Vehicle Tracks - Querying/Filtering in Dashboards Takes a LONG Time. Ideas?

Question

When I try to filter vehicle tracks in a dashboard, it takes 2+ minutes to display on the map and the browser tab memory sky-rockets.  If you happen to change the zoom level, buckle-up for a few more minutes, but most the time the features do not successfully refresh.  This specific vehicle tracks service has over 1.5 million records.  Is there a more efficient way to manage this, or is there a way in swagger to add an attribute index on the date field to help with this?  I have tried applying a filter in the webmap and that doesn't seem have an effect on query performance in the dashboard.

PeterNasuti · Answer

@ArmstKP If the layer in question is an ArcGIS Online hosted feature layer (this is what is created when unchecking "Store data in spatiotemporal feature layer"), the query performance in dashboards would be better suited to be reviewed with the ArcGIS Online team and an Esri Support Services case. In that configuration, ArcGIS Velocity is simply writing the data out to an ArcGIS Online hosted feature layer.

You are correct that currently ArcGIS Velocity spatiotemporal layers cannot add fields. This is an enhancement we are considering for the future of the product to enable this capability.

ArmstKP · Answer

@JakeSkinner this particular one is a hosted feature layer.  I specifically did not choose spatiotemporal because I had needed to edit/add fields, which you couldn't do with a spatiotemporal.  If I were to have chosen spatiotemporal, and my AVL provider now pushes additional fields, I would not be able to take advantage of those fields.  Or, if when I had chosen spatiotemporal and I had noticed I hadn't utilized a field I actually want to utilize now, I couldn't add that field.  Seems like a conundrum.

JakeSkinner · Answer

Hi @ArmstKP,

Where is the data be stored?  Is it in an Enterprise Geodatabase, Relational Data Store, or Spatiotemporal Data Store.  With data this large I recommend using the Spatiotemporal Data Store.