Hi all, In an effort to improve the experience of debugging resource allocation issues in a Mesos cluster, myself and several others would like to propose that we add a number of new per-framework metrics to the master.
These metrics are intended to increase the operator's ability to inspect the allocator state as well as gather information about incoming and outgoing messages on a per-framework basis. You can find our proposal for these new metrics here <https://docs.google.com/document/d/14aDm85SKMCX6RMJs0o1hRKhU2rABr4mnHIMNKfNdzuk/edit?usp=sharing>. Please take a look and comment! Thanks, Greg
