The Dockyard is used as a library (with no GUI itself) that collects data from the Fuchsia device through the transport system. It responds to requests from the GUI for digestible (i.e. small) sections of data that will be presented to the user.
This is an alphabetical glossary of terms.
A number used to uniquely refer to a Sample Stream or other named entity. The Dockyard ID is 1:1 with a specific Dockyard Path.
A UTF-8, case sensitive text string referring to a Sample Stream or other named entity. The path value is primarily used in the UI, internally the Dockyard ID is used. For example, cpu:0
, physMem
, procCount
are Dockyard Paths.
A graph rendering filter. Scales the graph output so that the high and low values are framed nicely on the screen. The data is scaled dynamically.
A sample is a piece of data from the device. A Sample has a time (when it was gathered), a path (like a name), and a value. For example, Sample z: (<time T>, "memory:used", 300000)
.
The named idea of where Samples come from. For example, cpu
, memory
, process
are Sample Categories.
A portion of a Sample Stream. For example, a Sample Stream might have 100,000 Samples and a Sample Set might only consist of 1,200 of those Samples.
A request for samples sent to the dockyard. It specifies the time range, sample streams, and filtering desired.
A message sent from the Dockyard in response to a Sample Stream Request. It includes the filtered sample data (a Sample Set) for each of the requested Sample Streams. A filtered sample is a mathematically generated representation of 0 to many samples.
An ordered set of Samples for a given Sample Category. For example, all the Samples for cpu0
.
A graph rendering filter that blends neighboring samples. The resulting graph will have have less dramatic highs and lows (with less noise).
Without smoothing, a set of samples like [100, 0, 100, 0, 100]
will look like a letter ‘W’. With smoothing, that data will look like a wavy line with values closer to 50. This makes overall trends easier to see, though it sacrifices detail. Smooth is like an average, but it includes data from wider range of Samples.