The following revision refers to this bug:
  https://chromium.googlesource.com/chromiumos/third_party/autotest/+/e587fd67bd5daccfc55474ddbeb3a25fa45ff066

commit e587fd67bd5daccfc55474ddbeb3a25fa45ff066
Author: Paul Hobbs <phobbs@google.com>
Date: Thu Sep 14 02:28:26 2017

[autotest] Fix use of RetryingConnection

The .Execute method returns the number of rows found, not the results of
Fetchall().

BUG= chromium:764822 
TEST=Tested locally

Change-Id: Ie0559245c2908c0c578bf9c67cde8f56486920e7
Reviewed-on: https://chromium-review.googlesource.com/665324
Commit-Ready: Paul Hobbs <phobbs@google.com>
Tested-by: Paul Hobbs <phobbs@google.com>
Reviewed-by: Shuqian Zhao <shuqianz@chromium.org>

[modify] https://crrev.com/e587fd67bd5daccfc55474ddbeb3a25fa45ff066/site_utils/stats/mysql_stats_unittest.py
[modify] https://crrev.com/e587fd67bd5daccfc55474ddbeb3a25fa45ff066/site_utils/stats/mysql_stats.py

fixed temporarily by restarting it.

The permanent fix would be to make sure mysql_stats resets its metrics when the mysql service restarts.

My solution to this was to change the mysql_stats upstart job
to restart with the mysql job.

I have that change now.  Is there an alternative fix that has
some superior characteristic?

No, that's perfect, thanks.

My plan to make the mysql_stats script catch the Exception that is thrown when the input to a Counter decreases, and manually .reset the metric, but this works too.

https://chrome-internal-review.googlesource.com/c/chromeos/chromeos-admin/+/531539

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/chromeos/chromeos-admin/+/fedaa4da1a68d5b8a32382ba36d615211d962c2f

commit fedaa4da1a68d5b8a32382ba36d615211d962c2f
Author: Richard Barnette <jrbarnette@google.com>
Date: Tue Dec 19 03:01:39 2017

For reasons unrelated to this change, I just restarted mysqld
on the master server.  And, Lo! the mysql_stats job restarted.