Skip to main content

Explain the search execution plan

POST 

/graph/:graph_id/search/explain

Explain the runtime characteristics of a search without performing the search. A section can be defined (defaults to / == root) to interpret relative property paths. Example: is(volume) and (reported.age>23d or desired.clean==true or metadata.version==2)

Request​

Path Parameters

    graph_id stringrequired

    The identifier of the graph

    Example: fix

Query Parameters

    section string

    Possible values: [reported, desired, metadata]

    The name of the section used for all property paths. If not defined root is assumed.

    at date-time

    The timestamp to use for the search. If not defined the latest version of the graph is used.

Body

The search to perform

    string

Responses​

The execution plan of the database

Schema

    estimated_cost integer

    Absolute number that shows the cost of this search. See rating for an interpreted number.

    estimated_nr_items integer

    This is the estimated number of items returned for this search. Please note: it is computed based on search statistics and heuristics and does not reflect the real number.

    available_nr_items integer

    This is the number of available nodes in the graph.

    full_collection_scan boolean

    Indicates, if a full collection scan is required. This means, that the search does not take advantage of any indexes!

    rating string

    Possible values: [simple, complex, bad]

    The rating of this search. A simple search is usually fine. A complex search might be acceptable in certain cases. Maybe an index can be used to make this search simple? A bad search is usually something where an index is missing or the search has to be rewritten.

Loading...