I wonder what game this was for.
To me this is validation that going the SQL route in almost 99% of new apps is the right way to go. It will be a rare case that you won't be able to scale out given how mature some of these technologies are.
To me this is validation that going the SQL route in almost 99% of new apps is the right way to go. It will be a rare case that you won't be able to scale out given how mature some of these technologies are.
There are very few things that structured NoSql can't do if you ignore reporting. Once you scale out traditional SQL, you aren't going to be using it for reporting either way though.
Automating master failover on MySQL without a human in the loop - in the topology GitHub used back then - is risky.
0: https://github.blog/2018-10-30-oct21-post-incident-analysis/