Report this

What is the reason for this report?

MySQL Upgrade from 5.5 to 5.6 fails and prevents a clean install.

Posted on August 26, 2015

Reference: https://www.digitalocean.com/community/questions/upgrade-mysql-5-5-to-5-6-ubuntu-14-04

Followed both sets of instructions from the link above to upgrade MySQL from version 5.5 to 5.6. I put a tail on the MySQL error.log file, and the error that appears each time is:

ERROR: 1062 Duplicate entry ‘innodb’ for key ‘PRIMARY’

Since my DBs were brand new, I was willing to lose data, so i performed several versions of a clean uninstall using remove, purge, and autoremove and even used regex on the uninstall attempts (ex. sudo apt-get remove mysql-server-*). After the uninstall attempts, i ran dpkg --get-selections to ensure that the packages were no longer installed. Unfortunately, every subsequent install of 5.5 or 5.6 produces the exact same error.

My guess is that there’s a file or config that I am not removing, but purge is supposed remove all configs. My other guess is that the packages under apt-get are bad and I need to download (wget) and use a local repository. I may be way off, but this is where I have landed.

One last thing to add…searching online for this error during upgrades, it appears that the package attempts to resolve DB names in a case insensitive way. There is a command to convert all DBs to LOWER({dbname}). So, if InnoDB exists in PRIMARY, the package attempts to add innodb to PRIMARY which would produce the duplicate key. Unfortunately, without being able to install or start the service, I can’t figure out where the duplicate entry is going.

I have plenty of tech experience, but a bit of a Linux noob, so don’t be afraid to go under the hood. I can handle it. Hopefully, the aforementioned shows that I have done my due diligence to research and find a resolution(s) to the issue(s) on my own before coming to the community. Any help would be greatly appreciated. Thanks.

  • possum

The developer cloud

Scale up as you grow — whether you're running one virtual machine or ten thousand.

Start building today

From GPU-powered inference and Kubernetes to managed databases and storage, get everything you need to build, scale, and deploy intelligent applications.