Investigate trace_exporter error logs

[zackattack01]

While validating trace shipping in the 1.3.2 release we found the following error log:

{"caller":"launcher.go:208","component":"trace_exporter","err":"failed to build resolver: passthrough: received empty target in Build()","level":"debug","msg":"could not create new exporter","session_pid":16021,"ts":"2023-12-19T21:57:31.347327917Z"}

This did not appear to cause issues (traces were eventually shipped) as long as it was followed by:

{"caller":":0","component":"trace_exporter","level":"debug","msg":"successfully replaced global provider after adding osquery attributes","session_pid":16021,"ts":"2023-12-19T21:58:02.026116048Z"}

Verifying the timeline for exactly when shipping starts working was difficult due to the sampling rate, and frequent restarts/re-enrollments throughout the release validation - but @RebeccaMahany also suspected there may have been one window where traces were not shipping until launcher was reloaded.

We will likely want to investigate this error at some point, especially if we continue to see large gaps in shipping windows.

[James-Pickett]

I believe that anytime we have a fresh install with a branch new launcher.db, we'll see the failed to build resolver message at least once. This is happening because we don't have a trace_ingest_url until we get the first batch of control server data. Should probably make apparent some how that we expect this.