Pro Licensee

It is MariaDB 10.3 from their repo.

Gotcha, thanks for the info!

We unfortunately don't have full support for that yet, MariaDB 10.3 support is something we're finishing up for Debian 10.

It should be ready soon, but at the moment there are indeed some issues.

We've also seen it cause problems with adding new database users as well.

I'm hoping that'll be ready soon though, we actually had hoped to release that last week, but it shouldn't be much longer.

pixel_paul's picture
Submitted by pixel_paul on Tue, 09/24/2019 - 12:26 Pro Licensee

Sure thing, well count this as a bug report :)

It's easily resolved by simply removing the output from the SQL file, just thought it worth mentioning.



Assigned: Unassigned ยป

Thanks for the info!

I'm passing this along to Jamie for further comment. Jamie, do the fixes you've been making for he new MariaDB handle this issue as well?

Hi all,

I think I ran into this last night while upgrading a number of servers from Debian 8 to 9 (MariaDB 10.1). The Debian upgrade went fine, but when the backups ran a short while later I got the events error along with a partial backup email. I believe I've fixed this by running mysql_upgrade.

I've been a long time Debian user and have done a number of major upgrades without issue. I thought this type of thing was normally taken care of by the Debian MySQL/MariaDB upgrade scripts, using the special debian-sys-maint database user (/etc/mysql/debian.cnf), but that didn't appear to run for any of my servers (even the non-Virtualmin servers). Either I broke something or Debian has made some changes to the upgrade process and expects the user to update the database schema post-upgrade. still references the debian-sys-maint user so I'm not really sure. I thought I had read something more recently that says this is no longer used.

I also have/had a number of CentOS/RHEL and Fedora servers, and with those I thought I always had to manually do a mysql_upgrade after the fact. CentOS no longer even offers a CD upgrade path so it's been a long time since I did an upgrade. Just rebuild everything now, dump databases and bring it over to the new server, etc.

If Virtualmin already keeps tabs on the database version I suppose it could run the mysql_upgrade command when it sees a change, or it could be something triggered by the [Update Detected Operating System] button / System Settings -> Re-Check Configuration as well. Or just tell the user to do it, note it on the index like the detected OS change, give them a button to click. Might be better than doing it automatically.