Fix 500 error from /consent form

Fixes #3731

Fix exceptions when a connection is closed before we read the headers

This fixes bugs introduced in #3700, by making sure that we behave sanely
when an incoming connection is closed before the headers are read.

Make the logcontext filter not explode if it somehow ends up with a logcontext
of None, since that infinite-loops the whole logging system.

Use get_cache_factor_for function for `state_cache`

This allows the cache factor for `state_cache` to be individually
specified in the enviroment

Fix logging bug in EDU handling over replication

Logcontexts for replication command handlers

Fix logcontexts for running pushers

Use a producer to stream back responses

Refactor request logging code

add new error type ResourceLimit

Merge branch 'neilj/limit_exceeded_error' of github.com:matrix-org/synapse into neilj/limit_exceeded_error

Neilj/resource limit block event creation

Update the admin register documentation to return a real user ID

Presumably this is the intention anyways. I've also updated the domain part to be something more along the lines of what people might expect.

Run the handlers for replication commands as background processes. This should
improve the visibility in our metrics, and reduce the number of "running db
transaction from sentinel context" warnings.

Ideally it means converting the things that fire off deferreds into the night
into things that actually return a Deferred when they are done. I've made a bit
of a stab at this, but it will probably be leaky.

First of all, avoid resetting the logcontext before running the pushers, to fix
the "Starting db txn 'get_all_updated_receipts' from sentinel context" warning.

Instead, give them their own "background process" logcontexts.