Multi-file comparison ability

It would be a very useful feature for the datasource to be able to load more than one source JFR file at a time, and plot the timeseries data from each on the same plots, but with some way to distinguish which file the datapoint came from. This could be simply a prefix on the timeseries key/name, ex. `JVM1.cpuLoad`. There could be some dashboard panel with a simple visualization like a table which maps `JVM1` to the filename that provided the data. These filenames would be the ones pushed into the datasource by Cryostat, so the Cryostat server would be responsible for naming the files in a way that encodes enough identifiable information about the recording file and the target JVM it came from.