Flow based data-collector and time-centric data-processor.
Faxe's inner core is based on a dataflow computing engine and it's components also called
nodes can be
freely combined into an acyclic graph.
Unlike other flowbased frameworks (node_red, ...) in Faxe computing graphs are built with a DSL called
FAXE can be managed via its rest api.
Data in faxe
In faxe we deal with
data-items are emitted by nodes
data_point consists of a
The value of the ts field is always: unix-timestamp in millisecond precision without a timezone.
tags are essentially
Valid data-types for field and tag values are:
string, integer, float, key-value map (also deeply nested) and lists.
The only valid data-type for field and tag keys is
A data_batch consists of a list of data_points ordered by timestamp.
Most faxe nodes can deal with both points and batches.
As field and tag values can be deeply nested maps and lists, it is possible to use a
JSON-path like syntax
to declare and reference these values:
averages axis.z.cur value.sub.data averages.emitted