I doubt if there’ll ever be a film version of ‘When Databases Go Bad’, but the people at Xeround have done a good job of making MySQL failures seem a bit more dramatic than usual.
The infographic from Xeround shows ‘the Biggest MySQL Disasters of All Time’, with earthquake magnitudes attached to the various failures.
Top at Magnitude’s 10 and 9 are two Amazon Web Services outages. Whether these are really the biggest disasters of all
time is debatable, but it makes interesting reading in a scary sort of way. Maybe there is scope for that movie, presumably with Jeff Goldblum playing the cool-headed database administrator who saves the day by rebuilding the indexes and getting things back on line.
Click for a full resolution PDF
The ‘lessons learned’ analysis of the database outages could well raise a hollow laugh if you’re a database developer; most of the companies say they’re going to put a lot more emphasis on doing proper backups. They are probably going to get more exercise, lose weight and be kinder to their mothers, too.
Xeround offers a cloud database service that guarantees high availability, so has a certain interest in making you feel nervous when the lights on the database server dim and flicker.
However, the outages described are real enough, and some more in depth analysis of problems is drawn from a Percona White Paper on research into causes of downtime in MySQL.