Remove excessive gRPC resolver logging

[mkeeler]

Overview of the Issue

Using the master branch, I spun up a dev agent and noticed a bunch of weird looking logs:

  2021-05-05T10:17:42.971-0400 [INFO]  agent: ccResolverWrapper: sending update to cc: {[{127.0.0.1:8300 0 hashimac.keeler.lan <nil>}] <nil>}
    2021-05-05T10:17:42.971-0400 [INFO]  agent: ccResolverWrapper: sending update to cc: {[{127.0.0.1:8300 0 hashimac.keeler.lan <nil>}] <nil>}
    2021-05-05T10:17:42.971-0400 [INFO]  agent: addrConn: tryUpdateAddrs curAddr: { 0  <nil>}, addrs: [{127.0.0.1:8300 0 hashimac.keeler.lan <nil>}]

These are periodically emitted and at the info level. From what I have been able to gather it looks like internal gRPC logging from here and a couple other locations.

Why the gRPC library thinks these should be info levels I don't know. But for our purposes these are really more like debug/trace level logging.

We do implement our own gRPC logger including filtering of gRPC log levels.

I am thinking we could do something with our logger implementation to downgrade the level set by gRPC at least in the case of the info logs.

Reproduction Steps

Run a dev agent from master.

[mkeeler]

Looks like there were upstream issues about this: grpc/grpc-go#3918

From my reading of that thread the grpc info level logs really are what we would classify as debug/trace logs.

I think we should probably update our grpc logger to send info level logs out as trace logs for hclog.