When deploying this against a non-demo opentelemetry collector (i.e. one that's not hosted on a Databricks cluster) the bearertokenauth blocks cause issues. If you don't specify a Databricks API Token env var the collector doesn't start. We could just have users set a fake API Token but that's not good UX. I propose that we don't include bearertoken auth in the primary init script and create a new init script specifically for working against the sample infrastructure.
When deploying this against a non-demo opentelemetry collector (i.e. one that's not hosted on a Databricks cluster) the bearertokenauth blocks cause issues. If you don't specify a Databricks API Token env var the collector doesn't start. We could just have users set a fake API Token but that's not good UX. I propose that we don't include bearertoken auth in the primary init script and create a new init script specifically for working against the sample infrastructure.