Point was accessed from multiple goroutines and there was a race on the the internal
cachedFields and cachedName fields. Accessing these fields is unnecessary work as it
requires the point to be unmarshal into Go types and then remarshaled back into protbuf
types. Instead, just send the line protocol version already available on the point via
the protobuf. This avoid accesssing these cached fields and eliminates some extra work.
Possible fix for #4069
With this change remote mapping no longer uses HTTP, as the HTTP ports
exposed by nodes on the cluster are not known cluster wide. The TCP
ports exposed by the cluster service are, so this change uses that
functionality. Each RemoteMapper has its own dedicated connection pool
for each node, and remote mapping TCP connections are in no way coupled
with query TCP connections.
With this change, the query engine code gathers information about
shards and tagsets by working with individual shards, collating the
information, and returning that to the client. It does not assume that any
particular shard is local, and accesses all shards through abstracted
Mappers, of which there are two types -- a Mapper type for Raw queries
and a second type for Aggregate queries. There are corresponding
Executors for each type of Mapper, but both types of Executors share the
same interface.
Jason Wilder
David Norton
Philip O'Toole
Ben Johnson
Paul Dix