In the case of using non-bb tooling the log parser is pretty opaque process. When it fails there is very little indication why and in the case of success (but incomplete parsing) there is no indication what is missing... Ideally we could have a "log parser log" that is an artifact (stored on s3 or somewhere with a short expiration) that would log the parsing state and dump useful information particularly in the case of an error.
Does the access to newrelic now solve this use case? If not, would bug 1087999 solve this? Or do we need additional logging in the log parser as well?
It really helps me personally but I think we should have some kind of artifact ("parser error logs") if/when the parser fails. This is mostly useful for people who are submitting data via the API. This goes with the team of (self-service API) if you cannot see the error its almost impossible to debug without access to all the right tooling.