This is the result of a general audit of how we fork threads, with a
detour into how we're using mutable state especially in websocket
codepaths, making more robust to async exceptions and exceptions
resulting from bugs.
Some highlights:
- use a wrapper around 'immortal' so threads that die due to bugs are
restarted, and log the error
- use 'withAsync' some places
- use bracket a few places where we might break invariants
- log some codepaths that represent bugs
- export UnstructuredLog for ad hoc logging (the alternative is we
continue not logging useful stuff)
I had to timebox this. There are a few TODOs I didn't want to address.
And we'll wait until this is merged to attempt #3705 for
Control.Concurrent.Extended
We upload a set of accumulating timers and counters to track service
time for different types of operations, across several dimensions (e.g.
did we hit the plan cache, was a remote involved, etc.)
Also...
Standardize on DiffTime as a standard duration type, and try to use it
consistently.
See discussion here:
https://github.com/hasura/graphql-engine/pull/3584#pullrequestreview-340679369
It should be possible to overwrite that module so the new threadDelay
sticks per the pattern in #3705 blocked on #3558
Rename the Control.Concurrent.Extended.threadDelay to `sleep` since a
naive use with a literal argument would be very bad!
We catch a bug in 'computeTimeDiff'.
Add convenient 'Read' instances to the time unit utility types. Make
'Second' a newtype to support this.
