TL;DR: If mysqldfails to start after running brew upgrade, try re-running the upgrade process manually.

---

A few weeks ago I went to install something via Homebrew, and it bugged me to run brew upgrade first, so I did, without really thinking about the consequences.

That upgraded all of the installed packaged to their latest versions, including jumping to new major versions.

I didn’t notice any big problems at the time, but after rebooting my computer today, I noticed that the MySQL service wasn’t running anymore. I checked the error log, and a lot of errors like these:

<br>
2018-07-19T15:45:04.652847Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBadLeechers is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.652872Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBadLeechers<br>
2018-07-19T15:45:04.653695Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBlockedCommentLog is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.653716Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBlockedCommentLog<br>
2018-07-19T15:45:04.654644Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBlockedIPLog is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.654664Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBlockedIPLog<br>
2018-07-19T15:45:04.655657Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBlocks is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.655677Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBlocks<br>
2018-07-19T15:45:04.656875Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBlocks7 is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.656895Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBlocks7<br>
2018-07-19T15:45:04.658150Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfBlocksAdv is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.658170Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfBlocksAdv<br>
2018-07-19T15:45:04.658952Z 2 [ERROR] [MY-012083] [InnoDB] InnoDB: Table regolith@002elocalhost/wp_wfConfig is not found in InnoDB dictionary<br>
2018-07-19T15:45:04.658972Z 2 [ERROR] [MY-010767] [Server] Error in fixing SE data for regolith.localhost.wp_wfConfig</p>
<p>[...]</p>
<p>2018-07-19T15:45:06.518136Z 0 [ERROR] [MY-010022] [Server] Failed to Populate DD tables.<br>
2018-07-19T15:45:06.518169Z 0 [ERROR] [MY-010119] [Server] Aborting<br>

In my case, I got those errors for all the WordFence tables for several sites, but not all of the sites that have those tables. I tried searching for information on those errors, but didn’t find anything that was helpful.

I then tried running mysqlcheck, and got an even more ominous error:

<br>
&gt; mysqlcheck -uroot -p --repair --all-databases<br>
Enter password:<br>
mysqlcheck: Error: Couldn't execute 'SHOW DATABASES': The user specified as a definer ('mysql.infoschema'@'localhost') does not exist<br>

Luckily, searching for that error did turn up a solution from Stack Overflow:

<br>
&gt; mysql -u root -p<br>
mysql&gt; SET GLOBAL innodb_fast_shutdown = 1;<br>
mysql&gt; quit;<br>
&gt; mysql_upgrade -u root -p<br>

After running that and restarting the service, everything was working normally again.

Sequel Pro crashing

On a related note, Sequel Pro v1.1.2 will crash when opening a MySQL 8 database. As of August 2018, that’s only fixed in the nightly builds.