Today we deployed the first in a set of large-scale infrastructure improvements, and the big win for everyone is that web apps are much faster, and much-better insulated from each other. It should also help reliability – we’re expecting that the daily problems where some apps had 502 errors at around mid-morning UTC will no longer occur.
There are a bunch more infrastructural improvements in the pipeline – stay tuned!