Instance index for all types now in logs

[Mike Youngstrom]

About a year ago there was a discussion about use of spec.index in release jobs [0] and the problems with that for those who's installation consists of multiple deployments or for those who's deployments for a given job templates type consist of multiple jobs.

Though the specific issue in that thread was resolved I recently noticed that loggregator has begun identifying the index of all components in its log messages [1].  This means that if we have 10 DEAs spread across multiple deployments or jobs then the index is unable to uniquely identify the component emitting the message if the component is spread across multiple AZs.  I'm fairly certain PWS having job templates spread across multiple AZs would have the same issue we are seeing.

Is there a tracker issue for this problem?  Are their any proposed solutions?  Whatever the solution I just want to make sure it works for our situation of an installation spread across multiple bosh deployments and not just the current PWS case of a template spread across multiple jobs.

Mike

[0] https://groups.google.com/a/cloudfoundry.org/d/msg/vcap-dev/XY5t7rlZD0o/R6bvPPaXqHQJ

[1] https://www.pivotaltracker.com/story/show/85484012