...with this change, it'll be possible to identify WriterFrontend's
based on (stream, filter_name, path) pairs in addition to (stream,
writer, path) pairs.

The header captures the enum values as well as the fields

If cluster::backend isn't broker_mgr, use the WriterFrontend's buffering
logic and send a whole batch of log writes during FlushWriteBuffer().
This is a different path than broker's own logging logic.

Preferably we adapt broker to a model where it isn't
buffering either.

This is a serializer for log records that is using SerialTypes
for serializing and un-serializing rather. Essentially, this is
similar to what broker does except for the envelope.