Improve Netdata Dashboards

[sanfordd]

Improve netdata dashboards for quick status checks,

• Implement more yes/no status checks

[sanfordd]

Finished interviews with relevant users and we'll be using the netdata badges to function as simpler displays.
The goal is to have a simple yes/no box to check the status of the application itself and then a few boxes with similarly simple values to help people determine where to start looking at a problem.

From netdata we'll be able to see
Resque-worker count
Redis seconds since last connection
Connections to the sufia database (1+ indicates it's there and working)
Connections to the fcrepo database (1+ shows the database is there and has a connection though normal use is many more connections)

We can enable the tomcat metrics as well, which should let us monitor that aspect.

Fedora and Solr don't have any current direct monitoring tools. We can look into writing them if we want , or we can use services they rely on such as monitoring Tomcat and Postgres for Fedora to serve as a basic option.

Application availability will be handled by taking the uptime data from Honeybadger directly, it doesn't seem worthwhile to integrate it with netdata.