what no blog post mentions and what happend before all this incident, was that they were running out of disk space.
No matter what happened after that, that was just a traversal of failures and laziness by the uber people.
sure postgresql has some weak points, but being lazy and then try to make a blog post about the weak points while ignoring the fact that you were extremly lazy in the written post is just curious.
for me their switch is more like a "we did bad so we use mysql since we can do everything from scratch" rather than "we will look what we did wrong and try to improve that". they just don't care what they did wrong, they just blame the implementation details about postgresql and move on.
somewhat stupid.
Probablly they should do it like netflix in the future. Use MySQL and use AWS RDS or any other managed MySQL solution so they can't screw up that hard in the future.
sure postgresql has some weak points, but being lazy and then try to make a blog post about the weak points while ignoring the fact that you were extremly lazy in the written post is just curious.
for me their switch is more like a "we did bad so we use mysql since we can do everything from scratch" rather than "we will look what we did wrong and try to improve that". they just don't care what they did wrong, they just blame the implementation details about postgresql and move on.
somewhat stupid.
Probablly they should do it like netflix in the future. Use MySQL and use AWS RDS or any other managed MySQL solution so they can't screw up that hard in the future.